Peak voltage

[John10b]

I should not have to ask this question, but if you don’t ask...!
Firstly my understanding is that with a half wave rectifier and capacitor the output voltage is almost RMS x ✔️2.
My readings using my basic digital meter is:
Output from secondary of Mains transformers is 226vac.
Output from rectifier with capacitor 210vdc @ 40mA.
So here is the question, why don’t I get a reading of about Peak which is about 326vdc?
This is the actual readings from a basic working audio amp using an EL84.
The rectifier is an IN5408 with a 32μF capacitor.

Cheers
John

[ms660]

DC out equal to AC peak in is only possible with no load current.

Lawrence.

[John10b]

Thanks Lawrence.
Cheers
John

[John10b]

Further to my previous post.
I replicated on the bench the exact circuit, but with a larger Mains transformer, giving out the same voltage.
With the secondary voltage out at 226vac, same as other transformer, IN5408 Rectifier and 32uF capacitor, with dummy load at current of about 50mA, which is about the same as original circuit.
Why then do I get an output of 280vdc, the original circuit only gave an output of 210vdc?

Cheers
John

[paulsherwin]

There is something amiss with your original setup. The smoothed DC voltage should be about 1.4 times the AC voltage in ideal circumstances. Your 280V is what I would expect.

[Lucien Nunes]

Your measurements may be in error because of the significant waveform distortion at the secondary terminals of the transformer. The transformer's equivalent resistance and leakage inductance form a divider with the load presented by the charging capacitor, which is grossly non-linear. Therefore the voltage waveform is no longer sinusoidal and the crest factor no longer sqrt(2). The larger transformer has lower effective source impedance and is less affected by the non-linear load.

The distortion is particularly acute with a solid state rectifier (which has low dynamic resistance, so it ties the secondary closely to the capacitor voltage when it is forward biased) and with half-wave rectification, where the capacitor must support the entire load for over half a cycle and then be recharged in less than half a cycle, and due to imperfect coupling can push the average flux in the transformer core away from zero.

What your meter makes of both the AC measurement (of a non-sinusoid) and the DC measurement (with significant ripple) will depend on what kind of meter and whether it is true-RMS reading. Scoping the voltages would probably be quite revealing.

Half-wave rectification is usually bad for multiple reasons. One of them is to reduce the useful VA rating of the transformer well below its nameplate rating, due to the much higher losses incurred by the pulsating DC load. Is there a reason you are not using full-wave?

[John10b]

Thank you both, Paul see post 2. I will investigate further when I have the opportunity.
Cheers
John

[snowman_al]

John Hi,
The statement in post no. 2 is correct.
If you try both set ups with no load you should see very similar results.
226 ac x 1.41 = 319 volts DC at the cap. (Assumes the caps are both the same in value and with no significant leakage of course.) If you do not, there is a problem somewhere.

Simplistically once you put a load on the end the 'resistance' of the transformer winding becomes significant. Just like putting a resistor in series with the load. The more resistance the lower the voltage at the load. So a bigger transformer has less 'resistance' therefore less voltage drop. Also it could have been designed to drop a few volts?

We do not know the relative sizes of the transformers to compare. But what Lucien says is also very important, 1/2 wave rectification only gives you approx. 30% of the AC current capability of the transformer.

The other factor is that 32uF is quite small, so again as Lucien says, the load may be such that the cap cannot charge to 'peak' voltage long enough and you end up measuring nearer the 'average' instead, which is about 90% of the RMS AC voltage. You could try adding another cap to the 32uF you have and see if that keeps the voltage up at all.

Phew, hope that helps a little, Alan

[paulsherwin]

I agree that half wave rectifying for a 40mA load is an odd decision. It's normally done where only a few mAs are required.

[John10b]

Let me expand.
The circuit, which is working fine, is from a basic EL84 audio amp from a Dansette Major delux 21. This has a small mains transformer.
The exact circuit I made up on the bench was identical except for dummy load and a mains transformer with same output but physically much larger.
I did some measurements purely for the record, as shown in my previous posts.
I’m wondering if my digital meter and my analog meter are giving false resdings.
Cheers
John
Ps I just ordered an AN8002.

[paulsherwin]

Just bung in a 4 diode bridge and see if anything changes.

[ms660]

I might be miles out but according to what I can figure out, Rs (resistance in series with the silicon rectifier) for your original set up (226 VAC, 210 VDC out at 40mA) is approx. 504 Ohms, and with your set up with the large transformer (226 VAC, 280 VDC out at 50mA) Rs is approx. 94 Ohms.

But I must admit I kinda got a bit confused with the workings out.

Lawrence.

[TrevorG3VLF]

Any resistance in series with the rectifier such as primary and secondary resistances and capacitor resistance will have a much bigger effect than a simple calculation will give.
If the pulse length is one fifth of the cycle than the resistance will appear to be five times its value and the dissipation will be five times that a simple calculation will give.
A high resistance will spread the pulse so this effect will not be so great. High power supplies often use a choke input filter to spread the pulse to almost constant current. Then you get average voltage, not peak.

[ms660]

It wasn't a simple calculation, well not for me anyways, still not sure that it's right though, maybe someone else more mathematically inclined can have a shot.

Lawrence.

[John10b]

Once again thank you all, this simple circuit is stimulating the brain cells, what’s left. I will investigate further and get back.
Cheers
John

[Lucien Nunes]

As I understand it, John's original query was not so much around why the DC voltage is low, but why the AC and DC readings taken on the same circuit did not seem to correlate. This, I suspect, is due to the way the meters respond to the waveform present.

I’m wondering if my digital meter and my analog meter are giving false resdings.

There's a difference between a false reading due to a defect in the instrument, and one that measures in a way that you do not want. For example, a moving-coil AC meter responds to the mean of the absolute value, but it's scaled in RMS as that is often the most useful way to indicate the magnitude of a sinusoid. However, if you present it with a non-sinusoid having a form-factor other than the 1.11 for which it is calibrated, the reading will not be the RMS value nor necessarily peak/sqrt(2). This is normal and predictable for a moving-coil instrument, although it might not be helpful for your application. Digital meters are more variable in the way they respond, so their readings can diverge with odd waveforms.

[John10b]

Lucien i was originally wondering why the dc out from rectifier, with capacitor, was not closer to the peak. This has been explained.
However I may have been deceived by using a combination of my very old digital and analog meters, which you and others have explained.
I’m going to redo my readings and record the difference I get between the meters.
I’m going to dig out my still older AVO 7, I have ordered a new meter which is supposed to read true rms, but this won’t be with me for a few weeks.
I’ll also get my scope working and take a look at waveforms.
Cheers
John

[G4_Pete]

Hi just to add to the variations if you do get a true rms meter the form factor of the waveforms under discussion will lead to different readings again. Best follow the oscilloscope path for peak voltages but take care with the earth loops and earth reference points!

Pete

[buggies]

Originally Posted by ms660

I might be miles out but according to what I can figure out, Rs (resistance in series with the silicon rectifier) for your original set up (226 VAC, 210 VDC out at 40mA) is approx. 504 Ohms, and with your set up with the large transformer (226 VAC, 280 VDC out at 50mA) Rs is approx. 94 Ohms.

I can't do calculations but fiddling around with some component values in PSU designer to give mean voltage and current gives series resistances of 480 Ohms and 75 Ohms so in same ballpark.
Ripple about 20-30 Volts p-p.

http://www.duncanamps.com/psud2/

[John10b]

Hi George, did you use two different transformers? I note the resistance of the secondary windings are different? I also not the slight difference in load resistance?
Cheers
John