Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

vintige car regulator

Status
Not open for further replies.

Thunderchild

New Member
Right the chalenge is on !

I did have a previous thread about this but now I know more of what I'm about we shall start from scratch !

I'm out to design a voltage regulator for vintage cars running on dynamo's my friend has a 1956 Jaguar (yes gets admiration wherever it goes) and other prewar cars he wants regs for. The electronic regs must fit into the original cases with no external modifications.

now as I understand it my working conditions are: keep output at about 14.5 volts (well his runs 13.6 but I say thats a bit low) and the maximum load is 30-35 amps so I want to build for 50 A to allow plenty of margin (apparantly 35 A is rare most are under 30).

Now I have to replicate each function of the old regulator being: voltage control, overcurrent protection and stopping the battery from discharging back into the generator when its output is low. At the same time as I'm building into original cases most of which are for 2 relay regulators so I have to account for how much heat I am generating and I'm not allowed to go putting holes in the original case

now I can easily stop the battery discharging into tha battery by using a blocking diode, I'm thinking again of a 60 ish amp capacity.

I can then control the voltage output by PWM the feild coil, I am thinking of a pic for this, that will produce a PWM output and will sample the output voltage and ajust the duty cycle accordingly.

For over current protection I was thinking of having a tiny series resistor and sampling the voltage over it and at a set threshold over ride the voltage control pulse value and reduse it to reduce the voltage and so reduce current, which might get to the point where it will cut off in the event of a short circuit.

Am I heading in the right direction ? does anyone know anything more and can point out a possible pitfall ?
 
Last edited:
I think that you are on the right lines.

I think that you would be best to use a hall effect current transducer to measure current, as currents of 30 ish amps will cause a lot of heat to be generated if you use a resistor. If you try to make the resistor very small, accuracy will be affected.

The original regulator had a different number of turns for the current limiting coil if the current was going to charge the battery or being used by the rest of the car. I think that the idea was that the battery charging wouldn't last very long so the dynamo could stand a bigger current for that.

I would suggest a relay or a big MOSFET instead of a diode to use for the cut out. A diode will drop about 1 volt so will generate 30 W or so. A relay or a MOSFET will be a lot less than that.

If you are using a MOSFET, the body diode will actually provide the cut out, but when you turn on the MOSFET, the diode will be bypassed by the MOSFET and will reduce the voltage drop. You can detect the voltage direction across the MOSFET to decide whether to turn it on.

You obviously need a freewheel diode or resistor for the field current.

Years ago, I just swapped a dynamo for an alternator and got rid of the control box.
 
Be wary of overvoltage . Jags had a frequent overvoltage problem in that one of the designers must have told Lucas they wanted 14+ volts on their regulators.

They installed this in many versions but had to change a lot in Australia as batteries overflowed acid onto the metal parts and caused a big mess.Bragging, but my new EType had this and I was not happy.

Stay under 13.8 would be my advice , why go any higher?
 
Fords go down to 13.3v in the summer, some cars up to 15.5v in the winter and the usual is 14.4v. I know at 17v the electrolyte boils vigorously and no charging current will flow at 12.8v.

Of course, you want to use a switching regulator with temperature as one of the controlling inputs.

The regulator setpoint, gen. power output capability, driving habits and battery capacity all dance with each other. Bosch has an excellent book on auto elec. systems.

My '04 Subaru has undersized something; on rare occasions the starter cranks very slowly.

The mechanical regulators I knew of had 3 relays; the cutout, the voltage sensing and the current sensing [the volt relay had fine wire and the current relay had thick wire and the contacts of these last two were in series so it was an "OR" function].
 
Last edited:
Try JPcycles.com, they sell the 35 amp "cycle electric" brand voltage regulator for Harleys that use a perm magnet alternator. I have worked with these regulators a bit. They can be bought in 14.5v and also 13.8v models, and operate in voltage disconnect mode so they open-circuit the alternator to regulate, not short it out like most systems.

I'm pretty sure they will work fine from wither AC or DC input, as this type uses a FET to disconnect the battery to regulate, not like th eold ones that used a SCR to short the alternator.
 
I would say that the original and all mechanical voltage regulators can be adjusted the way you want them to behave by tweaking/bending the spring loaded tabs for voltage and for current limiting.
It is done at the factory to be usually 13.8V; 30 A; or whatever the dynamo speed/pulley and ampacity be.

I see no need to design anything new. Just get/draw the mechanical regulator schematic/adjustment instructions from the manufacturer.

Miguel
'61 Caddy
 
I would say that the original and all mechanical voltage regulators can be adjusted the way you want them to behave by tweaking/bending the spring loaded tabs for voltage and for current limiting.
It is done at the factory to be usually 13.8V; 30 A; or whatever the dynamo speed/pulley and ampacity be.

I see no need to design anything new. Just get/draw the mechanical regulator schematic/adjustment instructions from the manufacturer.

Miguel
'61 Caddy

currently replacement regs are made in india cost £ 70 and break after 6 months to a year so I have a fairly large budget, my only worry being to keep heat down in the unit.
 
Last edited:
Just to recap the functions of the original regulator (correct me if I'm wrong)

1) connect generator to battery when generator supplies more volate than the battery,
2) Regulate the output voltage (by controlling the field),
3) protect against overload.

I can easily use blocking diodes to curry out the function of "cutout". I can regulate the voltage by PWM the field coil. I'm not sure on current limitation yet though, I can easily cut the power by turning off the field coil but I'm uncertain of the method of detecting the overload and then what to do about it as I can just cut power (but for how long ?) or i can lower the voltage (reduce field duty cycle) until the current is under control and keep it there. I could then have a warning light to signal under voltage which should make anyone looking for fault understand that there is a short, at the end of the day if there is a short the battery will supply most of the power
 
Just to recap the functions of the original regulator (correct me if I'm wrong)

1) connect generator to battery when generator supplies more volate than the battery,
2) Regulate the output voltage (by controlling the field),
3) protect against overload.

I can easily use blocking diodes to curry out the function of "cutout". I can regulate the voltage by PWM the field coil. I'm not sure on current limitation yet though, I can easily cut the power by turning off the field coil but I'm uncertain of the method of detecting the overload and then what to do about it as I can just cut power (but for how long ?) or i can lower the voltage (reduce field duty cycle) until the current is under control and keep it there. I could then have a warning light to signal under voltage which should make anyone looking for fault understand that there is a short, at the end of the day if there is a short the battery will supply most of the power

Did you read the very informative tutorial that Geko provided the link to?
 
Did you read the very informative tutorial that Geko provided the link to?

I'm working my way through it although thats for a slightly different situation,

after reading it is more or less what I'm after but I have to put particular emphasis on heat dissipation as unlike this design in an alumium case I am designing to use the original case so that the car does not loose its originality, a specific request of my friend and a vital requirement if he wants to fit these to other peoples cars as they care alot for the original look.

it goes as far as the crap indian manufacturer is allowed to put lucas on the case even though he is some cheap skate manufacturer and not lucas because he can argue on a like for like replacement.

I'm not being overly critical of these indian remakes, i looked at a failed unit and the quality is terrible bearing in mind that these things regulate power using very precise physical properties and ajustments, an armature that can actually make the contact miss because it sways to one side is not very helpful, the same as a current wire that is not even soldered to the terminals but relies on springed contact ! (or it was soldered so badly, there is a small quantity of something that I suppose is alegedly solder)
 
Last edited:
Cooling the circuit in the original case is certainly a significant concern. Perhaps a heat sink that contacts the mounting base and conducts the heat into the mounting surface would work. What surface does the original unit mount on?
 
most cases are bakalite and are clipped onto the cars body so no real chance of getting heat to dissipate that way, we are dealing with units made between 1930 and 1960, plastic and metal were a primium then.
 
Another option would be to keep the isolation relay intact and just make the voltage regulation circuit solid state.
To improve on current and voltage control by using an actual PWM control IC may work better. Some thing like a LM3525 or similar with current feedback.

For current feedback just use a length of heavy gauge wire like what was used in the original regulator (If it had it) for the current control circuit as the reference shunt if the voltage drop is within the working range of the PWM IC.

Its just a thought.
 
the other thing tcmtech is that this has to be a drop in replacement I am giving to a mechanic to fit so I need repeatablility, I was considering pic for control but not sure if something analog is best
 
Last edited:
is there a constant relationship between the amps going into the field winding and the anps coming out of the armater ? I could measure the lower amperage going in or in a way regulate this so that the output cannot soar to dangerous levels, from what I understand also if voltage is kept constant the amperage cannot increase by too much as there is an ohms law relationship ?
 
Last edited:
There is not an exact ratio being the speed of the generator always varies.
If you gave the field windings the full 12 volts you would see a fairly consistent ratio between RPM's and output voltage.
At idle you may only see 12 volts with a few amps of available current before the voltage drops off. But hat high RPM you may see 40 volts open circuit and need the full 35+ amp load on it to have 12 volts output.

The buzzing contact regulation was a crude PWM system but the field did not have an exact voltage or current control. It sort of just got the average of the many variable pulses the contacts created.
An actual PWM system would be able to give it a far more precise field power control and that would relate to a much smoother output from the generator.

Ideally the main control would be just a system voltage referenced PWM control that only regulates the average field power. A second feedback reference of the current would be only needed if the battery was real low or if the electrical system had a considerable load on it.

I am not sure of the reasoning for a uC charging regulator myself. It has far more capabilities but still the way the electrical systems are designed on the old vehicles you wont gain much. I am not sure how the program would work without extensive knowledge of the physics and actual system used in the vehicle. Theory doesn't work so well when applied to old fashioned ways things ere done. Back then 'it just sort of works' was good enough.
They did not have any battery voltage reference or thermal sensing for charge rates. They basically just tried to put out around 13 -15 volts on average at the generator and didn't care about much else. The system as whole was very forgiving so any system voltage between 10 and 16 was still well within its operating range.

The cheap low end comparator system I used worked well because it was basically a solid state version of the buzzing contact system. If I built a conversion system again I would likely go with a dedicated PWM IC and have better field control for slightly improved efficiency and charging stability. But still I would keep it simple and basic as much as possible.

If plug in and go is your intentions, simple is easiest to work with and replicate. A standard issue PWM IC mounted in a socket gives a rather easy replacement option should it go bad. A new uC IC will need a reprogram and other stuff that not just any one can do.

Ultimately it up to you as to whats practical and cost effective for the application.
 
yes tcmtech I am torn myself between uC and comparator, a uC would give a constant frequency where as the comparator will run free but probably at the original frequency of the relay system at the battery dampening effect and the pesistance of magnetic field in the field windings will be determining factors. As for replacement chips I can provide these (hm at a cost ?), I was thinking a uC could incorporate other battery management functions such as battery voltage indication for diagnostic purposes. it all comes down to how fast the chip can run, I was thinking a 1 KHz PWM and it would be a sase of sampling the battery/dynamo output with the ADC and then ajusting the PWM output, naturally it would have to run in a continuos circle, I'm not sure if it can cope with anything more or maybe I should use a comparator for basic running an a uC for any other function
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top