Monday, 15 September 2025

Vintage and Veteran Car dynamo power management

 Changed; 26/01/2026 - 25/01/2026

DRAFT

The mechanical cut-out or cut-out and regulator can be replaced by an electronic rectifier and an electronic regulator.  This blog develops electronics that will also deliver more of the dynamo's available power, not normally derived at lower and higher RPM.  Using a buck-boost switch-mode regulator and software similar to that used by solar power inverters, which uses a trial-and-error system to deliver the maximum available power, usually called Maximum Power Point Tracking.  Once that is done, calibration curves should be created by a rule of thumb, starting with an output of 5 times the dynamo field current at rated RPM, and then fine-tuned continually on the fly.  There is more power available from the dynamo at both lower and higher RPMs that could not be traditionally converted and used to reduce the car's battery workload, shortening its life. 

This blog also discusses the history of electrical systems, cars, and lighting, as well as ways to economise on the power required, so that the original car dynamo can be kept.  I have not made and tested the electronics on this blog.  So, take care with this design suggestion. 

Pitman's motorist library, Morris 8 and The Morris Minor to 1956, 
Dynamo, earlier cut-out and later cut-out and regulator control box, not as labelled.

The regulator and cut-out remained fairly unchanged between 1934 and 1970; the design was as good as could be achieved within the limitations of an electromechanical system, cheaply. 

Cars were once toys that the rich played with, but they became a useful tool, enabling transportation and allowing people to travel for leisure and work.  Cars remain toys, with their vroom-vroom noise and fun changing gears, being their most attractive features.  The noisy supercharged 1920s Bentley is preferred over the faster non-supercharged Bentley classic cars; they are heavy cars that the owner needs to manhandle rather than drive.  All this time, the bicycle was the better vehicle, public transport and walking were good common options.  My 1946 Lenton Sports bicycle was made during a time when bicycles were at their finest.  This blog discusses earlier car electrical systems and ways to optimise a veteran or vintage car that you wish to continue running using the original dynamo.  I have also included additional historical technical details about cars and bicycles. 

By 1890, bicycles and electric vehicles were quite good, and the horses, which suffered dreadfully at the hands of humans, as told in the story Black Beauty, started to be retired from human exploitation. 

‐---------------- The issues with the dynamo electrical system ------------

The battery was not charged at an optimal rate, it had to accommodate swings between being overcharged and running a high percentage of the car's electrical systems when the engine is idling.  This was improved upon by the electromechanical regulator, but resolved with the alternator, which powers all the car's electrical systems, even when the engine is idling.  For earlier cars, another blog is more relevant: Bicycle Dynamo, maximum power and battery charging manager
 
Automotive LED lighting can conserve power, increase the intensity of the lighting, but unbalance the charge rate unless a regulator is fitted.  Without a regulator, the charging rate would need to be recalibrated, and different field series resistors fitted for the medium and high current settings.  This blog proposes the buck-boost regulator and recommends LED lighting, but not recalibrating the original charging system by changing the field winding series resistors in the controller or the dynamo. 

---------------------- BUCK-BOOST SWITCH MODE POWER SUPPLY ------------------------

In this blog, I develop an improved electronic replacement for the cut-out and regulator that uses more of the dynamo's available power and reduces its temperature rise.  The boost switch-mode circuit will provide a little more power at a lower 600 RPM.  Buck mode is beneficial at higher RPM, the regulator can reduce the current and therefore the heating up of the dynamo, also provide a higher current output and possibly increase the life of the dynamo's brushes? Or just provide more power at high RPM.
  • The remnant magnetism in the dynamo is required for it to start generating, but an alternative can use the battery instead.  But in the conventional system, the dynamo also has to produce more than 5 times the power that the field requires to start generating.  That is, the field is only provided with ~20% of the generated power, whereas this circuit can deliver nearly 100% of the dynamo's power at low RPM to the field and thereby continue to operate at a lower RPM.  
  • Conventionally, the dynamo needs to generate more than nothing, but some dynamos need a bit of revving to start if not used for some time.  But setting a high charge and waiting longer should also work, unless it is an electronic regulator, which will require a 1V or 2V threshold to be crossed before it will start.  I have also included some transistorised control box circuits based on circuits likely to have been used in the 1970s, and fewer EMI prevention measures are included than modern automotive electronics have. 
  • The second and subsequent circuits use all the power generated by the remnant magnetism in the dynamo initially to power the field, so it builds up to full generating power quickly and provides extra power sooner.  The dynamo must produce at least 400 mV to start up, but there should be 1V produced.  Web search tells me 2V for a 12V dynamo and 1V for a 6V dynamo, I think at a modest RPM.  At higher RPM, the circuit can operate in buck mode, reducing the dynamo's power waste and temperature rise. 
  • At higher RPM, the buck switch-mode circuit can deliver a lot more power to the car's electrics, but the circuit limits that increase in power for 30 minutes, then reduces to say 0%, 20%, 50% or the maximum set.  The buck-boost circuit is likely to be more useful with pre-World War II cars.

Later Austin 7 dynamo, for example, is rated at 1,200 RPM, and the circuits will produce the same power as the original system.  So, the buck-boost circuit produces more power than the original car's electrics at both lower and higher RPM.  The circuits are designed for 6V electrical systems but have a switch for 12V electrics.  The dynamo will at least run more of the electrical systems' requirements over a wider RPM range with better voltage regulation. 

Some cars with magneto ignition can be hand cranked to start with a flat battery.  A flat battery can be high impedance and therefore will not charge, but the simpler charging systems of those cars which did not have a regulator would therefore supply the battery with a higher voltage that would recover the battery, but any lighting used is likely to pop bulbs and blow fuses due to that over-voltage. 

Expectation:

Graph showing the expected benefits for each
of the circuits AL-0060-??? is not to scale,
and curves may be the wrong shape.

  • Boost Mode:
 The region left of the dashed line is the
 boosted power and lower RPM operating. 

 The solid region is the lower RPM extra
 power operating. 

 The region is filled with horizontal
 lines is boosted power.

  • Buck mode:
 Extends the right of the graph where the
 dynamo produces the same maximum
 current, but at a greater voltage, which
 is used to create more current, but at
 the correct lower voltage is not used.

The exponential output curve does not
 continue above 6V field voltage but
 becomes almost flat 13A dynamo
 output in this case due to core iron
 saturating at a maximum amount of
 magnetism. 
At 600 RPM, somewhere between these two models:
  1. Maximum power transfer.  If the dynamo model were resistive, then it would deliver a maximum power transfer of 6.5A at 3.5V, boosted to 3.2A at 7V, which is 2A for the field winding and 2A for the coil supplied by the dynamo and the battery.  Thereby reducing the battery drain to 1A. 

  2. But if current limiting is the major factor.  In this case, the voltage will still have dropped to 3.5V, but at a current of 13A.  This can be boosted to 7V at 6.5A.  That is 2A for the field, 2A for the ignition, and 2A or so for trickle charging or some auxiliary circuits, such as some of the lighting. 
  • By comparison, without the boost regulator, the output would be 7V at zero current, so there is no current for the field, so inadequate or no voltage would be generated.  The dynamo could not start up or keep running.  When current is taken, the voltage will drop, and the current will drop, but depending on the combination of model (1) or (2), the power output will spiral down to nearly nothing.  But either way, there will not be enough power to run all the car's electrical systems at low RPM. 
At 1200 RPM, the dynamo would deliver a maximum of 13A at 7V or 8V, with 2A used by the field, leaving 11A.  That is the same as the dynamo's specification as the original system.   In which the charge rate would be lower than desirable when the battery needs charging, and all of the electrical ancillaries are in use. 

At 2400 RPM, the dynamo would deliver a maximum of 13A at 21V (28V - 7V).  The buck regulator could deliver over 26A = 13A * 14V / 7V, less 2A for the field winding, leaving 24A.  In this case, the average field current would be reduced so that less power circulates in the dynamo whilst delivering the same or more than the original electrical system would have.  But limit the output voltage to 50% higher, and thereby the buck converter will convert up to 7V, limited to 15.5A in this case. 

The graph shows a small section of the dynamo power output.  The upward diagonal lines shading region is the original power output, and the downward diagonal line region is the buck beneficial region.  The areas to the left are the boost mode region.  The solid region is the boost extra power region.  The graph extends further to the right and has the potential to generate significantly more power.  The electrical system should withstand higher current than originally would have, but for a short time, while the battery demands more current. 
-----------------------------------------------------------

These circuits do not suit a magneto-only car, such as the first car below, but are suited for a dynamo in the second car below, which may also have a magneto.  But the bicycle dynamo project:  Bicycle Dynamo, maximum power, and battery charging manager, may suit the first car below if the magneto has power output for lighting? 

   
The first car is a 1897 Vallée Vis-à-Vis, which is the earliest example I've discovered of a car with magneto ignition, so the operator did not need to carry a dry or charged-up battery with him, other than for battery mode ignition for starting, which this car might have.  Lucus Industries started making magnetos in 1896. 

The second video above is of a Swallow car (later renamed Jaguar cars) body placed on an Austin 7 rolling chassis.  Holden Australia produced a body for the same Austin 7 rolling chassis.  Dixi USA made Austin 7s from kits, BMW purchased Dixi, and the Austin 7 kit was BMW's first car, which they converted and made with metric parts.  Datsun also made a copy of the car, which they may have paid for a license to make it later. 

Both cars' engines can kick back and injure the operator severely or damage the car if the timing is not set to retarded before starting the car.  From about 1930, cars incorporated centrifugal and vacuum timing advance, so kick-back should not inadvertently occur, but placing your thumb on the handle so it faces the same way as your fingers is still important in case kick-back happens.  The technique required was to turn the engine over until the compression cycle was felt, then pull the handle quickly over compression.  A few veteran cars in museums that are kept running injure someone every 5 or 10 years, starting them. 

At this time, many car makers made rolling chassis, and carriage makers made the car bodies.  For example, at first, Rolls-Royce just made the rolling chassis; late in the car's life, the saloon car body was often replaced with, say, a fire engine body and equipment.  But whether it was Britain's best car at times, Rover at least shared that claim, and Queen Elizabeth II owned two of them.  Early Rolls-Royce car manufacture allowed two weeks for making and setting the rear differential gear; after that, the gears continued to get smoother over the life of the car.  Car and bicycle gears, of course, continue to get smoother during their lifetime or wear out, depending on the maker.

Rover made some of the first safety bicycles and used the very efficient Lloyd's roller shaft drive on some of their later bicycles.  There are conflicting claims for the invention of the first safety bicycle; it could have been made in the UK or France.  Similarly, with lighting, many people were involved in the development of incandescent lighting.  The horn, bell and lighting were important for bicycles and motor vehicles, and many people were involved in improving the lighting since the 18th century, see Incandescent light bulb - Wikipedia.  Swann's carbon filament light bulb was succeeded by the Tungsten filament lamp.  A major development was the chain drive, but there were lots of significant developments around the end of the 19th century.  Then lots of improvements in steel making and tempering, particularly after each of the two world wars. 

All Austin 7s had a dynamo, battery, lights and starter motor, plus tools including a crank handle.  But the earlier 1920s Austin 7s used magneto ignition, and the car starts well enough on magneto.  Magneto ignition has a reputation for poor ignition power at low RPM; therefore, cars using it solely suffer from poor starting.  On the other hand, you can start a small engine car with a flat battery on the starting handle.  Later, Austin 7s used the battery, dynamo, with the resistor, contacts, and capacitor Kettering ignition system.   This blog mostly covers Austin, but I mention Morris and Riley, for which I have some vintage books for 1930 to 1940 models of these cars and also Villiers motorbike engines.  Villiers used a metal or a selenium rectifier (which might be one of the earliest production uses of a semiconductor rectifier, when a cat's whisker crystal had been replaced by the point contact diode for the radio detector).  I also found information on early Ford cars on the web. 

 Cars like these, made between 1920 and 1970-73, are what this electronic unit can improve the electrical system of.  Because there is no commercial version, I know of, making a unit that has similar benefits to these circuits.  They are expensive to develop, so I have explored the option of using evaluation boards that can be easily adapted and assembled. 

 
The first practical electric motors and generators were developed between 1832 and the 1880s.  Two principles were demonstrated by Benjamin Franklin's electrostatic motor in 1748 and Michael Faraday's electromagnetic motor in 1821.  Practical electrostatic motors are being developed and used now; they are low-speed, high-torque, higher efficiency, lower material cost, but are lower density than a conventional electromagnetic motor. 

Raleigh bicycles, the Austin 7, and Mini cars are all classless icons of British manufacturing in their time.  The Austin cars were technically ahead of many other makers popular priced cars.  Classless up to a point, people were pushed towards cars rather than bicycles, walking and public transport, which they may be unwilling or unable to afford, or drive them, and those systems they could handle became more difficult to use. 

---------------------------------------------------------

The electrical systems that these electronic units can improve on
 

The mechanical regulator

Is a relay, which opens a contact to switch the field winding off when the voltage gets too high.  The snubber resistor across the contact reduces the back-EMF (voltage) across the regulator contact, thereby minimising spark erosion of that contact.  The field magnetism and, therefore, the dynamo output, decrease slowly, even more slowly with a semiconductor regulator that has a diode for snubbing, and the dynamo's field iron will remain magnetised to some extent. 

Therefore, the voltage control is slow, so the dynamo power output would be in excess or lacking periodically.  The battery has to work, absorbing these fluctuations in supply and demand, thereby stabilising the electrical system of the car. 

The contact hovers just slightly open, sparking slightly and continually, but contact erosion and electrical interference are more modest as a consequence.  Electrical interference needs to be minimised, but it won't be possible to eliminate it.  Fitting snubber, resistor and capacitor or capacitor units, and resistive high-tension cables helps.  

The two current sense windings cause the output voltage to increase when the battery charge rate is high to compensate for cable and battery resistance.  The output voltage increases more when the car is using a lot of power, which is faster charging, a less bad compromise than not charging faster and the battery being drained.  These dynamo and regulator systems regulate to 8V or 16V, for 6V or 12V electrics, whereas mass-produced alternators are usually 12V.  They regulate to a lower but more constant 14V, but do so all the time the engine is running, except after vehicle starting, when neither system can meet the battery's demand. 
 
I do not recall the regulator buzzing on my 1960s Mini or my Anglia but continually making small sparks, but I have read that some regulators buzz. 

The car wiring diagrams (left above) are difficult to follow.  I have expanded a section so you can see how the dynamo, battery, regulator and cut-out work.  My father redrew parts of these diagrams using coloured pencils; these diagrams have details about circuits and fuses that do not seem to be in published documentation.  This 1936-39 diagram is more like a picture with smooth curves rather than a modern orthogonal diagram.  Mechanical detail drawings for servicing are better and are very good.  

The basic Austin 7 never got the regulator that the Big 7 and other Austin cars got, but manufacturing of the Austin 7 ceased in 1939.  Manufacturers started including a regulator from 1934, which meant owners could add accessories, but the electrics were still not powerful enough to comfortably run a car radio at the time.  Riley made very desirable cars, but they had not introduced the regulator by 1940.  Ford UK introduced the regulator in 1939, but the earlier cars had the cut-out with the third brush, but not the different dynamo power settings coupled with the light switch, unlike Austin, Morris, Riley and other makes had.

Riley Manual, Instruction and Maintenance.  The diagrams have more black boxes and look nice, but are not so clear.  The book was published in 1952, but it includes pictures of cars as early as 1926, but not such early electrical diagrams.  The electrical section of many car maintenance books is generally the poorer chapter. 

Riley's top right circuit is an earlier magneto ignition before 1932; the other three circuits show the later coil ignition systems.  The last two circuits show two rather than three wires from the dynamo; therefore, two current settings are likely on the earlier cars, perhaps by different field winding tapping settings.  But the field current presumably is changed by introducing series resistors in the field circuit in the later diagrams, though this detail is not shown in these diagrams for later cars.  F1 field wire to the selector switch in the later circuits.  No regulator is included in the change list.  All of these Riley cars are negative chassis, 12V, and they never switched to positive chassis, unlike other makers. 


The mechanical cut-out was invented in 1915, web AI tells me.

The dynamo became available in 1832; public mains electricity was first demonstrated in 1864 in the UK, the web tells me, and public corporations started supplying DC electricity from 1879.  The lead-acid battery became available in 1859.  Ignition systems moved to electrical with a battery and a trembler coil, which works like a buzzer, generating a continuous spark for running or just starting.  A French car that used the magneto in 1897, generating a single spark per combustion cycle.  1907 Delage Type F, the car had a starter motor as original equipment.  In 1911, Cadillac developed and fitted the Kettering ignition system.  The cut-out made automotive battery charging systems possible in 1915.  From 1930, the selenium rectifier was used on Villiers motorbike engines for battery charging, the battery for the parking lights and used in battery chargers.  In 1920, starter motor, dynamo and batteries had become common in cars, and the internal combustion engine started to overtake other forms of motive power, but the bicycle was established, straightforward, and a supreme people's vehicle.  It was possible to park your car outside the shop you were visiting, and places visited were different until 1970, when that romantic era for the car came to an end.

Ignition adjustment was achieved in trembler coil ignition by a distributor which either switched battery voltage to one of the trembler coils or a high voltage distributor directed the spark to the spark plug.  Either the continuous sparks were directed to a spark plug or were dissipated elsewhere, rather than allowing ignition in advance.  In the low-voltage distributor system, the spark was not continuous but was disconnected if advanced further than the setting of the lever.  This means that a car may start without cranking it, but just the trembler coil turned on if the position of the piston is about right.  Later, a low-voltage magneto was included so the battery ignition could be switched to the magneto for normal running.  Alternatively, no battery was fitted, and a high voltage spark was created by the magneto, and the more sophisticated magneto has a contact breaker and capacitor, which may also be called the Kettering ignition system.  Please comment at the bottom of this blog if you wish to correct me.  I have filled in the gaps with guesswork. 

Steam cars were very quiet; some were modest speed, others were very fast, but like petrol cars, they needed a lot of maintenance, although the starting was safer than hand cranking an ICE.


 
The very wealthy's most desirable car might be Howard Hughes' 1925 Doble steam car, which uses a steam generator, but unlike the flash boiler always contains a little water, and similarly only boils enough water to use imminently, unlike a kettle that boils a lot of water.  The difference is that it makes super-heated steam and runs at higher temperatures and pressures, more efficiently than the earlier flash boiler type car.  Nearly turn on and go, it continues to build up pressure whilst approaching its top speed silently of 133 MPH and out-performing most cars of its time.  But this was the last of the steam cars; noise and gear changing were more popular, which the steam car had neither of.  The rear axle is more or less the crankshaft, and the wheels are directly driven by the engine. 

Whites, similarly, used super-heated steam, a steam generator and had a better condenser, so they used less water, and were an earlier and much more successful maker of probably better steam cars. 

Steam cars, Jay explains, could be damaged by steam hammering due to water in the steam circuits not being compressible.   But steam cars that required hours to boil a lot of water had mostly ceased to be made by 1890.  But I suspect scalding by steam was less likely than starting a handle kickback of an ICE?  Similarly, people became wary of being run over by a car, and the fatalities dropped due to crashes, road surface and regulations improved, but fatalities rose due to air quality as more cars were made. 

The difference, though, was that the famous Jaguar racing cars of the 1950s would also have better road handling, so be able to develop similar high speed but racetrack conditions.  Road handling of the Mini was amazing, hence its reputation, but when you owned one, you found the car to be understated; it was even better.  But it was not until the 1970s that many makes of car started to get dual, twin or two braking circuits and very good road holding.  Plus, seat belts, which improve survivability at speed in a crash, arguably only for the occupants of the vehicles. 


Electric cars were the first powered vehicles to be turn on and go. 

Like most technology at the time, it would be very expensive. 

 
Early cars required a lot of work getting ready to go, then cleaning up, and oiling after using a motor vehicle.  Steam vehicles were better, being much safer to start, and after 1890, when out for a ride, could go after 90 seconds or straight away if the pilot was lit.  An electric vehicle has always been fairly straightforward: unplug and go.  Bicycles were at their best between 1930 and 1960 for comfort and ease of cycling. 

Victorian electric cars, the Volks Electric Railway, Brighton, England, have been running since 1883.  I do not have a picture of the rail, but the picture shows where the rail runs in Brighton. 

Milk floats and forklift electric vehicles like this one, which I found on the web, were everywhere when I was a boy in the 1960s and 70s.  But they were a big investment with a good payback in low running costs, but the 1980s were the era of short-term profit or failure, and short-term cheap ICE vehicles replaced them. 

Regenerative braking is simple with a shunt-wound motor; the speed is set by the ratio of the armature voltage and the field voltage(current).  If the field current is increased, the motor slows, thereby applying braking that causes power to be generated.  If there is also a series winding to boost the torque when going up hill, it will need to be bypassed in the over-run condition, or it will reduce the regenerative braking.  There are two types of compound brush motor, one in which the series winding improves the torque under increased load and the anti-series type, which improves the speed regulation and, as a generator, would increase the braking and power generation even more on a steeper downhill run. 

Of all the types of vehicles above, the electric milk floats were used widely, and peak production was in 1970.  Electric vehicles and bicycles were the first to be usable by anyone, though they both needed oiling, and the EV needs to connect battery and be charged.  The soviet union developed many electric cars, but then made very few of them.  Czechoslovakian electric trams were used worldwide. 

There were also a small number of hybrid electric, internal combustion engine, and battery cars made in 1896 and1901: Armstrong Phaeton, Lohner Porsche Mixte, USSR also developed a hybrid car and tested it in about 1947 its fuel efficiency was very good, and despite the limited technology of the time managed its systems automatically but this was never put into production anywhere in the world despite shortage and high cost of fuel at the time. 

Electric transmission, but not hybrid, such as the Owen Magnet car, has no battery, and is a more efficient transmission system commonly used on trains and ships. 

------‐‐---------‐‐------------------------------------------

The cut-out and current selector switch. 

Earlier cars, AL-0055-01B, did not have a voltage regulator, but one, two or three current settings, high, medium, and low switch and the cut-out relay.  The cut-out is a rough approximation of the diode rectifiers used in a modern alternator.  Circuit diagram created using OrCAD Capture 17, limited trial version.  The generator starts because of the little remnant magnetism in the iron of the field electromagnets.  The dynamo also had the third brush adjustment of the power produced.

The three-brush dynamo has a simple cut-out only system.  A third brush was used to trim the charge rate and was set differently between summer and winter. 
  • The system is not accurate, and it has a lot of positive feedback, so when the dynamo voltage rises, the field winding voltage rises, and the dynamo would generate even more power when the electrical systems need less power.  The battery plays an important part in stabilising the voltage, preventing the voltage from rising.
    • The dynamos' iron may be sized so that the field winding saturates at, say, 6V, thereby limiting the dynamo's maximum output current, but the voltage would still increase as the engine speed increases, the field current setting and the battery prevent that.  The current limiting I described is true of all dynamos and alternators; it is part of their magnetic design. 
    • The field winding fuse protects against voltage and power from running away at the lower current setting.  In this case, a 1.5A fuse is probably about right, not the higher values I have found recommended on the internet. 
    • I have guessed the switch connection between D, F1 and F2.  By comparison, Ford did not include a dynamo power setting switch before introducing the regulator, but the third brush was also used to change the dynamo power output for summer and winter use.  Probably the best strategy is to always run an earlier Ford Model Y car with the headlights on, using the winter setting except when you need to fast charge the battery after starting the car. 
    • Voltage runaway could occur if the battery has become high impedance and so it is unable to regulate the voltage.  This occurs if the battery is worn out or the acid level has been allowed to drop too much.  But I have also found that if the acid level drops and has been topped up with water, the battery's impedance becomes low, and its voltage is very low, until it is charged up.  Note that distilled water does not mix with the acid easily, but floats on top, can freeze and crack the battery's casing in winter.

  • The cut-out connects when the dynamo voltage is above the expected battery voltage, perhaps 8V, then that current flows into the electrical system, which will enhance the magnetism, keeping the cut-out connected. 
  • The cut-out disconnects when battery current flows back into the dynamo, combined with the battery voltage dropping. 

Some control boxes have a third relay for current limiting; I have not seen one of those, and many car makers never included a third current regulator coil, but I don't know why they would include it because the dynamo current limits anyway.  In any case, lead-acid batteries are best charged with a constant voltage, which is increased slightly when the charge rate is high.  That is what the two current compensation windings on the voltage regulator relay do. 
Austin Cars - guessing the switch's
internal connection.  There are two
or three current rate settings. 
  • Austin 7's later wiring diagram differs in the power setting for the side lights. 
    • When lights and high charge are required, the side lights are evidently designed to be used instead of the headlights.  Otherwise, the headlights should be used. 
    • Side lights are very dim and best only used when parked, as required on a fast road.  That is, it is a bad but necessary compromise to drive on sidelights when the higher rate of charge is required.  This option is not available on earlier cars. 
    • Earlier Austin 7s, other cars and makes, side lights use the medium current setting.  A wire shown dotted is connected between the F1 terminals.  So, there was no high charge rate when any lighting was switched on, and there was no power to spare anyway when the headlights were on. 
  • The driver monitors the ammeter and reduces the charge rate when the battery has recovered from starting the car, and the charge rate has died back.  The electrical system has no protection if the charge rate is left set high for too long.
    • The driver needs to move the charge rate switch according to the ammeter, and what inspecting the battery electrolyte tells him.  Use a hydrometer, or if the battery needs topping up more often than once every two weeks, or if bubbles form after a long drive, you need to reduce the charging rate.  But if there were just one bubble in each cell after a long drive, the charging is about right.  [if I remember correctly].  It is necessary to adjust the third brush on the dynamo periodically. 
    • The manual says the ammeter reading should come down to +2A to +4A trickle charge, which is a necessary high rate to compensate for when the electrics are running on battery whilst idling.  Therefore, a car, as I said, was only suitable for a hands-on person. 
  • The Austin 7 electrics total was up to 70W; the dynamo was rated at 6V, 11A, but most other models and makes were rated higher than this, and earlier Austin 7s may have had a lower dynamo rating.  The battery was 50AH, probably 300A maximum?  Austin car club say 300-400A was the initial current of the starter.  This would be the short-circuit current of the battery, cables, and motor before the crank has movement.  100A, used to be said for starting, it could not be higher for maximum power transfer, allowing for battery wear, when cranking.  The spark voltage was about 25,000V, but the Austin Manual says 6,000V, no doubt with the spark plug fitted. 
  • Web search tells me that after starting, the battery will demand 30-60A, which, of course, is limited by the electrical system capacity of the car.  This is not fast charging but normal charging, in which cable and battery resistance can be compensated for, but without the system's limitation.  That figure of 30-60A is probably a limitation of the car's generator rather than of the battery.

  
When I was very young, from 1959 to 1963, I was taken on holiday 250 miles a few times, to Devon and Cornwall, from Kent in an about 1935 Austin 7 and one or both Morris 8s.  Mum complained that she had to keep kicking in the foot well of the passenger side to keep the electric fuel pump running on one of the Morris 8s.  Later, we went further, but in bigger cars.  My father also drove his Ford Anglia to Paris, a similar distance, with his mates.  Most cars had worm and wheel, or worm and pin steering box; some had rack and pinion steering and less well-developed damping and suspension, though sport bicycles, of the time, had much better active damping due to their apparent tuning with the cyclist, regardless of his weight, but sensitivity to placement of weight on a cycle rack.  Otherwise, the cars have a tendency to wander about on the road like the Austin 7 in this video.  The ignition timing adjustment placed on the steering wheel was replaced by about 1930 with the centrifugal advance and vacuum advance mechanisms within the distributor.  The second video features a much smarter car, a Riley, but it could be an MG or one of the less basic Morris or Austin cars. 

Most parts for cars were made by Lucas Industries - Wikipedia.  Ford used their own branding, though they would have bought in parts made by other suppliers, like all makers do. 

 
The mechanical cut-out: does not turn on or off precisely; by comparison, an alternator provides power even when the engine is idling, and the rectifier diodes switch precisely.  It is also likely that the early 4-pole type dynamo produced power even when the engine was idling, so that the inefficient cut-out only needed to close when the engine was running and open when it stopped.  The two-pole dynamo is no doubt cheaper to make and replaced the four-pole dynamo in the late 1920s. 

The regulator and cut-out:

The regulator, a higher-rated dynamo with forced air cooling and Morris switched from cable to hydraulic brakes in the 1930s.  These reduce the frequency of maintenance and improve those systems.  Battery life increased from 1-2 years to 5-7 years (web search).  In modern cars, the battery lasts as long as the car and requires no maintenance, which can be over 20 years.  The earliest regulator was used on a car in about 1929, and hydraulic brakes started to be introduced in 1921.  In about 1935, both Austin and Morris switched the battery polarity to positive earth; this probably reduced cathodic corrosion slightly of the car's chassis.  Riley never introduced hydraulic brakes, switched the battery to positive chassis or introduced the regulator until after 1940; eventually, Riley became a badge used on some BMC car models. 

AL-0055-02B veteran car regulator and cut-out, wiring diagram.  The three-phase alternator, electronic regulator and rectifiers replaced the dynamo, the electromagnetic regulator and the cut-out soon after 1970.  Many more circuits were fused, and relays were added for the high-power headlight circuits.  The component and circuit values differ depending on the car. 

The addition of the regulator improves the battery charging management.  The reduction of four to a two-pole dynamo, which probably occurred in the late 1920s, no doubt meant that there would be more periods when the battery is doing work running the electrical systems and the cut-out, which is not efficient, needs to operate more often.  I don't know how common the four-pole dynamo was in earlier cars, but it ceased being made for later cars. 

The mechanical voltage regulator is quite good, and the driver did not need or have the option to manage the battery by watching the ammeter and switching the charge rate, but the driver still needs to keep his tow on the throttle when idling in a queue in winter with the lights on.  No doubt the newer 2-pole dynamo was cheaper to make, using fewer assembly operations than the early 4-pole dynamo.  But the alternator has 12 to 16 poles, and a smaller, faster pulley sweeps about 6 to 16 times more poles per engine revolution and runs all the electrics even when the engine is idling by comparison.  The lead-acid battery has been very suitable for vehicles and electric vehicles since 1860. 
  • Auxiliary equipment became more viable, and A valve car radio of the time may require 6V, 5A (guess), which is more than the ignition system, so by adding accessories, the electrical system current balance could easily be very wrong, particularly if the car's electrics did not include a regulator. 
      • Some cars were negative earth, while others were positive earth. A valve radio could probably cope with either, but other cars used 12V, and this type of radio would not accommodate different battery voltage cars.  Heaters in the valves have different current ratings and usually run from 6.3V, but this issue could be resolved. 
      • Car radios could be made of high-quality cast aluminium with compartmentalised sections, a vibratory inverter for the high voltage section, radio and audio sections, and all had 0V bonding to the case by the shortest path at many points.  The valve base connectors had clips.  It would therefore be very expensive. 
      • The electrical system of modern cars has a high degree of interference screening and prevention, but at the sources of interference, such as the ignition, motors, and electronics.  So that the electrical systems of the car do not interfere with the in-car entertainment systems. 
    • The voltage regulator regulates over a range of current loading and allows the battery to charge as fast as the dynamo can deliver current.  But the battery is still used when the engine is idling.  The current winding in the regulator applies compensations;  A small increase in voltage when the charge current is higher,  A further increase when a lot of electricity is being used, as a compromise to somewhat compensate for when the battery is being run down faster when the car is idling. 
    Alternator supplies all the power all the time that the engine is running, including when the engine is idling.  Consequently, now car batteries rarely need topping up and can last at least a few decades, but they still provide voltage stabilisation and some over-voltage protection.  The electrical system must work from 5V to double 12V battery voltage, therefore 24V, withstand 60V and very fast high voltage spikes of around 500-800V. 

    The alternator has no brushes but uses slip rings that allow it to rotate twice as fast as a dynamo.  When introduced, they were rated at 30A or more, but the Lucas ones, like so many things then, were not well-made at first.  They used less material than a comparable but slower-spinning dynamo.  By comparison, a dynamo is an alternator combined with a synchronous rectifier, formed by the commutator and brushes.  The alternator, as far as I could tell, some have no current compensation, so the battery must be wired from the same junction point as all the other circuits on the alternator, regulator. 

    Two-pole two-brush dynamo, thank you for the
    picture     https://www.dynamosdynamos.com/ 
    Using just one field coil, as pictured, rather than
    two would limit the strength of the magnetism
     so limit the power generated.  Later dynamos
    had two field coils, one on each side.  Starter
    motors have four coils and four brushes.

    Dynamo pictured: -

    Notice that the field iron of this two-pole dynamo is solid mild steel, not laminated mu-metal. The switching losses used with a low-frequency switch-mode regulator would be high; consequently, that does not matter, because switching does not occur at full power.  Switching losses would occur at a lower power level during pulse width modulation. The high-frequency choke should address the electrical interference. Unless the normal low switching frequency is changed to a very low switching frequency, successfully?

    Changing the single winding to two windings would increase the power output, and replacing the windings with modern high-power permanent magnets should increase the power further.  In this case, the dynamo pulse width modulating voltage clamp would operate to protect the buck-boost inverter from 40V to perhaps 120V. 
     
    Control box with just a cut-out. 
     
    Control box with voltage regulating, current compensation windings, a cut-out and fuses;

    Dynamo information UK company

    Post World War II cars also included an instrument regulator which used a hot wire and a bimetallic strip to open a contact when it gets hotter, thereby delivering a regulated mean voltage.  It powered the hot wire, fuel and temperature gauges. 

    Dynamo replacement - flash the field winding, to ensure that the field is magnetised a little and with the correct polarity.  That is with only the battery chassis cable connected, and the dynamo field wire F disconnected.  Connect a wire to the battery live terminal and touch it to the field winding F connection for say a minute.  This will create a big spark when you disconnect, which will hurt you if you are touching the dynamo field winding F contact; you have no reason to do that.  Finally, disconnect the battery, then complete connecting the wiring, then lastly reconnect the battery, chassis cable last.  You can prevent the big spark by connecting something resistive, such as a headlamp or stop lamp bulb, across the field winding first; this will prevent compromising the field winding insulation with back EMF generated. 

    The regulator's voltage can be adjusted by moving a screw or by bending metal.  This wiring diagram/circuit with two fuses, but 12V instead of 6V before the War, is about what Lucas Industries made for British car makers after World War Two, but before the alternator system replaced them. 

    Coil winding insulation; pre-war, the enamel insulation gets brittle, cracks and falls off if the wire is moved.  Sometime post-war, probably in the 1950s, the enamel was replaced with a type of plastic that does not become brittle.  That modern enamel insulated wire enamel melts and smokes off, so the wire does not need to be scrapped to remove some insulation to make a termination.  A soldering Iron may just about be hot enough to melt the insulation of a thin wire. 

    Replacing electrical systems with electronics to improve a vintage or veteran car's electrical systems. 
    Three brush dynamo - The third brush should be disconnected, and the field winding third brush wire connected to the chassis, thereby setting the power to its maximum and removing one brush from the system.  Take care of what the brush mounting spring does without a brush fitted.  I think the spring without a brush might touch the commutator, which would be a bad thing.  Alternatively, set the third brush to its maximum setting. 

    The electrical system then needs to be switched to negative chassis if it is not already.  That is the battery turned around; any electrical systems that are polarity sensitive, the ammeter and the ignition coil should be turned around.  Circuit AL-0060-04 option A fitted can reverse the dynamo and start generating.  But the other two circuits require the dynamo to be flashed.  Circuit 01 must not be connected, but circuit 02 can or should be connected to the Chassis and Field.  It will then stop a high voltage spark from occurring when you disconnect the battery from the field contact of the dynamo. 

    A college tutor would say, "If all else fails, read the instructions" Installing the unit could still go wrong, and using the battery recovery mode can break parts on the car, so the default is best set not available, which is by placing a wire link initially. 

    Switch from a 6V to a 12V electrical system.

    Circuits -02 and -04 have settings to accommodate three of four combinations of 6V and 12V dynamo and battery voltage.  12V dynamo to 6V electrics is not an efficient option without adding a boost supply for the field windings, so it is not included. 

    Dynamo and battery charging; The boost only switch mode regulator diagrams are PDF AL-0060-01, and AL-0060-05 use electronic components rather than a microcontroller, which shows the software functions that need to be re-created.  The second and fourth sets of circuit diagrams, PDF AL-0060-02 and -04, I have not chosen a suitable microcontroller, but many modest speed 48-pin types will work well.  It is best to avoid using a multitasking operating system, but write software and interrupt handlers.  The second circuit, the buck-boost regulator, is the best option, but making PCBs with provision for a small number of different components for either 02 and 04 circuits could be included on the PCB, but not fitted, which of those different components, depending on which circuit you choose to make. 

    The final version AL-0060-04 is almost identical, but for some unintended differences in where the Vboot+ net is either rectified dynamo output or after the buck-boost or from the battery rectified.  That is, after splitting -04 away, they then became more similar as I continued to develop both, plus I discovered the better SPV1040 start-up boost IC.  Buck-boost ICs vary, but MAX20048 seems to be the most suitable; LM51772 may also be at least as good, but I can get no support or model to evaluate it.

    AL-0060-01A dynamo manager, Boost Regulator to extended low RPM operation, this is a 6V analogue partial solution.  The current sensor at the dynamo is used for maximum power transfer optimisation.  In this circuit, the setting has to be found by trial, measurement and adjustment.  That current sensor is also used to provide battery cable resistance compensation and switch-mode power supply current feedback.  The boost circuit requires a little higher RPM in order to start up, then it will operate at a lower RPM than the dynamo system would normally.  CADSTAR 18 used.  The circuits are draft, and some of the component values are incorrect.

    The circuits have screening and over-voltage protection measures.  With or without these circuits, the car should not be run without the battery connected; the battery stabilised the voltage in both cases.  The changeover relay selects between the field winding power source, the PWM regulator or direct connection to the dynamo for starting power generation.  This will start without the battery, such as with magneto-electrics and would introduce a more certain power generation start-up because there is no semiconductor voltage step to overcome before the field winding is powered. 

    Circuits AL-0060-04? and higher should work without a battery fitted, although they are not designed to work that way, and their output will drop with load increase briefly. 

    Electronics mostly start up above 3.5V, but the dynamo may produce less than 1 or 2V from the residual magnetism, so the field winding is connected by a relay to the dynamo initially.  When the dynamo is generating more than about 3.5V, the Field winding current is controlled by a pulse width modulator (PWM) to regulate the dynamo output voltage. 

    AL-0060-01A Field winding power and dynamo regulator. The field is turned off whilst the dynamo output is overvoltage clamped, therefore when the clamp is lifted, the field is briefly turned on, though it should not be until the dynamo voltage has had time to resettle again and be measured. 

    This does not provide any more power for the car's electrical systems at high RPM.  At low RPM, the dynamo will not start producing power, but as long as the engine has been running at a moderate RPM and the dynamo has started up, it will continue to produce a little power at a lower RPM with this circuit that was not provided by the original electrical system.  In addition, this circuit clamps the dynamo voltage using 10mS PWM so that the battery is also not severely overcharged.  This still does not resolve all issues, and using LED lighting should help. 

    Because of the relay, the dynamo will start up even with a flat battery in a car with a magneto.  In case of a faulty wiring, the fuse should blow before any semiconductors are broken.  The field should be flashed with the cable from the unit disconnected, this is different to circuit 02, -04, and -05, but the same as the original electromechanical system.

    But because the battery is connected to the output, if it were to become flat and low impedance rather than high impedance, this will stop this unit from starting up; this could be a safety feature or a disadvantage depending on the reason.  This is the same for all these designs, but differs from how a car with an electromechanical cut-out would behave. 

    Summary of this circuit version 01

    This circuit does not start generating at a lower RPM, the extra power region I describe in the graph further below, but it will continue to generate power in this extra power region once it has started generating at a higher RPM and continue to generate until the RPM drops to such a low RPM that not enough power is produced to over come all losses and deliver 7V to the field windings, with the boost SMP running. 

    The overvoltage protection, which will operate a lot when speed increases or when lighting is turned off, will put the input capacitors under more stress.  By comparison circuits -04? (buck-boost) and -05? (boost) have a separate power recycling transistor to put the excess power back into the dynamo. 

    There are, of course, losses in the system, so the charging system will stop producing enough power for itself when the RPM has dropped to where the dynamo cannot produce more than 100% of what the field requires.  This is better than the original system, where the field only ever got 20%, perhaps more of the total power generated, so the dynamo would have stopped generating at a higher speed. 

    This circuit does not have the buck converter that recovers more of the power that is normally returned to the dynamo at higher RPM.  So, with all electrical systems running, the battery would be trickle-charged, not the usual fast charge that occurs after starting up without the lights on and high charge selected or what the regulator does.  Changing the lighting to LED lighting may reduce the power used and resolve this issue with this circuit. 
    ----------------------------------------------------------------------------

    Replacing filament bulb lighting with LEDs;

    Using LED lighting will boost the light level and reduce the current taken, but an automotive unit can include a power-wasting load, so that the unit matches the load of a filament lamp unit. 

    • You may be able to remove the dummy load, which is a bank of resistors.  The electronics should still operate properly.  That is, if one LED fails by breaking the open circuit, then all the LEDs of that circuit are turned off, which would, on a modern car, illuminate a dashboard warning light.  But you may need extra electronics to pulse-width-modulate drive the red stop/tail light function.  The PWM is either 100% for the stop light function or 5% for the taillight function running at 100Hz. 

    • The tail light function can not be created using a series resistor; even if the LEDs were to illuminate, their colour may not be legally correct.  On the other hand, all 12V car systems, including the lighting, should work down to 5V but not at their correct intensity. 

    • A 5W LED can be equivalent to a 50W filament bulb.  An Austin 7 was fitted with 18W headlamp bulbs.  Other bulbs are 3W but would be 5W or 20W for indicator lights now; these probably would not reduce as much as 90% using cheap LED units, and less wasteful units are probably not available. 
      • The LED drive may be by a linear current source, therefore the efficiency could be reduced to 25% or perhaps or by a switch mode power supply, in which case the efficiency may be 80%.  Either way, LED lighting is beneficial for saving power. 

    If, by this change to the lighting, enough power at higher RPM, there is no need to proceed with the buck-boost regulator below, a boost-only regulator may be adequate.  You could try it using one of these boost regulator ICs demonstration kits;  
    • MP3433, 22V, 20A, no evaluation kit, will start up from 0.8V with an external 2.8V to 22V.  It also has a high-side low-loss diode function.  Unfortunately datasheet is not available without an appropriate company name.  The diagram on the website shows two audio inputs, CHL and CHR, that increase the output voltage when sound is detected. Monolithic Power Systems
    • MP3435, 19A, 22V. There is an evaluation kit, 3V to 20V, with a high-side low-drop diode function. 
    • Or MPT612, NXP includes MPPT software for this microcontroller. 
    • Most boost regulators or a microcontroller plus an SPV1040 to provide start-up power. 
    Limited function, minimal circuit; AL-0060-05A (PDF) Boost SMP evaluation board.  This has a much better overvoltage clamp, which does not put a high ripple current into the capacitors.  But it is only rated at 20V, so it may fail on a 12V system, but be just about okay for a 6V system.

    --------------------------------------------

    Buck-Boost Switch-mode power supply to give more of the available power: -

    The Dynamo's power output can be at least tripled at high RPM and start producing power at a lower RPM without having to have been running at a higher RPM prior, unlike circuit 01.  As in the previous circuit, the power available remains the same at the stated dynamo rating and RPM. 
    • This is not more power from the dynamo, but instead of returning potential extra power back as motive force, so that the dynamo motors a little, that power is converted to a useful voltage and current.  That is ordinarily, the turning force to the dynamo would have dropped with increased RPM, and the power taken remains unchanged.  But the dynamo voltage can be allowed to rise, the dynamo current can be reduced, and hence its power dissipation is reduced and/or more output current delivered.  If the dynamo voltage rises to 30V, it will be clamped by 10mS pulses shorting the dynamo, which then returns the power and makes the turning force drop.  Otherwise, the dynamo could be run at say 50% more than the battery voltage, 10V or 20V. 
    • Whether a unit is home-made or manufactured, this would be an after-market product, so it does not need to comply with each of the different manufacturers' own standards that they impose on their suppliers.  It needs to work with double battery voltage, 12V on a 6V or 24V,  on a 12V car electrical system. 
    • The voltage control response time is slow because of the magnetism stored in the iron, made slower by the diode across the field winding, so the dynamo output crowbar will necessarily do more work, consequently.  That is, the crowbar will operate normally when driving the car, and auxiliary circuits are turned off. 
    • But the average field voltage must be limited to the original rating of the dynamo, 7V.  The D45D dynamo, which is rated at 11A, is the figure given by Austin at 1,200 RPM.  0.5A for the third brush adjustment range, and 2A for the field winding.  So the output current from the armature (D) is 13.5A.  At 3600 RPM, 3x the specified RPM and outputs, the dynamo's output voltage can be allowed to rise threefold above the specified output voltage.  
      • That is, the voltage would be at least 21V = 7V * 3.
      • That is, a buck converter at least; 34A = 13A * 3 - 2A, which is much too high. 
    Other models and car makes used up to about a 125W dynamo; some were 12V.  Post-war, 1945, dynamos increased to 400W, and bigger cars had higher wattage dynamos.  Subsequently, alternator power and more electrically driven systems have caused the power budget to increase.

      Option (2) Nexperia - simplified buck-boost converter. 

      The buck-boost circuit version 2 can.
      1. Run a 6V dynamo and 6V car electrics. 
      2. Convert a 12V Dynamo's output to 6V.  But the output current would be half its expected value until the dynamo reached over 2,000 RPM without circuit option A.  And the extra power at low RPM is not available. 
      3. Increase the 6V dynamo to 12V electrics. 
      4. Run a 12V dynamo and 12V car electrics. 
      Using the microcontroller and software buck-boost converter is likely to have poorer efficiency than using a switch-mode converter IC, which will also include the four MOSFET drivers and MOSFET switching tuning to minimise switching losses. 

      AL-0060-02A dynamo manager, Buck-Boost Regulator to extend the low RPM operation - 6V or 12V or combination solution and fast charging with lights on, if necessary, at higher RPM.  CADSTAR 18 used.  Some of the components and their values are not resolved yet. 

      This is not quite an interleaved dual-phase, but it is dual-phase, and it shares some of the reduced EMC benefits.  Another sheet has the second phase; it is a partial duplication of this sheet.  A better EMC profile may be achieved by using one of the Buck-Boost ICs with extra features for EMC and efficiency.  The microcontroller is still used to optimise the power transfer and the field PWM control.  There would be benefits in changing to GAN MOSFETs in higher operating frequency, reduced gate switching noise, and reduced PCB size.

      Parts of the Field drive circuit sheet are similar for both the boost-only and the buck-boost circuit. 

      It took a while for me to work out this function, for which I have a few good, simple solutions. 
      • The Bipolar Junction Transistor (BJT) solutions have less protection, but it should be enough normally.  A variant need not require a battery, but won't start up unless the remnant magnetism is enough for the dynamo to produce 1V. 
      • The electromechanical relay will blow fuses if badly wired, but it is a better solution, starting up from a fraction of a volt out of the dynamo.
      • Two MOSFETs back-to-back resolve most protection issues.  It may run from Dynamo power or battery power via the ignition circuit, but in any case, the field winding power comes from the dynamo's remnant magnetism in the field iron. 
      • Boost regulator is good because a few solutions will start up from about 1V and then power the main buck-boost boost regulator.  But the best is SPV1040 solves most issues because it will operate from 300mV to 5V, and will power the microcontroller and will start up from <400mV from dynamo remnant magnetism.  That is, it should work reliably with a lower voltage 6V system.
      That is, if the battery is flat, the dynamo should start generating if the car uses magneto ignition.  Consequently, the dynamo will start up and run somewhere in what I describe as the extra power region of the graph, further below. 

      This circuit 02 has now been developed and is probably the best of the three circuits.
      • The cut-out function is more precise using a rectifier diode than circuit 04, which uses more parts and needs a lot of software development to do the same. 
      • It will start generating power at a lower RPM or start up more quickly. 
      • This circuit, by default, does not remagnetise the field negative chassis, unlike circuit 04. 
      • If it is necessary to reverse the car's electrics, including the field windings, the unit should be connected to the chassis and the field first, at least so that the snubber diode protects the field winding and the installer from a high voltage spark when the field is disconnected from the battery. 
      This circuit is better than circuit 01, and circuit 01 should be put aside. 

      AL-0060-02A There is now just one field driver; it uses power produced by remnant magnetism in the dynamo's field iron, which is boosted to a usable voltage for the microcontroller, main boost regulator gate drive and the field driver switch.  The start-up is different to the first circuit, because all the available power is used to power up the field more quickly, and the field PWM is set so that the field just gets up to the maximum available.  It starts up in the extra power region when there is adequate power available in the dynamo.  

      This circuit is reasonably well protected and should stand flashing the dynamo and prevent arcing that would compromise the winding insulation slightly.  The 5V0start net powers the uP and then contributes to the power for the gate drivers.  The Field winding PWM is set to take less than what is available.  The overvoltage clamping function protects the circuit.  Otherwise, the overvoltage clamp function should not be needed often because the buck regulator mode accommodates some variable voltage.  Reducing the PWM field frequency will reduce losses in the iron in the field, and the buck-boost can accommodate the fluctuation, provided the field current is set to produce at least enough power. 

      st.com community advise me that I can reduce R1 and R56 to 2 K.  I have reduced those resistors to 2K2, which improves the gate drive over using higher value resistors surrounding the SPV1040 boosted supply.  I have also added a diode between Vdynamo+ and Vboot+ nets so that the dynamo output can continue to increase, but without using the start-up boost supply or needing to supply the car electrical system at the same time; this diode is not beneficial (this came from circuit -04) when the battery is in good order. 

      No power is taken from the ignition circuit other than to run the status LEDs.  The Red/Amber/Green LED will flash slowly as a warning, not a fault, if the battery deep discharge recovery feature is turned on, but that feature only doubles the battery voltage, which does not go any higher, but at a very low current <0.5A.  So that normally the battery and other circuits, if used, take more power, and the voltage drops to 7V or less if that is as much as the dynamo can deliver optimally. 

      It took me a while to develop the Field Drive function.  Here are some alternatives that may have weaknesses, AL-0060-03A.  A problem arises that the dynamo will not generate unless D and F are connected together at start-up, but there is little voltage generated by the dynamo for the electronics to start up and turn on the connection.  The relay in AL-0060-01? resolves this issue, though it will build up power slowly, just like a car's standard system would anyway.  The electronic boost regulator increases the voltage to 7V, drives the field winding with up to 100% of the dynamo output.  The voltage, therefore, builds up more quickly if the dynamo is rotating fast enough, but it will start generating at a lower RPM than the original system could.  This is the extra power in the graph below: the anticipated advantage of AL-0060-02a, or AL-0060-04a.  When the dynamo is running in its extra power region, it generates a surplus that is delivered to the car's electrical systems sooner and at a lower RPM. 

      A 6V dynamo, based on what the web tells me, a 6V dynamo will produce about 1V from the remnant magnetism in the iron of the field magnets.  This start-up circuit powers the microprocessor and the status LEDs; then surplus becomes part of the Vboost+ net that also supplies the buck-boost regulator MOSFET gates and the field current.  Only the SPV1040 boost regulators listed below will start up at a low enough voltage to be relied on to start up on a car with a magneto and a flat 6V battery.  The complication, though, is that these ICs only produce 5V, so I have included a circuit that only takes pulses of current on the forward conversion cycle of the IC using a voltage multiplier diode-capacitor network, but with an output resistor.

      A flat battery can be high impedance and can be recovered by the deep discharge recovery method by charging it from a high voltage, say 5 times the battery rating, until it starts to draw current, which may happen after many hours.  Otherwise, it may recover after some weeks of charging normally.  Such a feature, if included, would require that it be switched off whilst any electrical system of the car is required, and may be of negligible value to the car user.

      If a flat battery is low impedance, which can also occur if the acid level has dropped a lot, then it's been topped up.  Then, putting some current into it will restore it again.  But if a plate has buckled and shorted, then the other cells will produce a lot of hydrogen, which is dangerous. 
      • Boost regulators that start up from a lower voltage but only work to the lower 5V.  That may have other features, such as MPPT managers.
        • LTC3105 0.4V to 5.5V. may start at (225mV) 1.1V due to #shutdown threshold. 
        • SPV1040 0.34V to 5.5V may start at 300mV (increase 270mV decrease to 140mV). 
        • SPV1050 0.58V to 5.5V <70mA or buck-boost from 2.8V to 18V <30mA has many other features. Boost works down to 150mV after starting. 
        • bq25570 0.6V to 5.1V works to 100mV 110mA includes a battery charger. 
        • bq25505 similarly 600mV start-up, 230mA. 
        • AP3015 or AP3015A  1V or 1.1V to 12V then shuts down <0.25V 70mA or 300mA 38V out. 
        • TPS61200,TPS61201,TPS61202, 0.8V enable threshold. 5.5V maximum, 1.3A inductor current. 
      The MPPT function ensures the boost regulator starts up without stalling, trying with a very low start-up voltage.  st.com calls the MPPT strategy "The perturb and observe algorithm is based on monitoring either the voltage or the current supplied by the DC power source unit so that the PWM signal duty cycle is increased or decreased step-by-step according to the input power trend."  Similarly, I propose for the dynamo management, trial and error, similarly, but also storing the optimum power characteristic and using a combination and introducing a slow ripple, to improve the characterisation and use the characterisation to correct for noise in the electrical system.


      Permanent magnets to replace the field windings. 

      The field could be continually magnetised using modern high-power permanent magnets.  The over-voltage clamp function will then regulate the output voltage using PWM running at 100Hz.   These circuits accommodate this option.  You leave the field winding's output disconnected.   The status LEDs will not indicate correctly, because there will appear to be a fault at times, causing a flashing green LED.


      Final circuit -

      The field windings can be powered from the battery, which will reverse the dynamo polarity as part of how it works.  Withstand flashing the field, although that is not required, will start the dynamo if the car has magneto ignition and the battery is flat; the dynamo should start generating power.   A deeply discharged high impedance state battery should be recoverable on a magneto ignition car, but there is a risk that if used in this mode, the unit may damage other components in the electrical system, so the mode is only available by switch option, maybe using the low charge setting? 

      The design AL-0060-04B encompasses AL-0060-02A, but with footprints for parts that can be added or removed to try features.  That is, the unit will start up anyway, reverse polarity or remagnetise, dynamo,  on battery power, flat battery on an early type magneto car.  Also has an ammeter display via USB charging port, because of the battery power option, which would have otherwise caused a bad cut-out operation.   This solution requires more software development, which could be avoided by adding more parts.

      Selection of buck-boost converter IC or Module:

      Likely buck-boost modules or ICs that may suit with some additional circuitry, that is, most of them need to work off spec. but not out of spec. But the detail is not in the data sheet or available by support request to people without a company email address.  In most cases, the input start-up voltage is below IC spec but easily within the power MOSFET capability: 

                                                                Vin        Iout           Note                                                                1 off
      ISL81401/ISL81601/ISL81801   4.5V  40V/60V/80V   controller                                                        £5.50/£6.50/£9.70 
      Unfortunately, a 5V input to ExtBias pin does not allow the IC to run at lower than the specified Vin 4.5V say 300mV.  The ISL81401A is unidirectional, but ISL81401 can be set to unidirectional (reverse direction can be disabled by tying FB-IN high or low) and, in addition, can operate in parallel with another of the same ICs, interleaved clock, the two parts are pin interchangeable provided the functions not provided are dealt with.  I do not know if the EVBs support these options.  £150 for the EVB can output up to 8A. 

      I have re-read the reply.  Vin must be 4.5V, but if I connect 5V from my boost supply, I have met that requirement.  Then ExtBias can come from Vout, which is 0V to 7V or 14V to increase the internal regulator to its 5.3V.  Significant other I/O can be 40V or 45V without restriction mentioned to Vin's voltage. 

      There are some useful features;

      • LDOs are 5.1V and 5.3V; they are supplied by either VIN or EXTBIAS. 
        • VDD of 5.3V is used for gate driving. 
        • VCC5V is probably cleaner, better for the analogue electronics?
      • IMON_OUT and IMON_IN are good measures of the input and output currents, where 1.2V is the current limit.  They must be buffered, then they can be connected to the microcontroller A/D inputs. 
        • But they don't tie in with microcontroller ports because - the current limit is set by a variable resistor, and the monitoring might be high gain and not really tell you, or it may be 860mV for 0A and 1.2A for maximum current, for -80mV to 150mV at CS. 
      • If required, the 60' or 120' clock options are not available.
      • PGOOD is provided by other controllers and is fairly useful for the generator light drive.
      • EN/UVLO The significant threshold is 1.8V when the other LDOs are on, and the PWM is enabled. 
      • MODE I need to find out what diode mode or CCM means - guessing that diode emulation is the correct setting?
        • This is potentially good news because other controllers operate in a bidirectional way rather than as simple output diode emulation.
      • The two current sensors have a common mode range that includes 0V to +40V.  That is, the IC may work where the MAX20048 model does not work? 
      • Diode mode emulation overcomes issues, if I've guessed right, that all other ICs have in preventing battery drain by the buck-boost converter.  That is, a high voltage won't be created at the input power by the battery in some circumstances.
      The outstanding question is though can the IC run from a lower input voltage with the boost supply added as I said above.


      LM5118                                        3V    75V                     "                                                                   £4.70
      The evaluation boards are only rated at 2A or 3A, although Vin requires 5V to start up, I cannot see any reason to not supply that from the auxiliary boost and VCCX for the transistor drive from 5V to 7V or 14V Vboot with a voltage regulator for over-voltage protection.  I did not get an answer on starting up with 0.5V at the high side input MOSFET D and an axillary 5V at Vin

      There are Tina-TI spice models, but I have not got them to work well enough. 

      Requires two rectifier circuits because they are not integrated into the IC.  Consequently, this controller does not work bidirectionally and won't produce a high voltage back into the input from the battery. 

      It is unlikely that there would be a few mS burst of high current drawn at start up because there are no cascode transistors in the reference circuit. 

       

       MAX20048                 2V (4.5V) to 36V EVB £230, up to 16A input, 5A output.              £6.00

      AL-0061-02A Buck-Boost The LTspice model, if it is accurate, this integrated circuit is different to other buck-boost ICs.  The IC requires 4.5V to start, then continues to work down to 2V input, so it cannot be used to start from remnant dynamo magnetism, which is only up to 1V or 2V.  Once running, it continues to work down to 2V.  and generates its own 5V Vcc for gate drive. 

      LTSpice 24.1.10 for Windows 10 or 26 for Windows 11, earlier 
      version VXII may also be suitable, it will run on Windows 7.
      Available from LTspice groups or analog.com 

      Blue Vi - Dynamo voltage.  The rectified voltage rises above the dynamo voltage to 21V during start up period, 10mS for ~1.7mS, using power drawn from the battery. 

      Red Vout - Battery voltage.

      Pink - compensation pin, I have not optimised the compensation, so the waveforms look thick due to instability.  It turns out that when the input voltage and power drop to zero, this pin's voltage is raised a little by a small current from the battery until the EN enable pin is held low, to stop it taking battery power when the dynamo is not running.

      Grey Ven - enable pin. This pin needs to be pulsed or pulled low to stop the IC from taking power from the battery.

      Green I(Vi) - input current after the rectifier.  An additional circuit simulating the microcontroller reduces the power drawn from the dynamo when the input voltage is 2-3V in order to prevent the controller from shutting down. 

      Emerald Green I battery - output current.  Is briefly reversed after ~6mS.  If the battery voltage is higher than the set point, the current does not reverse in this controller, in comparison with other controller ICs.  There is a small current drawn from the battery when there is no power input until the controller ICs' enable pin is pulled low. 

      To repeat the point - The controller needs to be disabled when the dynamo is not producing power otherwise once running it will continue to draw current from the battery.

      It seemed that the IC cannot be tricked into starting up using a boost regulator from 400mV Vin but it turns out that it takes a long time for the model to simulate and it needs to run for 10mS which in real time is very much longer before it has completed its start-up period.  That was putting 4.5V on pins IN and OUT and also ensuring CSP1 and CSN1 are less than 300mV above IN, the datasheet says VIN but there is no VIN pin. 

      CSN1 CSP1 what is their maximum voltage? - Q&A - Power Management - EngineerZone. these pins must be less than IN pin voltage although the datasheet does not say that clearly.  Input current sense must be below VIN but there is no VIN pin. 


      MAX25431                                  6V    36V, EVB is £120    initially start 7V, 3-25Vout                  £1.80

      This part looks very similar to MAX20048 except that it requires 7V to start and 6V to run.  EVB output is up to 16A.  It therefore probably cannot provide extra power at low RPM though. 

      Adapted working EE-sim oasis model abandoned. 

       

      LTC3789                                      4V     38V      £12, EVB 25A 6 to 36V DC2253A 2 phase interleaved. $229, there is also a single-phase 12A EVB.

       AL-0061-01A Buck-Boost LTC3789 LTSpice 24 or 26 simulation.  See the high input voltage (Red VIN) and high Battery current drawn for 1.7mS (Green I v4) after starting up; this does not recur subsequently.  The polarity of the Vi current displayed is inverted.  I have not optimised the under-voltage prevention control, so some of the currents oscillate, but show the point that the circuit works well enough to handle the function in software. 

      Green Vi dynamo voltage - The IC starts up when the input is <0.5V, provided the VIN pin is supplied with an auxiliary 5V, in this case, from a low-power boost supply.

      Red VintVcc - This is the internal 5V supply, which is switched on when RUN is true. 

      Light Blue Vout - Battery voltage.

      Grey Vcomp - Compensation pin is driven so that the maximum power is taken, or the output current is limited.

      Dark Green. Run - Used to disable the IC if the battery voltage is higher than the setpoint, consequently preventing the input voltage from rising by unintended back conversion. 

      Blue Vin, Input voltage after rectifier - Rectified input voltage rises above the input voltage, potentially harmful.

      Blue Iin+, Input current after rectifier -  There is a brief high current taken during start up, but this is much lower than the standard circuit that runs with a higher starting current.

      Red Id10+, Battery current - high current is drawn briefly from the battery during start-up, and also if the battery voltage is above the setpoint and the run pin is true. 

      The IC used within datasheet parameters differs; AL-0061-01A LTC3789 Buck boost  Diagram  Plot
      • On the other hand, despite the high current drawn, the model does work down to 100mV input.  But I don't know if the MOSFET switching optimisation, I presume at start up, would be compromised? 


      LTM4712                                     5V     36V   12A(Vin or Vout)         £44
      Works with low voltage input, provided that an auxiliary 5V power supply to VIN and say dynamo input to EXTVcc.  What I found in the LTSpice model differed from how I understood the datasheet.
      Adapted working LTSpice model and Input current LTSpice the models consume a lot of current (20-28A) as it starts up, briefly probably so that the power transistors drive is optimised.  This current is limited in the simulation with input current limiting control, and also greatly limited due to the very low input voltage from the dynamo.  There is always a current pulse taken for a millisecond from the battery then subsequently the cut-out function prevents that occurring.

      EVAL-LTM4712-A2Z four modules 48A for £160, 

      EVAL-LTM4712-A1Z two modules 24A out.  £142 + VAT. 

      These EVBs - may not be possible to wire them so that one current monitor is connected to the input and the other connected to the output.   I don't know if it would be possible to power VIN from auxiliary 5V and power EXTVcc from the input supply?   I have an outstanding support forum question at: LTM4712 Is lower voltage operation possible with EXTVCC supplied by 5 to 10V - Q&A - How to Use EngineerZone - EngineerZone 

      The field winding cannot be taken from SW2 pins and rectified, but the pins are output so that an snubber R+C can be fitted.  (TP15, 28, 36, 44). 


      The LTSpice model AL-006l-03A LTM4712 none of the models complete operation other than the original model, and that has very high current >20A taken during start up for 1 or 2 mS.  This circuit uses output rather than input current limiting, is shown above; it works better than other models.  All models begin to start up with 0.4V plus a low-power 5V supply connected to INTVcc.  In other words, I cannot trust that this module will work properly. 

      Starting the IC from below specified minimum voltage results in the model behaving differently.  In addition, these models ran too slowly to test the full cycle; start, run, turn off.  Question to the support forum with a company address me get an answer, and using an EVB would confirm any answer given. 

      The input current limited models look better, and even better when the soft start is increased at least 10-fold, but the power take clearly needs to be software managed using the input voltage and the current limit to get the best power transfer.  None of these models work so they have not been refined to test all that needs to be tested;

       

      LM51772 is suitably wide -40 to 125 °C.  0V to 55V with 3.5V bias I2C interface £7.30, EVB available for £212. 

      or LM51772-q1, which is automotive grade.    

      • Current limit is settable between 5mV and 70mV with 1% accuracy and 2.5uV offset, 2.8V to Vin.  Consequently, it can not be used to limit start-up current, but it can be used to manage cut-out function and to limit the output current to 2.5A to 35A, say.  For starting up at a low but maximum power, the COMP pin should be driven by an open-drain variable voltage clamp.
      • Cable compensation may be useful from the optional current sense input if it is placed on the output, which I guess might be the only place where current sensing will be useful.   But I don't think it can be set by programming through the I2C bus, but a resistor has to be used therefor the function is probably best not used.
      • BIAS alternative power for internal circuits and Gate Drive, said to be from 3.5V, but power MOSFETs need ~5V drive. 
      • Vin sense 0 to 55V, Vout sense 1V to 55V.
      • Under voltage range falling from 2.8V to rising 3.5V 
      • Output discharge - may not be possible to turn off? but can be set to 25mA, the lowest setting *** this needs to be resolved? ***
      • This controller is bidirectional - I do not know the best way of inhibiting this function but disabling it should work because it should also start up on low input voltage easily enough.

      These do not work down to 0 Vin. 

      LM5177, LM51770, LM517701 minimum 2.8V if Vias is >3.5V.

      LM251772 Bias or Input start up 3.5V to Vmax.   0-70'C so it is not suitable.

      LM5118 does not work below 3V and requires two rectifier circuits. 


      Microcontroller   Software needs to be developed to: 
      • Input crowbar dynamo over-voltage protection - optionally, there is hardware doing this function.  But the power generation needs to be managed so that the OV mostly does not need to operate, also by managing the field PWM. 
      • Power the field winding and the power supply from the battery, but ensure that the dynamo is generating enough power to power the field winding.  That is the circuit that doesn't include a diode to the battery, which must use Field and Output current monitoring to provide the cut-out function and also control power saving modes. 
      • Field PWM and output power must be managed to use the optimum power transfer.   
      Plus, in addition, veroboard to add fuses, an input over-voltage clamp circuit Field Winding drive by pulse width modulator (PWM) circuit and a low-voltage field start-up circuit.   The circuit below is the most refined, although there is work to do resolving issues and replacing parts that are placeholders with more suitable parts. 

      Circuit with options:

      AL-0060-04B This diagram shows the input and output power and switch options.  It is similar to the other circuits.  This circuit will use one of the buck-boost controller ICs: MAX20048, which requires additional circuit functions, but fortunately, the power circuit to the battery is not a compromise. 

      Not started, draft and incomplete controller IC circuit variations:
      AL-0060-04B buck boost ISL81401 
      I don't have a model for this IC, but it includes most of the function blocks that I have added to the other ICs.  This IC requires more discrete part to do the same probably as the other ICs and does not clearly offer current monitor and set inputs and outputs that it seems to offer which the LTM4712 module does.

      AL-0060-04B buck-boost LM51772 
      Looks promising and conveniently can be all microprocessor control via the I2C bus.  It is also rated from 0V if an auxiliary PSU is included to power the IC and gate drive.  No useful model available, and the datasheet is hard to understand. 

      AL-0060-04B buck-boost LTC3789
      This IC requires a diode before the battery, but this can be put beside the top right MOSFET, and the top right MOSFET's regular output is used to power the field and the boot circuits.  So, it works out well, the input voltage can rise much too high unless the part is disabled when this starts to occur. 

      AL-0060-04A buck-boost LTM4712 modules
      The startup current is high, but it will probably accommodate the dynamo limiting that current.  Similar to LTC3789.  Using two modules wired 180 out of phase sync, a current sense limit and monitor can be placed at the input and the output.   Drawback with this part is that the input voltage can rise very high, but I don't know if driving RUN to turn the modules off will protect them or let the circuit start up.

      AL-0060-04B buck-boost MAX20048 
      Similarly, the output current also needs to be monitored, and the field turned off before current is taken from the battery.  It draws a lot of current for a 0.5-2mS during start-up, more than the other ICs. 

      These circuits do not operate in multiphase, so the current is set lower at up to 25A, which is more than the wiring is rated for in a small pre-1940 car or would be needed if LED lighting were used instead.  If more current is needed, then the dynamo would be substituted by an alternator, and the cabling would need to be uprated.  In any case, the existing wiring will only accommodate a higher current for a few minutes. 

      The connectors should differ, and the power connectors should be pluggable and screw-in types.  The LED should be an assembly with a dashboard-mounted unit, cable and connector.  A junction box should also be offered that will replace the unit with the old cut-out regulator and fuse box units, accepting the same connectors, but for the generator lamp, which could be left in place regardless of which new or original system is used. The junction box should also drive the Red LED, as the filament lamp would have worked, that is, have an anti-parallel diode and series resistor to make it compatible, provided the system is not switched to a positive chassis.

      PCB switches;
      • The switches set -
        • 6V Dynamo, 6V Battery
        • 12V Dynamo, 6V Battery.  The power output would be half expected until the dynamo reached more than 2,000 RPM.  
        • 6V Dynamo, 12V Battery 
        • 12V Dynamo, 12V Battery.
      • The up and down buttons set the battery terminal voltage to compensate for the cable resistance.  So that the battery terminals' voltage is 7V when the dynamo meets the battery's need, when the generator light is green. 
      • The last switch could be used to offer a simple RED generator light.  Except under a fault condition.
      • Output current limit.
        • 100%
        • 120%
        • 150%
        • Maximum

      Dashboard.
      • The High charge (F2) or Medium charge (F1) Contact and pulled up to B+, then the low charge is not selected.   Low charge could be used to turn on flat battery recovery mode for a low or high impedance battery fault.   Otherwise, wire one of the unit's connector terminals to the ignition, and the feature will be turned on.   That is to bypass the warning and force battery recovery mode. 
      • The dashboard generator LED.  Flashing slowly indicates set to flat battery recovery mode is running. 
      • RED - Not charging, 
      • AMBER (GREEN and RED) - charging but not meeting the need. 
      • GREEN - Meeting the needed, and the battery voltage is correct. 

      • White or Blue LED - Extra power, buck mode, or extra power at low RPM, boost mode, is being generated.  If an LED is fitted. 
      I have not included another status LED, but a white or blue LED could be added to the dashboard to tell the driver that extra power is being generated.  I am not recommending this option; it adds clutter, but the light could be put on the unit, although the ammeter advises the driver anyway. 

      Error conditions displayed by fast flashing LED colours;
      1. Dynamo has not started up, so maybe the Field needs to be flashed to reverse or magnetise it. - Fast red LED flashing.

      2. No battery detected,  fuse or wiring fault. - Fast Green LED flashing.

      3. Open or shorted field winding, fuse or wiring fault. - Fast amber LED flashing. 

      4. GREEN and RED - Flashing alternately and slowly, Fault, wrong or low battery voltage setting.  Set to 6V, but a 12V battery was found.  But the cause may be that the battery voltage is very low; in that case, you can use the Low Charge setting to recover the battery by setting the battery voltage and continuing to run. 

      AL-0060-04B MAX20048 The input voltage is clamped using at least 10mS clamp pulses, which return the current to the dynamo, causing it to motor rather than simply waste a lot of power that would have caused a problem and the dynamo to run freely.  The buck-boost transistors' drive is simplified because many functions are integrated into the buck-boost controller IC, which also optimises the power transistor switching. 

      The Field is supplied from the battery supply, but the output current must be monitored to ensure that the y field does not draw current from the battery. 
       
      The field current and the magnetism will probably take 10 minutes to decline enough so that the voltage drops to a safe voltage for the circuit.  So, an overvoltage clamp circuit is included. 


      AL-0060-04B MAX20048 Buck-boost controller.  The compensation pin is managed by clamping to achieve maximum power transfer and to limit the output current.  This diagram shows a hardware dynamo overvoltage clamp, but this function could be carried out in software. 

      A cut-out function needs to be included in the software.  But the circuit will likely get by well enough with just input current sensing.

      MAX20048 The buck-boost controller offers some low-voltage operation, but needs 5V to start, which the relay provides.  It will start up from remnant magnetism.  It could also start up using battery power if software is written to let that happen. 

      It turns out that once the buck-boost has started up with at least 4.5V, then it can subsequently continue to work down to 2V input.  The relay connects the dynamo to the field winding, therefore starting the dynamo conventionally.  After that, all the time the dynamo can produce at least 2V, the extra power at lower RPM is delivered.  Therefore, the second low-power boost regulator option is not offered.  There is little benefit in using battery power to improve the start-up time.  
      • The microprocessor placed is not the optimal choice, but it shows that a 48-pin microcontroller should just about fit. 
      • The unit should withstand starting with jump leads to another battery, but not necessarily to a high-power fast-charging unit.  An inappropriate connection to the field or the dynamo without blowing the fuses. 
      • The unit starts up with a fraction of a volt generated from remnant magnetism in the dynamo.  If the battery is flat, it should start on the crank in a car with a magneto. 
      • Provided that the engine continues to run, the buck-boost will continue to run and can provide a small but extra power that the dynamo would not have produced originally. 
      • The extra transistors and circuitry for overvoltage are an improvement in circuit 02 and are similar to AL-0060-05A 
      • MAX20048 draws current from the battery at start-up, which causes the rectified input to voltage rises to ~24V, but I do not know if the voltage could rise enough to destroy itself and other parts? 
      The low charge switch on the dashboard, if the wiring allows, could be used to select remagnetise the battery if required. 

      Circuit -01, where more power is required only at low RPM, the boost only option would suit, but circuit -04A modified.  Or consider using circuit -05A.

      There are similar versions of this design, one: this AL-0060-04B use LM5118, ISL81401, MAX20048, LTC3789, and up to four LTM4712 modules, which are controlled with analogue and logic level signals or LM51772 buck-boost, which is controlled by an I2C bus. 
        
      LTM4712 modules. Use an evaluation board or complete my circuit. It may have the highest power output for the lowest electrical interference profile, but that mostly depends on the PCB design. It would be difficult to place the BGA package modules, though I have known people who manage it. The cost of having the boards made and assembled is many thousands of pounds for a small number.  Wilson Process Systems will do a super job for you and see it through any subsequent fixes, also some of the Chinese companies will do the job very well and more cheaply. This option may have the lowest noise profile, which is little benefit when a car of this vintage most likely will have a very high electrical noise profile.

      I've used Wilson Process Systems, which can do the smallest detail assembly in the UK and Minatron for many decades before I retired.  Pick a company near your location so you can visit it easily. 

      Make sure that all the PCB footprint origins are in the centre of each part.  Some footprints in CADSTAR, for example, are over 30 years old and use an earlier standard.  Some PCB assemblers can accommodate this; others cannot.

      The safest option, though, is to use AL-0060-04B buck-boost MAX20048 because the IC operates wholly starting and running within datasheet parameters.  In any case, use evaluation boards first.  The other circuits appear to run in simulation, but I do not know how accurate the simulations are off-datasheet parameters. 

      Conclusion.

      The circuit has been developed in stages; I have explained some of the improvements with each stage.  There are still plenty of PCB footprints and refinements to do.  MAX20048 or maybe LM517722 are the best options, but evaluation boards need to be used to confirm them.

      Completing and sending this design for manufacture and assembly would be expensive, costing thousands of pounds.  You also have to take care with the PCB libraries in CAD; the footprints could have been created in the early 1990s and have the origin set on pin 1 rather than as it should be in the centre of the footprint.  Companies can make adjustments for placement origin correction, but not as much as it being at the part's edge or corner.

      A much less expensive solution would be to use an Infineon and other makers' evaluation boards, say 12-60V 10A or 18A single-phase solar panel power supply board and adapt it.  It need not be the newest CoolGAN evaluation board because old technologies are still very efficient.  Then adapt it, which Infinon say can be done using different magnetics and transistors, but I guess two or three sets of boards would do the job with fewer changes to the board?  

      Use manufacturer's evaluation boards for microcontroller, SPV1040 regulator, plus veroboard for the field driver and other functions.

      A cheaper and easier option is to use already available evaluation boards, such as;

      Infineon solar panel optimiser
      OR
      • 15 to 60V needs to be adapted to optimise over; 2V to 8V or 4V to 16V
      • An internal power supply of 12V needs to be generated by a small buck boost regulator to 12V
      • Infineon advise that the wound components also need to be revised.  Though I think you can get by with what is fitted.
      • Other evaluation boards for SPV1040 or SPV1050 in boost mode need to be used.

      --------------------------------------------------

      Magneto for ignition, plus lights, plus battery charging;


      Villiers motorbike engines, magneto's - publications ~1950;

      1. Magneto with two lighting winding types and the ignition pair of windings at the top of the picture.  Pictured is the four-pole type, but a 6-pole with 6 lighting windings was also made. 
      2. Diagram the ignition coil with low voltage and high voltage windings, contact breaker points, and the capacitor.  Plus two coils for lighting, see top and bottom, 
      3. Diagram showing the working of the ignition magnetism, but the HT winding and the capacitor are not shown;  My text, a b c,  explains different aspects of the spark generation cycle.
      a) Contact open, is closing, poles are aligned, so the magnetism is at its maximum, 

      b) Contact is closed, magnetic polarity is changing, which causes a high current in the low voltage winding, and that current tends to resist the magnetism from changing. 

      c) Contact opens, magnetism rapidly changes, the capacitor slows the rapid rate of voltage rise, so that the spark occurs in the spark plug but not the contact breaker points. 

      I do not have experience with a 1920s car, but my 1970s motorbike, which used a selenium bridge rectifier, had no regulator but a switch to connect higher current windings into the circuit when the headlights were turned on, but it ran the battery down if ridden using the parking lights.  The battery was then used in the conventional Kettering ignition system, which also includes a current-limiting series resistor.  That is, it was different to a Villiers motorbike engine.

      Some Villiers engines had a selenium rectifier and a battery for the parking lights from 1930.  Riley was still using a magneto on a few of their cars until 1932.  Austin stopped using a magneto sometime in the late 1920s.  Cars with a starter motor, a battery that required charging, also had a dynamo and a cut-out. 


      The simplest spark ignition uses a moving fixed magnet and a stationary high voltage coil wound on soft iron.  A blunt wedge edge rotating magnet, and a stationary blunt wedge edge iron with the high voltage coil.  Parts pass each other closely, generating a high voltage spark.  The engine, such as a lawnmower, has to be cranked quickly past compression to start. 

       
      Two videos;
      1. This large, low-revving veteran engine has a spring-loaded magnet which was released on a trigger, which moved the magnet rapidly past the coil at the correct angle for ignition.  You might be able to see operating in this video at 6.50 minutes. 

      2. Earlier spark ignition describes the contact break inside each cylinder method.  In the video description of modern capacitor discharge ignition, I have put a better explanation below. 

        The Model T's ignition has a low-voltage distributor at the end of the camshaft with approximately equal sections of insulator, metal contact and a rotating roller contact connected to 0V chassis.  Its body is rotated by a lever on the steering wheel column to retard or advance the ignition, and the spark is generated continually for about half a piston stroke.  The later trembler boxes' contacts have a slight wiping action similar to very long-lasting post-office equipment relays and switches. 
      After-market electronic ignition units made by Jermyn Industries in the 1970s, also known as capacitor discharge ignition, because energy was stored in a capacitor, and the coil was used as a transformer, its inductance was not used to store energy for the spark.  I don't think car makers ever used this system, but they did use the magnetic contactless points system.
      • A magnetic sensor instead of the contact breaker points - this could be unreliable, 
      • A two-transistor saturated core high frequency oscillator and transformer to step the voltage up to about 300V - this circuit could be slow to restart, preventing the unit from working at a higher RPM. 
      • A capacitor was charged and then discharged into the coil, which was now used as a step-up transformer - the discharge was by an SCR, which was reliable.  The current is higher but limited by the leakage inductance of the coil, but could cause a more severe RF emissions issue with a car radio?

      Modern ignition systems probably do not use the inductive discharge Kettering system but have an oscillator circuit driving a small step-up transformer to produce a very short burst of high voltage sparks.  Or it could be a very high current, low voltage pulse and a step-up transformer?  Placed on top of each sparking plug, thereby containing the spark current pulses' path more closely with the spark plug and thereby reducing electromagnetic interference. 

      Non-electric spark ignition; using compression ignition (diesel) and a glow plug, before that, hot tube ignition. 

      Hand cranked starting;  You did not always need to crank a veteran car to start it.  It might start when the trembler coil (spark box) was turned on, but they did break a lot of people's thumbs or arms, hand-cranking them.  One car in the Beaulieu motor museum had injured 10 or so people in over a century of the car's existence.  Another manager spoke of several injuries starting veteran cars during his time at the museum.  

        
      AL-0060-03A Electronic Dynamo regulator and rectifier.  These circuits perform the same functions as a simple electromechanical regulator and cut-out, but there are no current-compensation circuits.  Both can be pulse-width-modulated low-frequency switching or operate in a linear mode, but in that case, they will get hot.
      • The first circuit is similar to an alternator regulator.  Although it has a slow response, it is robust. 
      • The second circuit requires a connection to both ends of the field winding and has a faster response time, similar to the electromagnetic regulator and snubber resistor. 
      An alternator has three-phase windings, A 6-diode rectifier stack for the main power, plus another 3 diodes for the field control and driver circuit, so there would be no output rectifier diode.  But the voltage sense circuit uses a Zener diode and does not have a voltage sense wire to the battery.  The protection would be by a much higher power press-fit mounted Zener diode; the 400mW one shown would not suit, but with input and output chokes added, it may be adequate. 

      An alternator's start-up voltage from remnant magnetism is higher.  Must be 1.2V for two diodes plus 1.4V and a bit more for the Darlington transistor and the bias resistor that is at least 3V.  This half-bridge configuration was not used in early alternator circuits but was something like the dynamo regulator and cut-out circuit further above, but using a Darlington transistor. 


      The Mini Metro was launched in 1980.  Cars had now become what the Lenton sport bicycle already was in 1946, a vehicle you could use with just some minor servicing points to observe and deliver when new without faults. 


      In the 1970s, people bought European car makes because of the better quality, but even though those were not delivered fault-free, the 1975 New Audi 80 now had front wheel drive, dual circuit brakes, and road handling like a Mini.  But my father's car had a temperature gauge that indicated too hot, like others the dealer had, a not worked out two-button control of the two sets of headlights, so off was with one button in and the other out, and both on could be another or the same combination of buttons in and out.  Lovely car, though similar to the BL Allegro, both delivered new with wiring faults.

      ------------

      Earlier steps towards easier travel

      After Richard Trevithick's high-pressure steam coach of 1801 and his steam train, came the velocipede, an early type of bicycle.

      The Karl von Drais of Karlsruhe, the human-powered bicycle, was an early step forward, being a two-wheeled velocipede of 1817 rather than his earlier 4-wheeled, or many subsequent developments that failed until about 1860.  The German eurocoin minted in 2017 to mark this inventor's significant development.  But the Veteran Cycle Club magazine's spring 2025 article does not believe the motivation for the invention was a shortage of horses around the time of the Tambora volcano or a concern over climate change. 



      Steam Coach services started being developed.  The picture is one of Goldsworth Gurney's steam road coaches from 1827, long before pedalled or motorised private vehicles.  The first vehicle was Richard Trevithick's high-pressure steam coach of 1801.  Mercedes Jellinek, whose name was used by Benz for his brand of car, to improve its sales in the USA, according to J Leno's TV show, if I understand correctly.  Pictures from the book, ISBN 0-9510625-0-6.  The first steam car was made in the UK and another in the USA between 1800 and 1810. 

      De Dion, Panhard, and Daimler were also early car makers.  Nicolaus Otto greatly improved the efficiency of earlier internal combustion engines by adding the compression cycle then started manufacturing the first commercially viable engines in the 1860s.  The Otto cycle or 4-stroke engine is unchanged.  The steam train services started in 1825 in the UK.  The hot air balloon travel goes back to the 18th century. 
       
      The first internal combustion engine developer 
      and maker. https://www.deutz.com/en/company/

      18th-century steam engines were made by heating, hammering metal and using a stuffing box filled with cloth and graphite to somewhat seal between moving parts.  These engines were used for pumping water often, but very inefficiently.  By comparison, a bicycle's frame and transmission efficiency had to be, and it is very high.  James Watt developed standardisation and methods of accurately making things, such as making three flat surfaces by grinding the three surfaces against each other in turns, from my school metalwork classes.  That is the wealth that the European nations had accumulated by conflict and exploration, which allowed wealthy of us to finance or develop technology, inefficiently though, found all over the world.  That is enabling the capabilities humans always had. 

      Steam engines date back to about Hero's engine, with a history spanning over 2,000 years.  An 18th-century French vehicle and a 17th-century Chinese toy are steps toward the first vehicles. 

      ---------------------------------------------------------------------

      i

      The designer of the Austin 7 in about 1921 introduced during a TV program broadcast in 1982.  MGs,  Austin's, Maurice's stopped being a cheap renovation hobby for anyone on a modest budget during the 1970s.

      Often, those doing the renovation needed the money back at the end of the project, and the only way to do that was to sell the car in parts.  Consequently, cars became more bits of this and bits of that.  My father and I looked at a few Austin 7s and Maurice's 8s as a restoration project, but what was a heap of rust for £50 in 1970 was £1,000 in 1980 or just not available if it was an MG, TC.  The TC was on dad's wish list, but a Riley 9 would have been nice as well. 

      Of those really nice cars they merged into bigger group's or like Jewett and Bristol they did not last for long after World War 2 as car makers. 

      Other developments

      Bugatti improved its engine's power density and longevity greatly in the 1920s.  This was made possible by higher precision manufacturing tolerance and metallurgy that were not introduced for decades in mass-produced cars.  
      https://www.youtube.com/watch?v=Vu0ffFAmocc&t=1448s

      The USA kept on making cheap gas guzzlers, seemingly carved out of solid iron; petrol or alcohol, either fuel was cheaper there, and the engines took a few hours to make, rather than a few weeks for a Bugatti engine to be made.  They used lower tolerance, but Ford car owners might hotrod their cars, for example, with an expensive cylinder head with better gas flow, and cooling that would double the engine power.  In one US car I rode in, the door was too heavy to open or close if parked on a slight hill. 
      https://www.youtube.com/watch?v=NaXztjI_ayA

      In Britain, petrol was expensive, and Austin or Morris developed a hoped-for better OHC engine, but its power output was poor.  It turned out that the turbulence in the old side valve head was important.  They brought in the designer from Jaguar Cars, who had already improved the Rolls-Royce Merlin engine.  After 1945, companies merged, and the once Austin 7 engine got a new head; the power output doubled.  [Some of the details of this episode I do not have correctly.]
       
      An owner's manual shows you
      how to service nearly every-
      thing on the car.  Car makers
      stopped publishing these, but
      The Haynes Manual replaced
      them also show you how
      to fix everything on the car.
      After 1945, all car engines got the improvements that had started in the 1930s, first in Jaguar Cars, the Rolls-Royce Merlin engine, then the designer went to Austin and did the same on a lorry engine.  The swirling, turbulent gas chamber design, OHV, better fuels, and higher compression doubled the output power.  Engines began to last a much better 100,000 miles.  Companies merged, got big, and management lost connection with the workers; bad industrial relations resulted, and a British Layland made a lot of bad headline news. 
       
      Up until 1979, car ignition contact breaker points would need to have the high point (spike) filed down, then the gap set every 3,000 miles.  Better oils that need less frequent changing, but brake hydraulic parts and fluid are rarely replaced as frequently as recommended, which is 20,000 to 40,000 miles or every two or three years.  Otherwise, cars needed much less owner maintenance and only required management of the gears, which people liked.  Hybrid cars manage the transmission more efficiently.  Now self self-driving cars are starting to do everything in 2025.
       
      When the Mini car became available, it soon became a people's car like quality bicycles made by Raleigh.  Its 850cc engine was still 3-bearing but now OHV, with a SU carburettor, forged crankshaft, and double the power output at 34BHP to the Austin 7.  Owned by ordinary people as well as the very rich.  Nations used to pride themselves on having good services, people took pride in working for a government service, and everything was well-made for their people.  Many manufacturers, including big companies, still pride themselves on making good quality things, but talking about profits is not vulgar anymore.  The Austin 7 cars were well known in sports, and the Mini cars were exceptional in rallying.  And very good sports bicycles used to be made to be used on ordinary roads, whereas they are now uncomfortable on ordinary roads.  One of the differences between a Raleigh Record Ace used in competition and the Lenton Sports is that the crank height is set 1" higher off the ground in the RRA.  Both bikes should be good enough to cover 1 km in a minute. 

      --------------

      Oil companies would send anyone who wrote, asked, booklets and lubricating charts in the 1950's, such as the one pictured right.  Here are some more: Ford, Morris, Riley, and Rolls-Royce.  Most things used to be made to be repairable, and ordinary people would be more empowered by having repaired things. 

      Note: * The Car's life was improved by adding oil and air filters to reduce engine wear. Later improvements to oils also meant thinner engine oil with both better, longer-lasting lubrication and better engine efficiency.  In the 1950s? CC Wakefield Ltd.'s  Castrol brand multi-grade oil became available, so summer and winter oil changes became unnecessary and over subsequent decades, car servicing frequency was reduced to once a year.  The number of types of oil was greatly rationalised at this time, and booklets like these had amendments reflecting the changes.  Many types of vegetable oil had stopped being used early in the 20th century. In the 1970s, improvements to carburation were made using a bimetallic mechanism to redirect air intake from the hot air flowing over the exhaust manifold to get the engine running at optimum temperature quickly, thereby improving efficiency and reducing cold engine wear.  Consequently, a small car's engine would last 15,000 to 25,000 miles, increased to 100,000 miles and now even longer, and engine, de-coking, and reconditioning are not so common. 

      Modern thin engine oil is very good for veteran and modern bicycles, so that you don't need to oil the bike so frequently, as once every 100 miles or every week.  Use grease on the bicycle's cables and slides, and levers rather than oil. 

       
      Riley 9 was a very desirable car with its pre-selector gearbox plus automatic overdrive, so you can accelerate from a stationary position, going through two gears before touching the pre-selector lever, just the selector pedal once.  The car has an automatic clutch, so you only need to use the selector pedal gearbox clutch for hill starting.  The wiring diagrams look nice, but are difficult to follow.  The designer must push the draftsman to make the circuit clear.  That is, avoid 4-wire junctions, lots of wires crammed together rather than grouped by function or labelled at the points they join and leave a bus. 



      My father's opinion of driving a pre-war car with synchromesh was that, as long as you waited for it, it was slow in operation but nicely paced and easy.   Imposing an un-paced, calm driving style, unlike modern life, for which those cars were not designed. 

      There are gaps in history, even during fairly modern times. My father told me that in the 1850s, there were many steam and electric cars around; they must have been just steam cars that were very slow to heat up or ran on rails and a pantograph for electric pickup.  Searching the internet, there only seems to be a replica car made, but I am sure Dad would have read this. 


      Posted by at No comments:
      Subscribe to: Comments (Atom)