Maker Pro
Maker Pro

JFET Power Amplifier With Output Transformer

Audioguru

Sep 24, 2016
3,656
Joined
Sep 24, 2016
Messages
3,656
EEK!
I am a hifi guy. I hate distortion if it is produced by clipping or class-B crossover, even harmonic or odd harmonics. You are even cutting low frequencies which makes the distortion sound louder and you have an awful output transformer that mutes high and low frequencies and has its own distortion. Your circuit does not even have any overall negative feedback so the distortion is all over the place. A buzzer is not a musical instrument.

Ban acid rock and bring out the pleasant sounding acoustic guitars, played through a modern amplifier design that has a wide bandwidth and distortion so low that it cannot be heard and is very difficult to measure.
 

rogerk8

Jul 28, 2011
179
Joined
Jul 28, 2011
Messages
179
Thank you for your input!

It is nice when anyone says anything.

However, I do not agree with you because my philosophy is that "Good Enough" is what is important, good figures on for instance THD or SNR does not impress me the least because it is ONLY if it is audible that it is interesting, otherwise the measured data is just totally meaningless.

You don't need to have a system with perfect parameters to have fun.

Best regards, Roger
PS
I am kind of tired of HiFi, I say as ZZ Top "HiFi, LoFi, NoFi is just fine with me"
 
Last edited:

Audioguru

Sep 24, 2016
3,656
Joined
Sep 24, 2016
Messages
3,656
Hi frequency hearing loss is normal for old people (men over 30) and for people who ruined their hearing with loud acid rock sounds, guns or loud motorcycles. Then they use way too much distortion because they do not hear all of it.

"Good Enough"? An AM radio or an old wired telephone is good enough for people who ruined their hearing.
I am 73 and my hearing aids make my hearing as good as when I was 18 and I can hear any awful distortion that is there. I enjoy listening to the music, not interfered with by the faults of the circuits.
 

Attachments

  • hearing loss.png
    hearing loss.png
    92.8 KB · Views: 51

rogerk8

Jul 28, 2011
179
Joined
Jul 28, 2011
Messages
179
Interesting graphs!

And I'm glad that you can hear as if you were 18, good for you!

In this perspective I kind of understand you better because normally you can't hear HiFi but with you hearing aid you can, right?

This must be very exciting for you and I really congratulate you!

However music is about music and not HiFi.

I think you can agree with me on that.

Many people who are audiophiles and loves extremely expensive equipment have lost there sense of what music really is, they measure it in dB not in feelings.

Moreover, many audiophiles have no taste in music whatsoever!

I am trying to design a guitar amplifier that clipps nicely and I think I have found a way to make that happen but what I wish to say with this is that this amp need not be able to reproduce frequencies below about 150Hz because the string on the guitar can not swing below that (sub harmonics are bullshit in this aspect).

So in regard to a HiFi-amp where you would need the transformer to withstand some 10Hz (at least) at full power my transformers of some 100Hz at full power is indeed bad.

It is useless for HiFi.

BUT it is useful for a guitar amp.

And this is what I mean with "Good Enough" because you need not have all the parametes correct, you do only need to have the parameters that MATTERS (and are audible) correct.

Best regards, Roger
PS
73 years old that is impressive!

Picture text: 5V/DIV, 1kHz, RL=4 Ohm (13W as shown)

DSCN4001.JPG
 
Last edited:

Audioguru

Sep 24, 2016
3,656
Joined
Sep 24, 2016
Messages
3,656
In this perspective I kind of understand you better because normally you can't hear HiFi but with you hearing aid you can, right?
Correct. I did not know that I was normally deaf to high frequencies. The hearing aids made a remarkable good change to my life.

I am trying to design a guitar amplifier that clipps nicely.
An amplifier that is clipping produces squarewaves. A buzzer also produces squarewaves but not music. An amplifier with vacuum tubes had to use an output transformer but amplifiers have not used an output transformer for about 60 years.

BUT it is useful for a guitar amp.
You are trying to make sounds that were normal to deaf people 60 years ago.
Your waveform shows about 30% distortion, pretty bad. Most people dislike 1% distortion.
 

rogerk8

Jul 28, 2011
179
Joined
Jul 28, 2011
Messages
179
I am not finished, the waveform will be corrected by use of some 12dB of feedback.

I am just starting up the system and is doing some preliminary tests.

I don't know if I have said this but I had some problems with VLF and HF oscillations which are now cured.

Finally I am right now winding a new small toroidal transformer which I think will work well, the core saturation frequency for this iron powder core (the other is a ferrite) is around 100Hz at full output power.

Have you ever listened to a push-pull tube amp while being slightly overdriven?

I love that sound because the distorsion never really gets bad, I think you can boost output power almost 3dB before audio is too distorted.

So if you have a tube amp of 6W you can boost it to at least some 10W using pp-clipping by transformer, this while of course hearing that it does not sound as it should BUT it is not that disturbing to the ear.

Best regards, Roger
DSCN4003.JPG
 

rogerk8

Jul 28, 2011
179
Joined
Jul 28, 2011
Messages
179
Today I applied some 12dB of feedback around my toroidal ferrite transformer.

The result was amazing!

The look of the signals is almost perfect.

I am attaching three pictures that show the maximum "pure" amplitude of 1kHz at 5V/DIV, a slightly overloaded version and a more overloaded version.

As you can see maximum output power is some 24Vpp over 4 Ohms which means 12Vp^2/2/4=18W.

The only thing that bothers me is that this core saturates at some 500Hz at maximum output level, on the other hand you do not always play at maximum level so at half the level you have 250Hz as saturation frequency, yet this is not really acceptable when the strings on the guitar can give a minimum of 150Hz, it is better to have core saturation below that frequency.

So I am winding the transformer above to switch to, core saturation is some 100Hz so in that regard it is much better BUT the copper loss is twice as high (which may not matter but I don't know).

Anyway, I am very pleased with these results because it means that I haven't thought completelly wrong.

Best regards, Roger
DSCN4005.JPG
DSCN4006.JPG
DSCN4004.JPG
 

Audioguru

Sep 24, 2016
3,656
Joined
Sep 24, 2016
Messages
3,656
Your clipping is soft like an old vacuum tubes with output transformer amplifier, maybe because the high frequencies are missing.

Since you like distortion, haver you ever built a Hendrix Fuzz Face distortion-making pedal circuit?

55 years ago I built a kit vacuum tubes stereo amplifier (25W per channel) and never had it clipping. It played an orchestra of many instruments well, never just one distorted geeetar.
 

rogerk8

Jul 28, 2011
179
Joined
Jul 28, 2011
Messages
179
Your clipping is soft like an old vacuum tubes with output transformer amplifier, maybe because the high frequencies are missing.
I missed to inform you guys that the new feedbacked (12dB only) bandwidth is 75kHz, so no there is not any problem with missing HF.

The soft clipping is due to the transformer (allowing only odd harmonics to excist).

Best regards, Roger
 

Audioguru

Sep 24, 2016
3,656
Joined
Sep 24, 2016
Messages
3,656
I missed to inform you guys that the new feedbacked (12dB only) bandwidth is 75kHz, so no there is not any problem with missing HF.

The soft clipping is due to the transformer (allowing only odd harmonics to exsist)
No.
Even harmonics are produced when the top half of the signal is not symmetrical with the bottom half like in an old room-heating class-A amplifier. Your signal's top and bottom halves are symmetrical so when clipping the distortion is odd harmonics (push-pull class AB). Some people think that even harmonics sound musical because they are exactly one octave apart (doubled frequency) but odd harmonics sound rough.

I think your soft clipping is partly caused by the output Mosfets, transformer and low open-loop gain produce a high output impedance. The output impedance of an amplifier determines its damping factor (to avoid a boomy and ringing sound) then the output impedance of a modern amplifier is 0.04 ohms or less.
 

rogerk8

Jul 28, 2011
179
Joined
Jul 28, 2011
Messages
179
I think your soft clipping is partly caused by the output Mosfets, transformer and low open-loop gain produce a high output impedance. The output impedance of an amplifier determines its damping factor (to avoid a boomy and ringing sound) then the output impedance of a modern amplifier is 0.04 ohms or less.

My open loop gain is about a 100, it is not that low.

BF is some 4 so BF+1 is some 5 which is the factor that reduces the output impedance and thus gives the damping factor.

Output impedance (while driven by pure current sources like MOS) is actually the loudspeaker impedance, that is some 4 Ohms, divided by BF+1 this gives some 4/5=0,8 Ohm, related to my 4 Ohm speaker the damping factor is some 4/0,8=5 or 14dB.

14dB of damping factor is more than enough, at least I hope so :)

Yet it is a damping factor separated from none.

Best regards, Roger
PS
Too much feedback destroys music.
 
Last edited:

Audioguru

Sep 24, 2016
3,656
Joined
Sep 24, 2016
Messages
3,656
Most excellent audio amplifiers use or are built like an opamp with an open-loop gain of 200,000 to a few million. Then when negative feedback is added to reduce the gain to about 30, the distortion is so low it cannot be heard and is difficult to measure. Audio opamps (OPA134 distortion is 0.00008%) and audio amplifiers have distortion at 0.003% or less and sound perfect. The phase shift of your output transformer is too much for much negative feedback.

The damping factor of your amplifier is so low then a speaker resonates, sounds boomy and rings with a poor transient response.
 

Attachments

  • Crown amplifier damping factor.png
    Crown amplifier damping factor.png
    24.6 KB · Views: 24
  • OPA2134 details.PNG
    OPA2134 details.PNG
    17.2 KB · Views: 19
  • LM3886 clipping distortion.PNG
    LM3886 clipping distortion.PNG
    29.5 KB · Views: 19

rogerk8

Jul 28, 2011
179
Joined
Jul 28, 2011
Messages
179
Most excellent audio amplifiers use or are built like an opamp with an open-loop gain of 200,000 to a few million. Then when negative feedback is added to reduce the gain to about 30, the distortion is so low it cannot be heard and is difficult to measure. Audio opamps (OPA134 distortion is 0.00008%) and audio amplifiers have distortion at 0.003% or less and sound perfect. The phase shift of your output transformer is too much for much negative feedback.

The damping factor of your amplifier is so low then a speaker resonates, sounds boomy and rings with a poor transient response.
The use of OP:amps is overrated.

The only thing they are good at is exact (DC) amplification, it is almost provable that their extreme feedback is awful for sound reproduction. Of course it looks good on the paper but tube amps has been with us for much longer times than OP:amps and they sounded perfectly for half a century ago so why would OP:amps with an enourmous amount of unneccesary feedback sound better today?

Best regards, Roger
If the SR of an OP:amp is not fast enough and you inject a sharp step into the OP:amp, the OP:amp may actually be destroyed due to the difference in spike levels. So transients using OP:amps...
 
Last edited:

Audioguru

Sep 24, 2016
3,656
Joined
Sep 24, 2016
Messages
3,656
Modern audio opamps (not a lousy old 741 or LM324) are perfect tor sound. Tube amps had a low open loop gain and not much feedback because the output tubes and output transformer had a high phase shift therefore a poor damping factor and distortion 100 times more than modern opamps when new but their performance dropped each day as they wore out.

Lousy old opamps had a poor slew rate for audio but modern audio opamps work perfectly up to radio frequencies.

Your guitar amplifier makes unusual noises (distortion and resonances) and not good music.
 

rogerk8

Jul 28, 2011
179
Joined
Jul 28, 2011
Messages
179
Your guitar amplifier makes unusual noises (distortion and resonances) and not good music.
Good music has nothing to do with dB or poor damping factor because it isn't really that critical.

Finally I don't se the resonances you claim.

Best regards, Roger
PS
"If you can't hear it, why go for it"
 
Last edited:

Audioguru

Sep 24, 2016
3,656
Joined
Sep 24, 2016
Messages
3,656
Decibels are used to measure noise (hum and hiss) and frequency response. A good amplifier has no noise and a response that is flat from 20Hz to 20kHz. 0.1% distortion is at -80dB.

An undamped speaker sounds like a bongo drum and also has higher frequency noises.
A well damped speaker does exactly what the amplifier tells it to do, it plays what and when it should and stops when it should.

"If you can't hear it" then you are deaf. I like to feel my body vibrating from a sub-woofer and like the "sizzle" of very high frequencies that many people cannot hear.
 

rogerk8

Jul 28, 2011
179
Joined
Jul 28, 2011
Messages
179
My KJA is not a HiFi-amplifier, that was never my intension (but can obviously be with a better OPT).

I'm curious of one thing, my KJA is a Class B amplifier YET there is no cross-over distorsion :D

Bets regards, Roger
PS
Why is it that most musicians prefere tube amps?
 

rogerk8

Jul 28, 2011
179
Joined
Jul 28, 2011
Messages
179
This is the way my system looks right now.

Please observe how small the OPT is.

The coin is a swedish 5kr but it is almost the size of 10pence.

Best regards, Roger

DSCN4013.JPG
 

Audioguru

Sep 24, 2016
3,656
Joined
Sep 24, 2016
Messages
3,656
Musicians prefer a hifi amplifier that is wideband and has very low distortion.
Electric guitar players prefer the muffled old distortion of a tubes amplifier so the noises sound like 60 years ago.

Your schematic in the first post shows a class-AB amplifier with the output devices biased so that they never turn off until the signal has crossed over to the other device.
 
Top