Using 50W Zener to regulate motorcycle alternator?

[tylernt]

I have an old motorcycle from the early 70s with a 6V electrical system and no regulator -- as engine RPM rises, so does the alternator's output voltage. I would like to add a 6.8V, 50W zener diode shorted directly to ground, after the rectifier output, to keep the alternator from overcharging the battery.

I understand this is the same principal used in modern motorcycle regulators, but wanted a sanity check on the actual implementation. I found a NTE5247A zener which can be bolted to the frame, which would act as a very large heatsink.

Thoughts?

 

[Externet]

I would rather use a LM196 regulator set to 7V and advise against the zener.

 

[mneary]

The NTE5247A (6.8V +/- 5%) can have a knee voltage anywhere between 6.46 and 7.14 volts. 6.46 is too low and will prevent your battery from charging. 7.14 might be about right.

The one NTE5247A that I found is $27. You can build a better regulator for a lot less.

 

[Hero999]

How about using a smaller zener and boosting it with a few large BJTs?

 

[Nigel Goodwin]

This was the system used on many 60's and 70's motorbikes, with a large zener diode on a heatsink mounted between the fork legs.

As suggested, a smaller zener feeding a large transistor would be a cheaper and easier soloution, or even an opamp feeding a transistor, where you could adjust the exact voltage.

As a matter of interest, it's called a 'shunt regulator'.

 

[BrownOut]

What's going to limit the current through the shunt device?

 

[Nigel Goodwin]

What's going to limit the current through the shunt device?

The available current from the alternator. Essentially with the headlight ON the alternator only just provides enough current to drive the headlight (at reasonable revs) and to trickle charge the battery. During daylight running (without the headlight ON) the zener dissipates the energy the headlight takes, minus the current the battery is charging with.

Motorbike alternators are generally permanent magnets, so there's no way to control it directly like a car alternator.

 

[BrownOut]

Since you're making your own regulator, you can use a schemem like the one attached. R2 is chosen to trun on Q2 at a current above the maximum current used by your motorcycle's systems. As the zener turns on, Q1 turns off, making the requirements for the shunt lower, and the regulator more efficient.

 

[tylernt]

Thanks for the feedback. The thing I like about the zener is the dead-simple installation, so a complex IC-based solution is rather less attractive. If it will cost less, I suppose I could bolt a large transistor or two to the frame and feed them with a little zener, but I'm not sure what part to spec -- any suggestions on what transistor(s) would be appropriate?

 

[MikeMl]

I have an old motorcycle from the early 70s with a 6V electrical system and no regulator -- as engine RPM rises, so does the alternator's output voltage. I would like to add a 6.8V, 50W zener diode shorted directly to ground, after the rectifier output, to keep the alternator from overcharging the battery.

I understand this is the same principal used in modern motorcycle regulators, but wanted a sanity check on the actual implementation. I found a NTE5247A zener which can be bolted to the frame, which would act as a very large heatsink.

Thoughts?

View attachment 39113

Here is how I would do it. I'm simulating a situation where the alternator is capable of 10A, there is a minimum fixed load of about 3.5A, so the Shunt Reg has to pass about 6.5A. Using the TL431, the regulation band is very narrow. I'm suggesting using a large TO220 NFet. It should have a voltage rating of >30V, and will need to be mounted on an insulating kit where there is lot of metal to dissipate the heat. The NFet is more readily available than a large Zener, and likely much cheaper.

The regulation voltage is adjustable over a narrow range with a trim pot. In the simulations that follow, the five traces correspond to five different positions of the pot wiper. First plot shows the regulation of the system voltage. The second plot shows the current being shunted through the NFet. The third shows the power dissipation in the NFet.

 

[MikeMl]

What's going to limit the current through the shunt device?

Just the max output current from the alternator. The shunt reg only has to sink the excess current over and above what the minimum loads in the motorbike draw....

 

[Hero999]

Here's how I'd do it.

I modified the circuit on the datasheet.

It works on the same principle but the big PNP transistor has been replaced with a Sziklai pair with the NPN transistor split in to two transistors. You could use a Darlington pair but I chose the Szkai because it's easier to get hold of high power NPN transistors.

If correctly designed, most of the power can be dissipated in the emitter resistors which takes the strain off the transistors.

 

[MikeMl]

...
If correctly designed, most of the power can be dissipated in the emitter resistors which takes the strain off the transistors.

Good Idea! I added a 1Ω 50W resistor. I also jacked up the voltage slightly for better float charging the battery. This sim shows the regulation voltage with the pot centered (green trace), the resistor dissipation (blue trace) and the NFet dissipation (red trace) vs shunted current. Only the OP knows how much current needs to be shunted to ground with the headlight off.

Peak dissipation in the FET is only 12W, which is quite doable. The FET and wire-wound resistor should be mounted to the motorcycle frame.

 

[Hero999]

Whoops I didn't post a schematic but can you probably guess what it will be.

 

[MikeMl]

Whoops I didn't post a schematic but can you probably guess what it will be.

You will have to eliminate Rint, or at least make it no more than the wiring resistance.

 

[Hero999]

That's what it represents, I just forgot to say - int is short for internal resistance of the alternator and wiring.

 

[RCinFLA]

I assume this is permanent magnet alternator. Have you thought about adding a couple of extra side mount headlights and at at least put excess power to some good use. Modulate the brightness on the auxilary headlamps for voltage stabilization.

 

[spu]

I have a similar challenge with my 1939 Chris-Craft. I have an 8 volt battery based electrical system. The output from the 3rd brush generator is 11.54 volts at cruising speed. I have 12 excess amps I need to sink. Any suggestions on modifications to this circuit for a nominal 9.5 volt output?

 

[tylernt]

FYI I ended up using a 7.5V 50W Zener, but it doesn't work: system voltage exceeds 8V when the engine is revved up. I don't know if I installed it wrong or if I got a bum part or what. Ah well.

 

[crutschow]

FYI I ended up using a 7.5V 50W Zener, but it doesn't work: system voltage exceeds 8V when the engine is revved up. I don't know if I installed it wrong or if I got a bum part or what. Ah well.

It may be that the impedance of the diode at the breakdown knee is high enough to that that voltage increases to 8V with the excess current going through it. You may need diode with a Zener voltage of about 6.8V.