Rectified voltage a lot higher than expected

[hojnikb]

So i have this moped, which i'm fully converting to DC voltage.

Originally, everything was powerred by AC voltage (lights, siren, etc..) but in effort to modernise and add digital gauges, i've decided to rectify AC input to DC and power everything this way.

Without any regulation, it produces 10-40VAC (idle to max rpm).
Now my plan was to use a full bridge rectifier and convert whatever it spits out to 12VDC using this
https://www.aliexpress.com/item/1005002295226200.html?spm=a2g0o.order_list.0.0.7b461802j3KKmZ 13-60V version.
Theoretically, this should work fine (40V*1.41=~56,4V). But in practise, i've notice i'm getting more than double the voltage on the DC side of the rectifier, even though measured AC voltage is half (for 10VAC i get ~20-22VDC).

What am i missing here? Would a smoothing capacitor help ?
Could secondary ignition winding cause issues (this setup has one low voltage winding for lights and one high voltage for ignition).

[eugene]

Without load, when AC is rectified and smoothed with a capacitor, the cap will charge to the peak AC voltage. If the AC is a pure sine wave, then the peaks will be 1.41x what your DMM says the AC voltage is. OTOH, if it's not a pure sine wave but has higher peaks, then the cap on the DC side of the rectifier will charge to those peaks. This is true whether your voltmeter is true RMS or not.

It's likely that as soon as there's a load on the circuit the DC voltage will drop. That's true even with a sine wave, but it will likely drop even faster if your AC waveform is peaky. 

[Doctorandus_P]

On top of the rectification rms / peak levels...

Specifications for regular old fashioned sheet metal transformers are under full load. at a low load the output voltage is always higher. this can be between a few percent for big transformers, to 30% or even more for small transformers.

[hojnikb]

Without load, when AC is rectified and smoothed with a capacitor, the cap will charge to the peak AC voltage. If the AC is a pure sine wave, then the peaks will be 1.41x what your DMM says the AC voltage is. OTOH, if it's not a pure sine wave but has higher peaks, then the cap on the DC side of the rectifier will charge to those peaks. This is true whether your voltmeter is true RMS or not.

It's likely that as soon as there's a load on the circuit the DC voltage will drop. That's true even with a sine wave, but it will likely drop even faster if your AC waveform is peaky.

I not sure if it's pure sine wave (dont have a scope) but i would assume it is, since it's a alternator type power generation. Frequency in my case is around ~400-1,500Hz, depending on rpm.
It's really interesting to see such a difference in voltage between AC and rectified DC in this instance.

If i take the same rectifier and hook it up to a 10.5VAC desktop lamp (give or take), i get the expected ~13,5-14V on the DC side.

[hojnikb]

On top of the rectification rms / peak levels...

Specifications for regular old fashioned sheet metal transformers are under full load. at a low load the output voltage is always higher. this can be between a few percent for big transformers, to 30% or even more for small transformers.

I don't have a transformer or it isn't part of this equation. I'm rectifying AC generated by the "alternator" and measuring the voltage on the DC side of the rectifier. And that DC voltage is higher than expected, as it's 2-2-2x instead of the usual ~1,4x.  This is the wierd part (for me anyway).

Circuit actually works just fine, but i'm not testing it under high rpm, due to risk of blowing up the DC-DC converter.

[hojnikb]

And just for reference, this is how the AC voltage is generated.
https://www.mopeddivision.com/moped-make-model/puch/electrical/puch-e50-12v-cdi-stator-magneto-box-with-spark-wire-boot/

[Zero999]

The alternator obviously doesn't output a sine wave.

Does your meter read the true RMS voltage?

Either way, the rectified DC will not equal V_RMS√2, as it's non-sinusoidal.

Buy a DC-DC converter which will run off the full voltage range.

[Terry Bites]

The windings share a common ground so its unlikely that the ignition winding is responsible for your observation.

The magneto uses separate sectors of the shaft rotation for different windings- this gives a wierd AC waveform.
You need a very good meter to extract the actual RMS from the waveform. The output from a bridge rectifier will be pulsating DC.
The DMM will fail to measure this pulsating volatge correctly. It would seem that its a DMM problem rather than physics.
It doesn't help that magnetos have no regulation.

I have seen AC to DC regulators used on mopeds. You wont need a second wasteful pre-rectifier eg LVC-12ACDC-5A

Have a look at https://www.mopedarmy.com/forums/read.php?1,3827317

How is your bridge wired, not like this...? Try not to upset Mehdi.

[hojnikb]

The windings share a common ground so its unlikely that the ignition winding is responsible for your observation.

The magneto uses separate sectors of the shaft rotation for different windings- this gives a wierd AC waveform.
You need a very good meter to extract the actual RMS from the waveform. The output from a bridge rectifier will be pulsating DC.
The DMM will fail to measure this pulsating volatge correctly. It would seem that its a DMM problem rather than physics.
It doesn't help that magnetos have no regulation.

I have seen AC to DC regulators used on mopeds. You wont need a second wasteful pre-rectifier eg LVC-12ACDC-5A

Have a look at https://www.mopedarmy.com/forums/read.php?1,3827317

How is your bridge wired, not like this...? Try not to upset Mehdi. (Attachment Link)

I have one of these bridge rectifiers
https://www.amazon.com/10PCS-D10XB60-Bridge-Rectifier-Original/dp/B07KX63VL1

AC input connected to the magneto winding and DC +- is connected to the DC-DC converter i've mentioned above (accepts 13-60v and regulates to 12V).

So if i understand this correctly, because of the wierd waveform my magneto is making, the multimeter gets tripped out and measures DC voltage on the rectifier "incorrectly"?

[Zero999]

Did you have a smoothing capacitor? If not, your meter will show the average voltage. The peak will be even higher.

[hojnikb]

Did you have a smoothing capacitor? If not, your meter will show the average voltage. The peak will be even higher.

Nope, no smoothing capacitor for now. I just wanted to test the whole deal.

As for previous post; Is there even a DC-DC voltage converter, that works in such voltage range?

Or would a better bet be to use a transformer, that stepds down the AC voltage, so it's within the voltage range of the DC-DC converter?

[Zero999]

Did you have a smoothing capacitor? If not, your meter will show the average voltage. The peak will be even higher.

Nope, no smoothing capacitor for now. I just wanted to test the whole deal.

As for previous post; Is there even a DC-DC voltage converter, that works in such voltage range?

Or would a better bet be to use a transformer, that stepds down the AC voltage, so it's within the voltage range of the DC-DC converter?

You need to add a smoothing capacitor to measure the voltage properly. There's no point in addressing your other questions, until you have an accurate voltage measurement.

[BeBuLamar]

When you add the filter capacitor the voltage would actually rise further in no load condition.

[mclute0]

Guessing, since you did not say what voltage meter you are using ... IMHO, your stated DC Frequency of ~400-1,500Hz may be high for your multimeter.

RMS voltage is an average. A cheap meter will be totality wrong in assuming 50/60hz, a more expensive cheap meter may do ~400+hz and will integrate to report fake true RMS above its max and be wrong. Some DMM are just badly programed and might have a sufficient ADC to read at high frequencies but not programmed to actually do it.

Reducing frequency with a properly sized capacitor will be a lot closer to a true RMS voltage measurement.

[Zero999]

When you add the filter capacitor the voltage would actually rise further in no load condition.

In theory yes, but we don't know if his meter is measuring the voltage properly. If the high frequency harmonics are interfering with it, then it could actually measure a lower voltage.

[Uunoctium]

tl;dr
in my humble opinion an magneto acts same principle as an bicycle dynamo. Ist is an current source!
So without sufficient load, voltage tends to raise up to infinite
Please repeat your measurement with proper load connected ant tell us the result

[hojnikb]

tl;dr
in my humble opinion an magneto acts same principle as an bicycle dynamo. Ist is an current source!
So without sufficient load, voltage tends to raise up to infinite
Please repeat your measurement with proper load connected ant tell us the result

Yes, i think this is the right course of action. I'll get a filtering cap and some typical load (lightbuls are usually alway on) and see what happens on a good true rms multimeter.

If not, i think i'll just use the existing AC regulator that came stock and regulate that to DC just for the accessories. That seems like a much easier route.