Remember These?

What people are working on at the moment
User avatar
pre65
Amstrad Tower of Power
Posts: 17946
Joined: Wed Aug 22, 2007 11:13 pm
Location: North Essex

Re: Remember These?

Post by pre65 » Fri Nov 22, 2019 11:17 am

Cressy Snr wrote:
Fri Nov 22, 2019 9:19 am
I’m off to do some creative writing.
In the style of Alan Davies ? :lol:
Life is what happens to you while you're busy making other plans. John Lennon

G-Popz THE easy listening connoisseur. (Philip)

Cressy Snr
Social Sevices have been notified
Posts: 8693
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

Re: Remember These?

Post by Cressy Snr » Fri Nov 22, 2019 11:27 am

No, Ian Mc’Millan is more my style.

I’m only Alan Davies on this forum. :mrgreen:
On every tree there sits a bird,
Singing a song ...of love...

User avatar
Nick
Site Admin
Posts: 12152
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

Re: Remember These?

Post by Nick » Fri Nov 22, 2019 11:35 am

Not sure I would trust anything that calls them self an Institute when they are not. The restriction on NFB in a valve amp is the transformer, which oddly a OTL doesn't have, so al the above is not of any relevance to Steve. The other effect that NFB has of course is to reduce output impedance, which is of great interest to a OTL.

Oh, and from that link, where he feels it worth referring to Einstein, "Physics is the science of measurement." No its not Metrology is.
Resistance isn't futile it's V / I.

Wolfgang
User
Posts: 174
Joined: Mon May 07, 2018 3:08 pm

Re: Remember These?

Post by Wolfgang » Fri Nov 22, 2019 3:21 pm

The restriction on NFB in a valve amp is the transformer, which oddly a OTL doesn't have, so al the above is not of any relevance to Steve.
That's definitely a very interesting view. So, in order for me to understand this better, do you want to say with your general comment that:

NFB in OTLs doesn’t have to be less than 180 degr.out of phase with the input signal which limits its effectivity to reduce distortion?
All frequencies of the audio frequency range are treated equally by one NFB setting in OTLs and only the OPT makes it impossible to reduce audible distortion equally and effectively ? If this is the case how do you explain the dramatic change in sound character like warmer midrange, less clinical sound in the treble, if the NFB is reduced in OTLs? I am not talking about the "warmer bass" around resonance frequency caused by less output impedance.
The SQ of OTLs is the same with and without NFB?

And if Physics – as all exact sciences - wouldn't have to prove its postulates and theories by exact measurements in order to achieve accurate quantitative expression how else would it get there?

User avatar
Nick
Site Admin
Posts: 12152
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

Re: Remember These?

Post by Nick » Fri Nov 22, 2019 4:03 pm

Wolfgang wrote:
Fri Nov 22, 2019 3:21 pm
The restriction on NFB in a valve amp is the transformer, which oddly a OTL doesn't have, so al the above is not of any relevance to Steve.
That's definitely a very interesting view. So, in order for me to understand this better, do you want to say with your general comment that:

NFB in OTLs doesn’t have to be less than 180 degr.out of phase with the input signal which limits its effectivity to reduce distortion?
All frequencies of the audio frequency range are treated equally by one NFB setting in OTLs and only the OPT makes it impossible to reduce audible distortion equally and effectively ? If this is the case how do you explain the dramatic change in sound character like warmer midrange, less clinical sound in the treble, if the NFB is reduced in OTLs? I am not talking about the "warmer bass" around resonance frequency caused by less output impedance.
The SQ of OTLs is the same with and without NFB?

And if Physics – as all exact sciences - wouldn't have to prove its postulates and theories by exact measurements in order to achieve accurate quantitative expression how else would it get there?
Not really a interesting view, just the correct one. Look at the ACF that Phil was asking about, a valve circuit that works by using loads of feed back, and it works because there is no group delay that will cause the phase response of the circuit to lag 180 deg before the gain drops below 0dB.

I have no interest getting into a discussion about subjective sounds you may have heard with a amp I have no experience of. But "The SQ of OTLs is the same with and without NFB?" Of course not, negative feedback will change the distortion spectra and the output impedance, so I would expect it will sound different.

As to Physics, yes of course experimental proof of theories concerns its self with accurate measurement, as does Chemistry and Biology, so yes, measurement is important to all science, but that doesn't make "Physics the science of measurement". I told you what the science of measurement was called.

As it happens Einstein was a very poor experimental physicist, which was not a problem as he was a theoretical physicist. SR, ad GR were theories that were not originally derived by experiment, that's one of the wonderful things about them. They were later tested against accurate measurement, but Einstein didn't perform those experiments, others did. His Nobel on the photoelectric effect again was the result of explaining the experimental results of others he was not one for the lab coat.
Resistance isn't futile it's V / I.

Cressy Snr
Social Sevices have been notified
Posts: 8693
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

Re: Remember These?

Post by Cressy Snr » Fri Nov 22, 2019 4:46 pm

Here is a text grab from Morgan Jones “Valve Amplifiers”
This is the measurement data produced, with and without feedback, by a 6528 single ended amplifier he built
3403929C-D9BB-422A-9F69-2B66C0260381.jpeg
He added a small amount of cathode feedback and the distortion was reduced across the harmonic spectrum.
I strongly suspect that the same thing is happening with my SEOTL, but the effect is greater as the feedback is stronger.
That negative feedback reduces Zout I thought was a given.
On every tree there sits a bird,
Singing a song ...of love...

User avatar
Nick
Site Admin
Posts: 12152
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

Re: Remember These?

Post by Nick » Fri Nov 22, 2019 5:11 pm

That negative feedback reduces Zout I thought was a given.
Not sure why its any more or less of a given than reducing distortion as they are both part of the same process.
Resistance isn't futile it's V / I.

Cressy Snr
Social Sevices have been notified
Posts: 8693
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

Re: Remember These?

Post by Cressy Snr » Fri Nov 22, 2019 5:56 pm

Nick wrote:
Fri Nov 22, 2019 5:11 pm
That negative feedback reduces Zout I thought was a given.
Not sure why its any more or less of a given than reducing distortion as they are both part of the same process.
Well that’s what I thought I meant. It’s just that one was somehow being separated from the other by the argument.

I suppose if we are wanting to strictly define what happens with negative feedback, we could say that an ideal negative feedback signal doesn’t “reduce distortion” ..depends what we mean by “reduce” It’s the same distortion, but shifted to a lower level relative to the output signal, so 1% becomes 0.1%, becomes 0.01% by that mechanism.

So negative feedback widens the signal to distortion ratio and the signal to noise ratio might be the better definition, or am I being thick?
On every tree there sits a bird,
Singing a song ...of love...

User avatar
Nick
Site Admin
Posts: 12152
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

Re: Remember These?

Post by Nick » Fri Nov 22, 2019 8:17 pm

Cressy Snr wrote:
Fri Nov 22, 2019 5:56 pm
So negative feedback widens the signal to distortion ratio and the signal to noise ratio might be the better definition, or am I being thick?
I don't see how "reducing distortion" is a bad description. Ok, you could argue that the distortion is still there but the input is mixed with a special signal that magically cancels some of the distortion. You could equally argue that the output impedance of the amp remains the same, its just the signal is again magically altered to cause the combination of the original (and remaining) output impedance when driving the load to behave as if it was a source with a lower output impedance.

But IMHO, that’s all just playing with words when lowering the distortion and output impedance is perfectly fine.
Resistance isn't futile it's V / I.

Cressy Snr
Social Sevices have been notified
Posts: 8693
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

Re: Remember These?

Post by Cressy Snr » Fri Nov 22, 2019 8:46 pm

Nick wrote:
Fri Nov 22, 2019 8:17 pm

But IMHO, that’s all just playing with words when lowering the distortion and output impedance is perfectly fine.
That was the description I had been perfectly happy with too.
On every tree there sits a bird,
Singing a song ...of love...

User avatar
Mike H
Amstrad Tower of Power
Posts: 18624
Joined: Sat Oct 04, 2008 5:38 pm
Location: The Fens
Contact:

Re: Remember These?

Post by Mike H » Fri Nov 22, 2019 9:40 pm

Nick wrote:
Fri Nov 22, 2019 4:03 pm
I have no interest getting into a discussion about subjective sounds you may have heard with a amp I have no experience of.
Image

My bwain is starting to 'urt!

:D

Meanwhile, back to the 13E1 — yes fab looking valve! :thumbleft:
 
"Entrenched convictions cannot be changed by facts." – Dick Francis

User avatar
jack
Social outcast
Posts: 3400
Joined: Wed Dec 29, 2010 8:58 pm
Location: No longer living somewhere stupidly hot...
Contact:

Re: Remember These?

Post by jack » Fri Nov 22, 2019 9:53 pm

Interesting. My understanding of NFB is that it relies on a constant180° phase shift, so if the DUT is not phase linear across the working bandwidth (which pretty much no power amp is), distortion will be introduced....

Phase linear amps are very specialised and expensive...
Nick
Ars longa, vita brevis

User avatar
Nick
Site Admin
Posts: 12152
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

Re: Remember These?

Post by Nick » Fri Nov 22, 2019 10:28 pm

distortion will be introduced
I agree, but I would say that distortion reduction would be reduced as the phase shift moves away from 180. Look at a typical SS amp, distortion will rise with frequency which is a combination of the group delay shifting away from 0 deg and the gain failing with frequency. But if a amp is not generally phase linear over the range required IMHO its broken in several ways without feedback making it worst. The effect of a transformer is to introduce such a frequency dependent term, hence the valve amps can't use much feedback. The Williamson has a specially designed output transformer to allow 20dB of feedback to be used. Of course that was in the days where previously interstage would have been used ruling the use of global feedback out altogether. Of course the phase shift in a amp is in itself a form of distortion.

WRT output impedance, there were amplifiers with adjustable output impedance designed, adjustable from +ve through 0 to -ve impedance.
Resistance isn't futile it's V / I.

User avatar
Nick
Site Admin
Posts: 12152
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

Re: Remember These?

Post by Nick » Fri Nov 22, 2019 10:30 pm

is not phase linear across the working bandwidth
And of course feedback networks can be designed that are themself not phase linear to counter this to some extent.
Resistance isn't futile it's V / I.

Cressy Snr
Social Sevices have been notified
Posts: 8693
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

Re: Remember These?

Post by Cressy Snr » Fri Nov 22, 2019 10:35 pm

Perfection is impossible, so all we can do is give it our best shot.
I mean OTLs really have to be the some of the stupidest amplifiers ever created, and are without doubt a complete and utter bastard for an amateur to get right. I’ve been at this thing a year now, and I’m just about starting to see its true worth as a music machine.

It sounded like nothing I’d ever heard when I fired it up as a breadboard and once I’d got over the fact that it didn’t blow up or catch fire, and I’d dealt with the hum, I was smitten. I stepped backwards somewhat, when I starved the original 5687 input/drivers of HT and Nick’s analyser results showed the badness of that decision. Since then, little by little as I’ve got used to what works and what doesn’t, I’ve been able to refine it and at each step in the process, the sound quality has got better.

I’ve done a lot more work on the OTL, than I’ve documented in this thread, and a year is the longest I’ve kept my attention on a project. For example I’ve quietly found out that carbon film resistors sound better than any other type I’ve put in the feedback line. I never get bored with working on it. There’s always something new to find out; a tiny nuance here, a ‘did I really just hear that’ there. Seeing the twin cathodes of each of the 13E1s coming up to temperature before switching their HT on is always a treat.
On every tree there sits a bird,
Singing a song ...of love...

Post Reply