With suitable heatsinking, a pair of 1N3312B diodes (50W each) should work fine.

Robert,
I cannot find any stock around here for those Zeners as they are not stocked by any of our electronics suppliers. I only saw one add for those on ebay and they were for $26 for two, without international shipping.

The complete aftermarket AC regulator available in the USA goes for $20 without shipping.

That does not make such a DIY build cost effective as it still needs a heatsink and sealing against moisture.

Thanks for the suggestion, every bit helps even if this is a simple schematic, but it seems to be quite a grey area. Much appreciated.

 

Am I understanding correctly that you want to regulate a variable 12-14V AC to around 12V DC at at least 35W?

If you don't mind inefficiency you can use a linear regulator. But it would get hot and would require a big heatsink.

Otherwise you can take advantage of the fact that the peak of the AC voltage (assuming sine wave) would be 16.8 volts min. and 19.6 volts max., smooth that with a capacitor and then use a simple step down regulator (I suggest a SMPS buck regulator) to 12 volts will work well. Bonus is that it will often be cheaper and is more efficient. It will last longer because less stress is placed on components.

35W at 12 volts is just under 3 amps. Try the LT1070 or LM2677 both of which have a internal 5A switch, otherwise, you can get the pretty well standard MC34063A (but you'll need external transistors.)

If you must have a linear regulator, the LM338, a 5 amp linear regulator offers a high current solution. Beware, the IC will burn up if not heatsinked properly.

 

Am I understanding correctly that you want to regulate a variable 12-14V AC to around 12V DC at at least 35W?

No, actually I want to regulate a variable 0-30V AC to 12-14V AC and use a very low component count.

Thanks for your input.

 

It's going to be difficult to regulate 0V AC.

But it would be possible to step your 30V (max.) to say 4V (min.) down to 3VAC and then step it up again. Bit inefficient, though. If you don't mind the lamp dimming when the voltage is low (below 10VAC or so), the idea I posted would work fine.

I don't see if you're running a lamp why you need AC... DC should be fine.

 

It's going to be difficult to regulate 0V AC.

I don't see if you're running a lamp why you need AC... DC should be fine.

I want to keep it as original as possible so that the owner does not have future problems if someone else tries to figure out what I have done. Honda has used this approach from the late 70's right up to 2007 (XR650R) so I do not really want to reinvent the wheel. Space on an offroad bike is always a problem and it also has to be sealed properly. The one unit shown in the pictures I posted a few posts back shows it as the same size as a 25A rectifier bridge and most likley uses the same heatsink/case. Obviously there cannot be much inside there and it works. It regulates AC quite good enough for the headlamp bulb.
I will post more detail as this project rolls out a bit more.

Thanks for your inputs

 

Another Honda XL500 has been dropped off at home. As you can guess headlight problems!

I found a regulator for a chinese scooter that can regulate the AC to the headlight and when fitted it worked perfectly. It is very small and has a heatsink but it does not seem to get warm at all. Over and above the bulbs not blowing now, the lights have more of a flicker at idle which disappears as the engine RPM's go up slightly.

Now I am more determined to find a DIY solution for this, using the same OEM approach.

I am sure someone is going to come up with an idea. Maybe I should try to find a faulty one and try and get the potting out and see how it is done.

Please, if anyone has a schematic or an idea on how to regulate the 12V AC to the lights on an offroad Honda XL250S please share
Thanks again.

 

A simple capacitor may take care of the flicker a bit. But most likely, the alternator is just not producing enough voltage for the rectifier and regulator at such low rpm.. But I do think a cap should help.

 

The flicker at idle is not a problem and can be expected with a light fed directly off the magneto, so it is not a bother I think.

I have just opened up a failed AC regulator, unfortunately the potting was like rock, nothing would soften it, neither heat or any thinners nor acetone.
I chipped it apart and unfortunately could not prevent damage to the components, but there was one small surface component on a small white ceramic square, two diodes and a small electrolytic capacitor. The one side of the surface component was grounded to the case, another went to the AC tag, another to the anode of a diode. The diodes were the size of the 1N414 glass type.
It was impossible to identify any of the components and I cannot be sure that the surface component did not have a resistor on it or not. So component count is 4 maybe 5 parts only and definitely only one large semiconductor and two small glass diodes with a capacitor.

The question now is, do they use a triac for regulating AC or a SCR?
It seems as if the schematic from Solcar is close to what the OEM's may be using, component wise.

I will appreciate any further help with the circuit below. At the moment it just pulls or regulates the AC to the headlight down to 1.8V. I have used two 15 V zeners as in the schematic. Does anyone have any further suggestions on this?

 

you could use an 'amplified zener' in place of the actual high power ones.

Something like a 2N3055 with a 100 Ohm resistor between base and emitter and a decent size (5W?) 12V zener between base and collector, then put the whole thing inside a bridge rectifier to make it independant of polarity.
Any decent rating power transistor should work.

You would get 12V + 3 x diode drops, so around 14V or so.

High power encapulated bridge recs are quite cheap and the other bits are almost trivial, so the total cost should still be low.

 

A cheap-and dirty way to do it, would be to rectify the 30vac with only one diode, so the rms value won't be so high, and hopefully the lamp won't burn.

I don't know how that shunt regulators are able to be sealed. They should dissipate a lot of power. The most sane option, would be a switching approach, but isn't likely to have a low component count.

 

Is the motorbike a small Honda 70cc or somethink like that?

If so, even if the bulb is running on AC, it could be a battery problem.

Running the bike with no battery or with a bad one will burn the headlamp.

I cannot really explain wy but this problem is documented on some Honda forums.

Alain

(Edit: Reading back, I saw that you were talking about a Honda Xl500. I don't know if my post still apply.)

 

It would most likely use the battery as a buffer and a smoothing cap to handle spikes and such.

Thats why it wont work, at least for long, without the battery.

 

It seems that very few people are aware that it is not unusual for manufacturers to use AC directly for a motorcycle or scooter headlight and that regulating the AC is not a big deal and does not generate lots of heat.

I was really hoping that some suggestions on the schematic will be forthcoming as I am stumped on what to do at the moment.

Anyone with suggestions on a schematic to regulate AC to the headlight, your inputs will be appreciated.
Thanks

 

What can be easier than putting a single diode on the wire to get rid of half the AC? If you do that, you will reduce the risk of burning the lamp a lot, and is as cheap and easy as it gets.

By the way, the schematic you put in the original post will short-circuit the alternator, and whatever is connected to the same wires when the triac is triggered. Not a good idea to do that, and even if it doesn't explode instantly somehow, will dissipate lots of power.

Some manufacturers seem to regulate the voltage by regulating the alternator itself: http://jhau.maliwi.de/mot/voltreg.html , but that depends on the motorcycle.

I suggest a search like this: http://www.google.com.ar/search?q=m...icial&ei=mDklTJXrLMOB8ga3-Oj8Cg&start=10&sa=N . Adding the motorcycle brand and model would likely help.

*IF* the output voltage from the alternator is reasonably stable, you could try with a circuit used to dimm ac lights: http://www.epanorama.net/documents/lights/lightdimmer.html , but it will require some modifications.

 

rjenkins you could use an 'amplified zener' in place of the actual high power ones.
Something like a 2N3055 with a 100 Ohm resistor between base and emitter and a decent size (5W?) 12V zener between base and collector, then put the whole thing inside a bridge rectifier to make it independant of polarity.
Any decent rating power transistor should work.

This is going to be a large unit and as I said before I really want to regulate AC not DC, exactly what the manufacturers did after a few years.

Norfindel A cheap-and dirty way to do it, would be to rectify the 30vac with only one diode, so the rms value won't be so high, and hopefully the lamp won't burn.

I don't know how that shunt regulators are able to be sealed. They should dissipate a lot of power. The most sane option, would be a switching approach, but isn't likely to have a low component count.

That is an idea, but the AC varies continuously with engine speed and may be too weak at very low RPM's. I really want to regulate the AC, AC in AC out.

AlainB Is the motorbike a small Honda 70cc or somethink like that?

If so, even if the bulb is running on AC, it could be a battery problem.

Running the bike with no battery or with a bad one will burn the headlamp.

I cannot really explain wy but this problem is documented on some Honda forums.

Alain

This approach is used on many off road Hondas, scooters and some of the other manufacturers as well. At the moment I am looking at Honda XL250 and XL500. The battery has no effect on the lights which are driven by AC from a dedicated winding in the alternator and does not use the battery or charging winding. On some Honda forums the posters have no idea how this works exactly and compare with cnventional systems.

retched It would most likely use the battery as a buffer and a smoothing cap to handle spikes and such.

Thats why it wont work, at least for long, without the battery.

The battery does not come into the equation here at all. There is no capacitor installed. The battery has it own rectifier and regulator for the DC part of the bike. The lights AC part is separate and only for the headlight.

Norfindel What can be easier than putting a single diode on the wire to get rid of half the AC? If you do that, you will reduce the risk of burning the lamp a lot, and is as cheap and easy as it gets.

By the way, the schematic you put in the original post will short-circuit the alternator, and whatever is connected to the same wires when the triac is triggered. Not a good idea to do that, and even if it doesn't explode instantly somehow, will dissipate lots of power.

Some manufacturers seem to regulate the voltage by regulating the alternator itself: http://jhau.maliwi.de/mot/voltreg.html , but that depends on the motorcycle.

The alternator will not explode if shorted out. If the regulaor works as it should, it hardly gets warm, it just shunts the excessive voltage to ground
The regulating you are refering to is a regulator for a field regulated alternator and is much different than a fixed magnet alternator that uses shunt regulation.


Thanks for your inputs, but it is not helping me at all.

 

Hi, I think I will rather close this thread down. The original intention of this thread has been lost and we are going in circles.

Thanks again to Sgtwookie and Solcar who gave some usefull information and schematics as well.

Thanks again for all your good intentions.

 

I have to add this:

You are after a circuit that honda already used, ac in ac out.

Just because honda did that does not mean it is the best way to do it. Just look at toyota. And notice that honda no longer does it this way.

Why you are getting upset, when you have received may ideas that you can use, that WILL work fine, I do not know.

You are the Original Poster, the thread starter, and I understand your urge to get what you want, but there really are better ways.

 

Just because honda did that does not mean it is the best way to do it. Just look at toyota. And notice that honda no longer does it this way.

Why you are getting upset, when you have received may ideas that you can use, that WILL work fine, I do not know.

You are the Original Poster, the thread starter, and I understand your urge to get what you want, but there really are better ways.

Not really, just a bit confused as I thought I had explained quite a few times why I did not want to deviate from the proven track and asked for some help for a specific approach. As I said I appreciate all the help, but I was looking for a specific solution.

I want to keep it as original as possible so that the owner does not have future problems if someone else tries to figure out what I have done. Honda has used this approach from the late 70's right up to 2007 (XR650R) so I do not really want to reinvent the wheel.

 

I understand. Sorry we couldn't be of more help on this one.

 

To me, trying to patch the problem this way is a bad idea anyway. After all, there must be many thousands of other motorbikes of this type that do not burn headlamps.