It is about time I built a push pull amplifier

What people are working on at the moment
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8863
Joined: Mon May 21, 2007 9:42 pm

#106

Post by Paul Barker »

Yes the signal generator when puting out 1v rms produces distortion which varies between .7% and 1%.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#107

Post by Nick »

Hard to tell from the scope trace, but the -ve parts of the cycle look heavily distorted.
Now a friend of mine that is not an audiofool says if you can't see the distortion on a sign wave there isn't any.
Well, for some values of "isn't any"

Does he have a view on the sound of a tree falling in the woods?
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8863
Joined: Mon May 21, 2007 9:42 pm

#108

Post by Paul Barker »

Here is the sig gen directly into the distortion analyser.

Image

Yes the bottom half in the picture is out of shape..

Will have to wait for better source first.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
Andrew
Eternally single
Posts: 4206
Joined: Thu May 24, 2007 2:18 pm

#109

Post by Andrew »

What about an inexpensive sound card and a copy of Audactity, most sound cards will be better than than 1%, much better.

Next get some BNC and a selection of adaptors and ditch the clip leads...sorry Paul I know you're fond of the pld clip leads, so am I, but they're not ideal, you'll be picking up all sorts of environmental crap.

I can bring some test gear to Owston, if anyone wants to do get a benchmark done... :?:

Andrew
Last edited by Andrew on Sun May 26, 2013 6:49 pm, edited 1 time in total.
Analogue, the lost world that lies between 0 and 1.
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8863
Joined: Mon May 21, 2007 9:42 pm

#110

Post by Paul Barker »

Nick wrote:Hard to tell from the scope trace, but the -ve parts of the cycle look heavily distorted.
Now a friend of mine that is not an audiofool says if you can't see the distortion on a sign wave there isn't any.
Well, for some values of "isn't any"

Does he have a view on the sound of a tree falling in the woods?
Of course I am not listening at anything like those levels so I can't here the 10% distortion as my ears would be bleeding. Even 1 watt is too loud to me. 1/2 a watt for Diana. It would sound very nasty at Owsten though wouldn't it?

We'll wait until we can put a clean sign wave into it before we dismiss it completely.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
Mike H
Amstrad Tower of Power
Posts: 20157
Joined: Sat Oct 04, 2008 5:38 pm
Location: The Fens
Contact:

#111

Post by Mike H »

Nick wrote:Hard to tell from the scope trace, but the -ve parts of the cycle look heavily distorted.

FWIW I was thinking that too. This is the sort of thing I look for in LTspice traces.

Re the sig distortion oh yes, I was staggered how much distortion there is in a computer generated sine, which is supposed to be like a digital file after all.

IIRC what you have to do is first measure the sig gen, make a note, then measure the amp (+ sig gen), subtract the first number. I gave up after finding the difference between the two (the actual distortion of the amp) is so small that I can't really see it. So still don't knwo what it is accurately.

Don't forget also THD reading on the meter also includes noise and hum. Unless you have filters to take all that stuff out.


HTH
 
"No matter how fast light travels it finds that the darkness has always got there first, and is waiting for it."
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8863
Joined: Mon May 21, 2007 9:42 pm

#112

Post by Paul Barker »

So I only have 9% distortion; single figures, yey :wink:

Bring your guitars boys :D
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8863
Joined: Mon May 21, 2007 9:42 pm

#113

Post by Paul Barker »

I knew I should never have bought a distortion analyser.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#114

Post by Nick »

I was staggered how much distortion there is in a computer generated sine, which is supposed to be like a digital file after all.
Just how were you generating it? The PC is the lowest distortion source I have by far.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
Neal
Shed dweller
Posts: 2299
Joined: Fri Aug 03, 2007 10:57 am
Location: From the land of the Bodgers

#115

Post by Neal »

If its of any consolation Paul, commercial figures for distortion are usually measured at 1W output at 10Khz and also 1Khz into either an 8 or 4ohm load...

Drive the amp input to get 1W out, measure the sig gen distortion at that input voltage and take it away for the amps output at 1W...it wont seem so bad. :wink:

The distortion will be mostly second harmonic I would think which is not too bad if kept under ~3% or so...third harmonic is nasty and you'll hear it as a glassy edge and sharpness if much above 1%
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8863
Joined: Mon May 21, 2007 9:42 pm

#116

Post by Paul Barker »

Neal wrote:If its of any consolation Paul, commercial figures for distortion are usually measured at 1W output at 10Khz and also 1Khz into either an 8 or 4ohm load...

Drive the amp input to get 1W out, measure the sig gen distortion at that input voltage and take it away for the amps output at 1W...it wont seem so bad. :wink:

The distortion will be mostly second harmonic I would think which is not too bad if kept under ~3% or so...third harmonic is nasty and you'll hear it as a glassy edge and sharpness if much above 1%
Thank you Neal. I'll do that later today. I have the glassy edge and sharpness compared to SE:? But I like the sound at my listening levels, because the bandwith assessed auraly is better and it has more power than my little 750mW SE amp. You can't achieve with it the elements of the SE amp I described earlier in the thread, but as said later now the LDR attenuator is controling volume it is coming good. It is not an amplifier to throw out with the bath water.

I was hoping that up to the output from the interstage transformer I have a Phil style module from which I could drive a variety of output stages both SE and PP. So when it gets built properly it shall be modular. I can't see the CF producing any of this distortion. The VA stage would definately clip, but shouldn't clip at the voltage swings required for this amp. I have to remain open to change on this first stage, to could be taken away and replaced with a pentode for instance, as almost anything will drive the CF. Perhaps the VA will become a GG pairing... then out of the GG into the CF driver valve.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8863
Joined: Mon May 21, 2007 9:42 pm

#117

Post by Paul Barker »

OK got more sensible results. Did as Neal suggested and checked input distortion at each level change deducted it. It mainly stayed at 0.6% just 0.55% at lowest possible setting.

I changed the Dummy load to 8 ohm because I am using the 8 ohm tap, so less distortion now. Just shows you how much that matters.

Very interesting linear progression. Pick off your power output at the amount of distortion you want. Could this be mostly 2nd harmonic generated in the VA stage? Something which I wanted to create from the outset?

Image

I think Fritz Langford Smith works to 5% distortion which would give 3.33 watts. That makes plenty of sense. But notice distortion doesn't get out of hand until after 6 watts. I recall earlier in the thread calculating 6 watts power output for the 6V6. This valve is not a 6V6, but it is betwixed that and the 6L6. With my filament starvation it is more on the 6V6 scale than the 6L6 according to these results. It is not a bad valve. If you see a batch going cheap get them. I note they are all over the pond and I bought mine before the rest of the world decided to start useing them.

Further more the listening levels I employ at home it has no distortion at over 1 watt. that is why it sounds so much less distorted than the 750mW SE amp. The reason being the SE amp has the added 2nd harmonic of the output stage.

Now I am sitting happy with this amp.

Future possibility is as mentioned less distortion from the VA, by various strategies.

Wouldn't it be interesting to do the same tests on an SE 300b amp?
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#118

Post by Nick »

Wouldn't it be interesting to do the same tests on an SE 300b amp
Been there, done it, can't find the graphs at the moment though.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8863
Joined: Mon May 21, 2007 9:42 pm

#119

Post by Paul Barker »

Nick wrote:
Wouldn't it be interesting to do the same tests on an SE 300b amp
Been there, done it, can't find the graphs at the moment though.
the only thing is I think the 300b would have the advantage because the 2nd harminic it creates is mitigated by the 2nd harmonic of the VA. Whereas in the PP output stage all of the VA distortion passes through.

The solution is lessen the VA distortion, or use a cascade. But part of the objective is to produce some 2nd harmonic. Just not quite so much.

Because of the interstage transformer I can't really apply global feedback in moderation, I already have the benefit of degenerative feedback with the unbypassed cathode resistor. I can't bootstrap it as that takes away too much of the gain. I am left with a serious rethingk. I always knew the VA stage of the 6em7 was it's limitation, I have just always been pig headed about not wasting it. But it may have to go in this instance. But probably after Owsten. every weekend is fully booked and I am full time in the week. Plus only driver in house, Sam with broken leg, blah de blah de blah. But I don't have to walk feed and toilet 30 lurchers :shock:
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8863
Joined: Mon May 21, 2007 9:42 pm

#120

Post by Paul Barker »

I suppose chasing away the distortion might entail something like DHT mu 10 or so vA then replace Cf with another of these 1619's as a triode right way up. That kind of thing. Then we'll show those fully balanced jockeys a thing or two. :D :D :D

This is a good platform to build on and learn from. It can improve and the transformers can be altered for niche ones later. But these sound fine as it is.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
Post Reply