Page 52 of 54

#766 Re: Remember These?

Posted: Wed Nov 20, 2019 1:46 pm
by RhythMick
coupling caps are 470nF ? Am I reading that right ? Should it be uF ?

#767 Re: Remember These?

Posted: Wed Nov 20, 2019 1:49 pm
by RhythMick
Thanks for sharing the circuits. I would love to try an OTL in a year or so, watching this with interest.

What speakers do you intend to drive with this?

#768 Re: Remember These?

Posted: Wed Nov 20, 2019 1:56 pm
by Nick
RhythMick wrote: Wed Nov 20, 2019 1:46 pm coupling caps are 470nF ? Am I reading that right ? Should it be uF ?
470nf into 39k gives 9Hz, about right. Maybe could be lower, but its parts box stuff.

#769 Re: Remember These?

Posted: Wed Nov 20, 2019 3:02 pm
by Cressy Snr
RhythMick wrote: Wed Nov 20, 2019 1:49 pm Thanks for sharing the circuits. I would love to try an OTL in a year or so, watching this with interest.

What speakers do you intend to drive with this?
I’m driving my 8 Ohm Fane single driver speakers. These are 100dB efficient have a flat impedance characteristic and never drop below 7.4 Ohms. They spend most of the frequency range hovering around the 10 Ohm mark.

#770 Re: Remember These?

Posted: Wed Nov 20, 2019 3:12 pm
by RhythMick
Nick wrote: Wed Nov 20, 2019 1:56 pm
RhythMick wrote: Wed Nov 20, 2019 1:46 pm coupling caps are 470nF ? Am I reading that right ? Should it be uF ?
470nf into 39k gives 9Hz, about right. Maybe could be lower, but its parts box stuff.
Again I'm showing where my understanding may be flaky - I just love doing that ! :D

I thought the equation for the -3db frequency corner was 1/(2pi*R*C)... where R is the parallel resistance of anode load and grid leak. Anode in this case being the 150H choke (+ 3400DCR). At 10Hz the inductive reactance of the choke will be about 9K5, plus DCR makes about 12K8 whcih in parallel with 39K is about 9K75. Do I have that right ?

#771 Re: Remember These?

Posted: Wed Nov 20, 2019 3:15 pm
by RhythMick
Cressy Snr wrote: Wed Nov 20, 2019 3:02 pm
RhythMick wrote: Wed Nov 20, 2019 1:49 pm Thanks for sharing the circuits. I would love to try an OTL in a year or so, watching this with interest.

What speakers do you intend to drive with this?
I’m driving my 8 Ohm Fane single driver speakers. These are 100dB efficient have a flat impedance characteristic and never drop below 7.4 Ohms. They spend most of the frequency range hovering around the 10 Ohm mark.
When you finish I'd love a listen.

#772 Re: Remember These?

Posted: Wed Nov 20, 2019 4:12 pm
by Nick
RhythMick wrote: Wed Nov 20, 2019 3:12 pm
Nick wrote: Wed Nov 20, 2019 1:56 pm
RhythMick wrote: Wed Nov 20, 2019 1:46 pm coupling caps are 470nF ? Am I reading that right ? Should it be uF ?
470nf into 39k gives 9Hz, about right. Maybe could be lower, but its parts box stuff.
Again I'm showing where my understanding may be flaky - I just love doing that ! :D

I thought the equation for the -3db frequency corner was 1/(2pi*R*C)... where R is the parallel resistance of anode load and grid leak. Anode in this case being the 150H choke (+ 3400DCR). At 10Hz the inductive reactance of the choke will be about 9K5, plus DCR makes about 12K8 whcih in parallel with 39K is about 9K75. Do I have that right ?

I don't think so, the anode load in parallel with the valve Ra gives the source impedance, but the corner of the RC formed by the coupling cal and grid leak are the bits that matter when calculating the transfer forward. The choke does create the load the valve sees which is in parallel with the grid lead as you describe.

#773 Re: Remember These?

Posted: Wed Nov 20, 2019 4:15 pm
by RhythMick
Nick wrote: Wed Nov 20, 2019 4:12 pm
RhythMick wrote: Wed Nov 20, 2019 3:12 pm
Nick wrote: Wed Nov 20, 2019 1:56 pm

470nf into 39k gives 9Hz, about right. Maybe could be lower, but its parts box stuff.
Again I'm showing where my understanding may be flaky - I just love doing that ! :D

I thought the equation for the -3db frequency corner was 1/(2pi*R*C)... where R is the parallel resistance of anode load and grid leak. Anode in this case being the 150H choke (+ 3400DCR). At 10Hz the inductive reactance of the choke will be about 9K5, plus DCR makes about 12K8 whcih in parallel with 39K is about 9K75. Do I have that right ?

I don't think so, the anode load in parallel with the valve Ra gives the source impedance, but the corner of the RC formed by the coupling cal and grid leak are the bits that matter when calculating the transfer forward. The choke does create the load the valve sees which is in parallel with the grid lead as you describe.
OK understood Nick thanks for correcting... FYI this is where I got the previous understanding from (among other sources)...

https://www.diyaudio.com/forums/tubes-v ... alues.html

#774 Re: Remember These?

Posted: Wed Nov 20, 2019 5:15 pm
by RhythMick
Capture.PNG
Page 483 of the RDH v4 confirms... not that I've yet grokked the other formulae on that page

#775 Re: Remember These?

Posted: Wed Nov 20, 2019 10:38 pm
by Mike H
Cressy Snr wrote: Wed Nov 20, 2019 1:33 pm I’ve also after the fact, seen this on DIYAudio...“NEVER put an interstage transformer or plate choke inside a global feedback loop” but there don’t seem to be any problems with my plate choke setup, probably because there is no output transformer sticking its oar in. Apart from the potential to tweak the driver bias, the audio circuit performs quite nicely.
LOL, well my germanium output p-p amp does, it has a (modified) interstage acting as phase splitter and driver to the output OC34's, admittedly it doesn't like the speaker load going AWOL and tends to throw its toys out its pram and go a bit unstable.

#776 Re: Remember These?

Posted: Thu Nov 21, 2019 4:17 pm
by Cressy Snr
Mike H wrote: Wed Nov 20, 2019 10:38 pm LOL, well my germanium output p-p amp does, it has a (modified) interstage acting as phase splitter and driver to the output OC34's, admittedly it doesn't like the speaker load going AWOL and tends to throw its toys out its pram and go a bit unstable.
Not tried taking the load off it. :)
Increased the feedback to 20dB late last night. I’ve been reading the section on feedback in both Bruce’s ‘Tubes & Circuits’ book, and Morgan Jones, and after dealing with past misunderstandings, I think I might now know enough about feedback to be dangerous: hours of fun with the feedback equation, gain, resistor ratios and gee, aren’t decibels really useful things?

OK, so what has been learned? Well for one thing, it’s no good blathering on about your ears, with an amp requiring feedback; not until you’ve at least set a level that gets you almost there. And to be able to do that, you need to be competent with the maths and how to apply it. Being skilled in creative writing and poetry plays no part whatsoever in the process. Bugger!

And using feedback is not always a criminal offence. :wink:

#777 Re: Remember These?

Posted: Thu Nov 21, 2019 8:39 pm
by Cressy Snr
Below is the sepia toned audio circuit schematic again, but with the revised feedback level. Feedback resistor is ringed on the diagram:
04478A14-ABC7-4DC3-BD80-E453C7316FA5.jpeg
Now I have the feedback levels worked out with maths and stuff, and set up right, what I can say is that it is interesting the level of distortion you will quite unwittingly accept if the distortion spectrum stays the same with output level. You do, however, notice when the distortion is not there.

#778 Re: Remember These?

Posted: Thu Nov 21, 2019 10:56 pm
by Wolfgang
Now I have the feedback levels worked out with maths and stuff, and set up right, what I can say is that it is interesting the level of distortion you will quite unwittingly accept if the distortion spectrum stays the same with output level. You do, however, notice when the distortion is not there.
Interesting stuff from the "Lenard Audio Institute" regarding NFB. According to this NFB hardly reduces distortion but mainly hiss and hum and it negatively influences SQ. Of course one would never find critical comments on NFB in a book that features OTLs which depend by principle on a lot of NFB.

Here is the excerpt:
"External negative feedback is the process of feeding back a small amount of signal from the speaker output to the Cathode of the pre-amp valve (in opposite phase) to reduce the overall gain of an amplifier. External Negative feedback has a limited effect on reducing distortion because the output cannot be exactly 180 deg out of phase with the input at all frequencies. 6dB to 12dB feedback is the maximum that can be achieved before most valve amps are caused to oscillate.
Almost all valve amplifiers have external negative feedback. But, external Negative feedback is not required to enable a valve amp to function. If the external negative feedback is removed, the amp will appear to sound more responsive with approx 6dB to 12dB more gain. The immediate experience is sound quality has improved.
What is the purpose of external negative feedback in valve amps?
The primary purpose of Negative feedback in valve amps is to quieten the amplifier. Valve amps are naturally noisy and in the past (before solid-state) there was nothing to compare them with, therefore a small amount of hum and hiss was considered normal. Instead of stating "Negative feedback reduces the noise of the amp" the emphasis was to state the benefit as "Negative feedback reduces distortion".

Thermionic hiss and mechanical rattling of the valve internal elements is clearly audible. Tapping a valve with a finger reveals this. The valve filaments in the majority of valve amps are fed with 6.3V AC from the mains transformer. Hum leakage from the AC current flowing through the filaments is imposed on the Cathodes. EI power trannys induce eddy currents into the chassis which is heard as a residual background hum. Valve amp circuits do not have common mode rejection, therefore any ripple from the power supply, no matter how small, introduces a small audio hum which also causes amplitude modulation of the audio signal.

Distortion reduction Negative feedback reduces the hum and hiss of a HiFi valve amp, but has little to zero effect on reducing audible distortion. Many so called valve amp experts will have a hostile reaction to this statement. The theory and education behind electronics technology of the past concentrated on telecommunications and not HiFi amps. Valve telecommunication repeater amplifiers are very different, where Negative feedback had a large effect in reducing distortion. This is where much of the confusion about Negative feedback in HiFi valve amps came from.
The most significant effect of Negative feedback is reducing the output Impedance, providing more damping to the speaker. Many high quality valve amps made during the 60s allowed the Negative feedback to be adjusted to tune the damping of the speaker."

#779 Re: Remember These?

Posted: Thu Nov 21, 2019 11:08 pm
by Wolfgang
Distortion reduction Negative feedback reduces the hum and hiss of a HiFi valve amp, but has little to zero effect on reducing audible distortion.
For hiss and hum reduction we don't need big amounts of NFB but can use regulated psus and a second housing for the transformers.But of course this is more expensive than a NFB resistor.

#780 Re: Remember These?

Posted: Fri Nov 22, 2019 9:19 am
by Cressy Snr
I’m off to do some creative writing.