overvoltage protection for motorcycle

[EEBlackSheep]

I have a dirt bike that has a single phase alternator and an AC regulator that shunts excess voltage to ground.   The problem with this configuration is if the regulator fails it will then allow high voltage downstream and fries expensive parts.  I would like to put somekind of overvoltage protection in place so that if anything over 14.4V it will shutdown.  There is a rectifier so I can put in a DC circuit. 

Any off the shelf suggestions?  I need about 5Amps.  All the automotive regulators I see work with the alternator.

[amspire]

Is the battery being fried too, or is there no battery in this kind of bike?

If the battery voltage is going over 15V, then it is damaging the battery as well, so it will need protection too.

[EEBlackSheep]

no battery. 

[amspire]

No battery?

Normally a battery does a major job in regulating the voltage swings.

With no battery, I am not experienced in how they make it all work.  Is there a big capacitor somewhere to smooth out the DC, or do they just have a very crude rectified AC going to al the accessories?

What sort of parts need protecting. Is this things like lights, electronic ignition, or is it electronic accessories that you have added.

This is old technology. It just sounds odd that the regulator can fail so easily.  There will be ways you could design an extra regulator or a crowbar + fuse protection, but it sounds like there is a problem that is not being fixed to me.

[Simon]

yea I'd say a crowbar and fuse but I'd be inclined to rebuild the regulator

[amspire]

yea I'd say a crowbar and fuse but I'd be inclined to rebuild the regulator

It looks like there are lots of aftermarket regulators available.

I just don't understand why a regulator would be failing, unless it is a very bad design.

Get a different one if the original is so bad.

Adding a second regulator because the first regulator is so lousy sounds a bad solution.

[hacklordsniper]

Get used regulators from Japan bikes (except older honda) and your problem is solved. You could also use tvs and a fuse to open the fuse in case of overvoltage

[Zero999]

It's strange that the regulator is on the AC side. It's normally on the DC side and is often integrated into the rectifier module.

[EEBlackSheep]

Crowbar and fuse - thanks!


And for the curious here is the schematic of the 2005 motorcycle (they are from Spain and called GasGas)

[RCMR]

It's strange that the regulator is on the AC side. It's normally on the DC side and is often integrated into the rectifier module.

I don't know what circuit you're talking about but the voltage regulator with an alternator-based system usually regulates the voltage by varying the current to the rotor.  The more current in the rotor, the higher the voltage out of the field windings (which are the ones which are rectified to produce the actual voltage output).

This allows you to regulate a very high current (in the field winding output) using a relatively small change in "control/exciter" current whch feeds the rotor/armature.

This will ultimately be more efficient and produce less heat than simply throwing a DC regulator on the rectified output of the alternator.

[IanB]

No battery?

Normally a battery does a major job in regulating the voltage swings.

With no battery, I am not experienced in how they make it all work.

The schematic shown later makes it clear--it is a magneto system. Magneto systems were used on old bikes before they had batteries or electric starters. Apparently magneto systems are still used in special applications like dirt bikes. They are presumably simpler and more robust than battery systems, since I'm guessing a dirt bike would be a very harsh environment for a battery.

[Rufus]

And for the curious here is the schematic of the 2005 motorcycle (they are from Spain and called GasGas)

The alternator powers lights with ac and a fan via a bridge rectifier. Are they expensive parts?

The 'regulator' is probably just back to back power zeners. You could probably wire a second in parallel as a backup.

[amspire]

I am trying to make sense of the circuit.

If the magneto or alternator was basically fixed magnets on the shaft and fixed stator coil, then the output voltage swing would be huge as the motor speed changes.

The "regulator" is a dinky little thing, so it cannot possible regulate this voltage by absorbing the extra power. That means there must be some other regulation process happening, and the "regulator" is designed to absorb transient spikes, say, when you switch the headlights off.

So my guess is that there is a magneto that provides enough power for the ignition control module via the red wire, and it in turn is controlling an alternator rotor current via the red.white wire to regulate the alternator output (yellow wire).  Because the is no battery or capacitance in this circuit, any sudden drop in load will cause a big voltage spike, and the $2 worth of back to back zeners in the $70 regulator module clamp the spike. Lets say the headlight is using 4A of current and you turn it off. Then the regulator module has to momentarily absorb that 4A of current until the energy in stray inductances are dissipated  the ignition control module winds down the alternator output down a little. If these zeners are underrated, or if a wiring fault (like a bad earth on the headlights) causes extra transient spikes, then the regulator blows, which can in turn kill the fan, and along with it, the $0.50 full wave rectifier in the "DC Converter"  - probably worth at least $70 again - might die in sympathy.


If my guesses are right, then I assume that you are not worried by the headlight blowing - which it would if the voltage wasn't being regulated by the ignition control.

So you could take the regulator, attach it to a power supply via a 1K resistor, and determine the zener volts. Then if you wanted to you could get some industrial strength zeners and make your own module.  Also you could obviously take the "DC Converter" and replace it with a rectifier + LDO regulator module that would eliminate any chance of the fan being burnt out by voltage spikes.

But I return to why the regulator is blowing. There just shouldn't be lots of transients, and so something is causing this. A wiring fault, or I have seen a report from someone saying that water in the alternator housing was somehow causing transient spikes blowing the "regulator". Don't know how that works, but we don't know what is actually in the magneto+alternator housing.

One senario is that the rotor is connected via slip rings and carbon brushes. A drop of water could push the brush away from the slip ring breaking the rotor current. So if water is splashing around and can get to the slip rings, then you may start getting more transients then the "regulator" can thermally cope with.

All speculation of course. 



Richard.

[Rufus]

I am trying to make sense of the circuit.

If the magneto or alternator was basically fixed magnets on the shaft and fixed stator coil, then the output voltage swing would be huge as the motor speed changes.

Shunt regulators are commonly used with permanent magnet alternators. The coils and magnetics can be designed to provide a soft current limit so the power dissipation in a shunt regulator is also limited.

[EEBlackSheep]

The alternator powers lights with ac and a fan via a bridge rectifier. Are they expensive parts?

Yes.  The whole reason for some kind of protection is because the fan is about $150 a shot (I don't run a headlamp).  I would rather it blow a fuse and let it overheat rather than burning out the fan again.  I have a temp gauge, as well as the rad cap will let go before anything bad happens to the engine.

I just don't understand why a regulator would be failing, unless it is a very bad design.

Its a dirt bike (well really, a trails dirt bike, but I digress).  Nasty environment for electronics. Its 10 years old. Lots of heat from the engine.  Lots of bashing around. Lots of water, mud dirt, rocks, sticks.  Everything breaks sooner or later.  Always.



I am liking the crowbar circuit.  It looks like it will suit my needs nicely.  I found a precision circuit that I can put on a perf board then dunk in epoxy: http://axotron.se/index_en.php?page=26  This will give me 14V +/- 0.46V

Here is the schematic:

[amspire]

I am trying to make sense of the circuit.

If the magneto or alternator was basically fixed magnets on the shaft and fixed stator coil, then the output voltage swing would be huge as the motor speed changes.

Shunt regulators are commonly used with permanent magnet alternators. The coils and magnetics can be designed to provide a soft current limit so the power dissipation in a shunt regulator is also limited.

There has to be some sort of regulation from the dynamo/alternator before the "regulator".

Otherwise if there is enough voltage to power a headlight at 2000 rpm, then there will be 3 times the voltage (6 times the power) at 6000 rpm.

By the size of the "regulator", it couldn't handle 20W continuously, and it would probably have to sink well over 100W at 6000 rpm+ to limit the alternator voltage.

There must be some regulation process before the "regulator" - either from the ignition control module, of they have done something very clever with saturating cores in the static armature. Some kind of  regulation is happening.

Richard

[amspire]

I am liking the crowbar circuit.  It looks like it will suit my needs nicely.  I found a precision circuit that I can put on a perf board then dunk in epoxy: http://axotron.se/index_en.php?page=26  This will give me 14V +/- 0.46V

You will have to change R3 to something like 50K.

Also you could use a resettable polyfuse/polyswitch or PTC instead of a wire fuse.  That way it will protect the fan, and it will reset every time the engine stops.

Richard

[EEBlackSheep]

Also you could use a resettable polyfuse/polyswitch or PTC instead of a wire fuse.  That way it will protect the fan, and it will reset every time the engine stops.

Great - thanks!

[EEBlackSheep]

I found a circuit that uses NTE7172 which is a replacement for the obsolete MC3423 - http://www.onsemi.com/pub_link/Collateral/MC3423-D.PDF

Might be a dollar or two more, but easier for me to build.  Thoughts on this?

[amspire]

How the fan is burning out - just by heat from to much voltage, or is it that the fan has an electronic commutator that is fusing due to voltage spikes?  If it is failing due to voltage spikes destroying transistors, then you need more of a genuine voltage regulator instead of a crowbar clamp.

If it is just by overheating due to  too high a voltage, then how precise does the cutout voltage actually have to be?

If you use that zener circuit on that axotron page with a 15v zener along with a 1v gate turn-on SCR, it would clamp at 16V. I like the zener circuit in it is very fast and reliable, and there is no possibility of stability issues.

Now I just remembered that the volts is AC, so you would really need to add a bridge rectifier in front of the clamp, and a capacitor + series resistor across the SCR chosen so the minimum latching current is maintained at the lowest RPM.  So you have to take the bridge rectifier voltage drops into account.

Richard.

[EEBlackSheep]

The fan is DC after the bridge rectifier - so I plan on putting the clamp after the rectifier - so I think I am good there.

Unfortunatelty the fans are sealed so I cannot tell what cause the issue - commutator fusing or coil burning out.  I suppose putting in a DC regulator circuit after the rectifier would be another option as well.  But I am not 100% sure if there is enough voltage coming from the alternator. I will put the bike back together and get some measurements.