I've researched the forum before posting, but it seems my question is a bit specific and not answered before. Please excuse me if I'm wrong.

So what we have here is a chinese scooter alternator, which gives 70W max. The alternator is a half wave stator, half of the power is converted to DC through a regulator to charge the battery, the other half is also regulated by an extra regulator I put there and drives a 45W LED headlight. So far so good. But the problem is LED flickering due to the nature of the regulator/rectifier.

So I'm trying to decide what smoothing capacitor to add after the regulator. I noticed by trial and error that the flickering stops completely after I put a 10000uF electrolytic cap in there. I also after that, calculated that a minimum 4700uF cap is needed for an approximately 5% ripple voltage, for the current drawn (2V ripple at around 14VDC).

But the main problem, which I think everyone misses (or I am wrong and I'm trying to figure which of the two stands here) is the CONSTANT current draw (Irms) for the smoothing capacitor, in order to maintain a smooth output.

I've run a simulation with a full bridge rectifier and it seems a 10000uF, if inrush current is left out, it draws a constant current Irms of around 9A!!!!

Full bridge smoothing cap

So, unless I'm missing something here, the alternator which can only give 70W max (so 5.8A at 12V), will be constantly overloaded and eventually will burn out.

Of course half wave stator will give half of the power and the available power will be half of 70W, but I have no problem with that, I can fix it to run full wave by removing the center tap. My problem is the constant current drawn from the smoothing cap.

Could anyone please help me on that and tell me if my calculations/simulation are wrong or if indeed that is the case here? And if it is so, what could I do in that case?

Thank you all very much in advance!

  • \$\begingroup\$ My comment that "everyone misses out constant current" is referred to that most of the times people try to calculate the inrush current, which of course is also important, but I've seen almost nothing about the constant current draw for the smoothing capacitor to do its job properly. So please don't misunderstand that line. :-) \$\endgroup\$ – Nick G. Apr 4 '18 at 12:19
  • \$\begingroup\$ It's something wrong with the simulation, I cannot find a datasheet for MDA3551 , where does this Irms current goes? V is 15V and I through R1 is 3 Amps. The only explanation would be that the current goes back through the bridge due it's reverse recovery charge. At 500 Hz nobody can see any flickering, either the frequency is not 500Hz or the variations of V1 with the speed are the source. You should include the internal impedance of the generator for a more accurate simulation. Wouldn't be better to place a second battery for the LED since there is a ready tested circuit on the scooter? \$\endgroup\$ – Dorian Apr 4 '18 at 13:51
  • \$\begingroup\$ Thank you Dorian! You are right my friend, I can't find any datasheet for the bridge too. I picked one in Multisim since it showed high current rating but other than that I don't have any other info. I will try with a different one and come back. Also at high rpms (500Hz) no flickering happens. The flickering happens when the scooter is idling. I will change the components, run the simulation and post the new results here. About the battery, there's no space my friend, it was the first thing that crossed my mind!Battery is the best smoothing tool but it's a small scooter... \$\endgroup\$ – Nick G. Apr 4 '18 at 20:50

A smoothing cap is loaded only when the rectifier voltage output is higher than the capacitor voltage. This gives you current spikes at the top of the rectified sine, and no current at all for the other times.

You are right the rectifier needs to be able to handle the current spikes, or it will burn out.

A better design used another single diode, and put a substantially smaller smooothing cap behind it, just big enough for powering the headlight.

BTW: A 45W LED headlight is much brighter than other people's normal headlights, and most likely, than allowed. The scooter has no license and you will run into insurance problems if you are involved into an accident.

  • \$\begingroup\$ Thank you very much for your quick reply my friend. So you mean that the Irms calculated above is wrong? P-p current (spikes) is much higher of course, something around 30A I think. But isn't the Irms we need to calculate here? And from what I understand, you mean putting another diode, after the bridge in series, before that a smaller cap and after that another larger smoothing cap? Also the headlight is 20W+20W, 20W for the main lights + 20W for the far distance lights, so it's legal. Thanks for mentioning it though! :-) \$\endgroup\$ – Nick G. Apr 4 '18 at 13:08
  • \$\begingroup\$ RMS is heat related, so the only part where the RMS value becomes interesting is the series resistance losses of the capacitor. \$\endgroup\$ – Janka Apr 4 '18 at 13:26
  • \$\begingroup\$ But isn't also related to the heat produced at the alternator's side? I don't mind the rectifier can handle the amps for sure. The alternator is the problem here... \$\endgroup\$ – Nick G. Apr 4 '18 at 20:35
  • \$\begingroup\$ Only if you assume the internal resistance of the alternator is anywhere as worse than the ESR of an 10000µF Al cap. \$\endgroup\$ – Janka Apr 5 '18 at 1:21

A generator voltage is proportional to the rotation speed. You say that the flicker is happening at lower speeds. At 250 rpm the voltage should be 6V rms, , maybe less than 7v after the bridge. Very hard to find detailed information about LED headlight bulbs. I don't know the type of your led bulb but some are specifying 10 to 40V This might be the source of the flickering at lower speeds, a voltage under the minimum specified, the capacitor can do nothing about it. You might need a led bulb that works at lower voltage or a higher voltage from your generator.

  • \$\begingroup\$ Thanks again for spending time to deal with my problem Dorian. No, it's not a 6V, it is a 12V, 8 coil stator. So when idling (without the rectifier) the voltage is 12-13VAC and at high RPMs it goes up to 60VAC! With the rectifier/regulator in, it stays at 12,5VDC but with tons of ripple voltage! I'm sure this is the cause of the flickering, because when I put a smoothing capacitor after the regulator everything is fine. I'm just trying to find if the capacitor draws too much constant current from the alternator. But Janka claims it doesn't, it's only the spikes that matter and I tend to agree \$\endgroup\$ – Nick G. Apr 5 '18 at 14:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.