Excited Field Motorcycle Voltage Regulator using BJTs or MOSFETs

Thread Starter

crlsprz

Joined Mar 9, 2023
13
First of all, as this is my first post in this forum, I would like to introduce myself and thank the rest of the participants for their contributions and help.
I am glad to take part in this community and hope to be able to help in return, in anyway I can.

Anyway, I have a rather old motorcycle which uses an excited field alternator.
For this type of alternators, voltage is dependent on the rpm and the current in the coil. As rpm in the alternator are dependant on engine speed, the only way to control the output voltage is through current control.

The regulator in this motorcycle, therefore, acts like a "current tap", which allows current to pass to a coil until a certain voltage is reached (around 13,5V-14V if working properly). If that voltage is reached, the regulator "closes" and allows no more current to pass.
Well, when restoring the bike, I decided to switch to a lithium battery, which the original regulator could not fully charge.

I made some research and came across an interesting website (http://gpzweb.s3-website-us-east-1.amazonaws.com/RegRec/GPZvRegExciteField.html), so I built the excited field voltage regulator using the following diagram:
(I can only acknowledge the merit of the person who created it and who unfortunately is no longer with us)

1678379319950.png

In summary, by varying the R4 resistor, the output voltage can be adjusted. (The whole circuit analysis can be checked in the original website)

It works properly but from time to time (excessively often should I say), it blows the indicators or the headlight bulbs. It also blew an USB charge socket that I fitted, including the capacitor.
Upon some inspection, everything seems to be in good conditions, all BJTs are OK, Zener, resistors, etc.

Instead of tunning the regulator with the engine running, I used a laboratory power supply to feed the regulator (with it connected to the motorcycle coil). I set 14,5 V in the PS and then tuned the potentiometer until the current in the coil started to drop (measuring it with an amperemeter).

Do you think this could this be the cause of error? Should I proceed in another way? It’s an old bike and it’s quite hard for me to set the bike at the desirable rpms, and I don´t want to cause unnecessary wear to it.

In addition, I have some questions about the circuit though, please forgive me if they are stupid, but my knowledge in electronics is limited.

1) First of all, what is the point of using 2 resistors in line with the Zener? Wouldn’t it act as a single resistor of a higher value? Keeping in mind that one is 100 ohm and the other 0,8-1k.

2) The other question I have is about the Q1 BJT. The base current is quite high, so a lot of power is dissipated and it gets rather hot. Could a MOSFET version be made in order to make it run cooler? Can someone enlighten me with this or make a little sketch? I'm completely lost here...

Thanks a lot
 

Alec_t

Joined Sep 17, 2013
14,263
Welcome to AAC !
There are two issues here: regulating the alternator voltage and charging a lithium battery. Although the circuit shown provides a measure of voltage control it does not provide a safe way of charging a lithium battery. If a lithium battery is not charged and discharged correctly it can catch fire or explode. Check out Battery University.
What is the point of using 2 resistors in line with the Zener?
R3 prevents excessive base current in the event that R4 is adjusted down to very low Ohms.
Could a MOSFET version be made in order to make it run cooler?
Yes. That would be a big improvement.
 

crutschow

Joined Mar 14, 2008
34,201
1) First of all, what is the point of using 2 resistors in line with the Zener?
R3 is added so that the pot wiper can't go all the way to ground and zap Q2.
2) The other question I have is about the Q1 BJT. The base current is quite high, so a lot of power is dissipated and it gets rather hot. Could a MOSFET version be made in order to make it run cooler?
The power dissipated equals the voltage drop across the transistor times its current, which is determined by the coil resistance and the coil current when regulating.
That power will be the same no matter what type of transistor you use.

Do you know the coil resistance and the current at the regulation point?
The maximum dissipation occurs when half the battery voltage appears across the coil resistance.

One way to reduce that dissipation is to use a PWM switching circuit, which rapidly turns the transistor full-on to full-off, with the duty-cycle determining the average coil current.
Since the transistor is switching and doesn't spend significant time in its linear region, its dissipation will be much less.
A typical circuit to do this is reasonably simple, using one IC a MOSFET, and a few passive parts.
 
Last edited:

Thread Starter

crlsprz

Joined Mar 9, 2023
13
Welcome to AAC !
There are two issues here: regulating the alternator voltage and charging a lithium battery. Although the circuit shown provides a measure of voltage control it does not provide a safe way of charging a lithium battery. If a lithium battery is not charged and discharged correctly it can catch fire or explode. Check out Battery University.

R3 prevents excessive base current in the event that R4 is adjusted down to very low Ohms.

Yes. That would be a big improvement.
Thank you for your reply!
Your advice is very much welcome, I chose a battery with a built-in bms just in case, but every futher info is appreciated.
About the R3 now I understand. I was not taking into account that R4 could indeed be too low when doing the adjustment.
 

Thread Starter

crlsprz

Joined Mar 9, 2023
13
Do you know the coil resistance and the current at the regulation point?
The maximum dissipation occurs when half the battery voltage appears across the coil resistance.

One way to reduce that dissipation is to use a PWM switching circuit, which rapidly turns the transistor full-on to full-off, with the duty-cycle determining the average coil current.
Since the transistor is switching and doesn't spend significant time in its linear region, its dissipation will be much less.
A typical circuit to do this is reasonably simple, using one IC a MOSFET, and a few passive parts.
Thank you for your answer!
I measured coil resistance at 3,9 ohms.

About the power dissipation, perphaps I didn't make myself clear.
I meant that the current through Q1 base was rather high, so the power dissipated in R1 was high as well (around 4,3 W).
Even with a 10 W resistor it gets way too hot, I'm afraid it can burn itself or other components...
I was thinking on a MOSFET as there is no gate current, so those 4,3 W could be saved.
 

crutschow

Joined Mar 14, 2008
34,201
I measured coil resistance at 3,9 ohms.
So the Q1 dissipation can go as high as 12.5W with 14V battery voltage.
I meant that the current through Q1 base was rather high, so the power dissipated in R1 was high as well (around 4,3 W).
Okay so yes, a P-MOSFET for Q1 would allow the use of a larger value, low power resistor (e.g. 1kΩ, 1/4W), for R2.
But I'm concerned about the regulation with that change.
The circuit is a poor design as the pot basically varies the current through the 12V Zener to adjust the voltage level, and that's not likely to give good regulation.
Going to the MOSFET and higher value resistor, would greatly reduce the base current of Q2 and thus affect how the regulator works.

I'll give some thought as to how the circuit might be improved.

Does the coil connection have two separate leads or is one of them internally connected to ground?
 
Last edited:

Dodgydave

Joined Jun 22, 2012
11,277
Here's a circuit i quickly drew up using a TL431 zener and N-mosfet , the voltage range is 12.3V to 16.7V, . If anyone has simulation software they can tweak it.

IMG_20230309_184708.jpg
 

Ian0

Joined Aug 7, 2020
9,618
A couple of points:
1. I've seen that circuit before, but with a resistor between Q1 collector and Q2 base to give it some hysteresis - which makes is oscillate, and gives a PWM output.
2. A lithium battery will have a BMS, which can disconnect the battery from the charger when it gets fully charged. If it does so whilst the engine is running there is nowhere for the energy to go except into the lamps. To prevent it happening you need a big shunt regulator set at 15V capable of dissipating the entire output of the alternator - or, better still, stick with the lead acid battery - there are plenty of vintage bike connoisseurs that would frown upon your using anything but the original type of battery (or transistor, for that matter).
 

crutschow

Joined Mar 14, 2008
34,201
2. A lithium battery will have a BMS, which can disconnect the battery from the charger when it gets fully charged. If it does so whilst the engine is running there is nowhere for the energy to go except into the lamps. To prevent it happening you need a big shunt regulator set at 15V capable of dissipating the entire output of the alternator - or, better still, stick with the lead acid battery - there are plenty of vintage bike connoisseurs that would frown upon your using anything but the original type of battery (or transistor, for that matter).
I agree.
A lithium battery is not a good choice for a motorcycle as it's much fussier about charging currents and voltage.
Stay with a lead-acid battery.
 

Thread Starter

crlsprz

Joined Mar 9, 2023
13
Does the coil connection have two separate leads or is one of them internally connected to ground?
So, in this link you can find the original regulator diagram, which is electromechanic. I am aware that the original regulator has three "states" (http://kz400.com/Workshop manual bi...ual/Manual s 157-165/maintenance page 159.JPG)

1678397007312.png

This is for a motorcycle regulator with sla battery.
Thank you, I will have a look in any case!

I agree.
A lithium battery is not a good choice for a motorcycle as it's much fussier about charging currents and voltage.
Stay with a lead-acid battery.
A couple of points:
1. I've seen that circuit before, but with a resistor between Q1 collector and Q2 base to give it some hysteresis - which makes is oscillate, and gives a PWM output.
2. A lithium battery will have a BMS, which can disconnect the battery from the charger when it gets fully charged. If it does so whilst the engine is running there is nowhere for the energy to go except into the lamps. To prevent it happening you need a big shunt regulator set at 15V capable of dissipating the entire output of the alternator - or, better still, stick with the lead acid battery - there are plenty of vintage bike connoisseurs that would frown upon your using anything but the original type of battery (or transistor, for that matter).
About the use of lithium batteries in motorcycles, I mean, it's not that they are DIY or hand-built. There are sold commercially (LiFePo4), and they are also OEM in some newer motorcycles. I don't know why it should be a problem.

This is the circuit of a deconstructed Bosch altenator regulator. Its quite simple & works.View attachment 289339
Thank you very much!
 

Reloadron

Joined Jan 15, 2015
7,480
I will only suggest the following. My ride is a 1992 Harley Electra Glide. A few years back at a HOG Chapter meeting we were introduced to the new LiPo batteries. Everyone marveled at how light the batteries were as we passed them around. That's when I asked about the charging regulators. All the SLA battery regulators would need changed for the new batteries. Those who opted for change for the most part had problems. My advice is stay with the original OEM system and the SLA battery. Unless you can see substantial gain in changing battery chemistry I would leave well enough alone. :)

Ron
 

debe

Joined Sep 21, 2010
1,389
These are the wave forms on the Rotor of Bosch altenator under 2 different charge rates looks like.ALT ROTOR. 20A. Waveform.JPGALT ROTOR.5A. Wave form.JPG Shouldnt be to hard to build.
 
Top