Gentlemen,
The last post of my 6384 SE amp thread includes comparative measurements of transformer primary distortion vs secondary distortion on an amp with ~93% parallel applied voltage feedback from plate to grid of output tube.
All were at 1kHz, so wouldn't shed light on low frequency limits, but they do illustrate how well distortion tracks from primary to secondary at a few power levels on the common HiFi Edcor SE transformers at mid band. (Short answer: not so well at higher levels)
https://www.diyaudio.com/forums/tubes-valves/354046-shunt-feedback-6384-se-amp-2.html#post6240091
The last post of my 6384 SE amp thread includes comparative measurements of transformer primary distortion vs secondary distortion on an amp with ~93% parallel applied voltage feedback from plate to grid of output tube.
All were at 1kHz, so wouldn't shed light on low frequency limits, but they do illustrate how well distortion tracks from primary to secondary at a few power levels on the common HiFi Edcor SE transformers at mid band. (Short answer: not so well at higher levels)
https://www.diyaudio.com/forums/tubes-valves/354046-shunt-feedback-6384-se-amp-2.html#post6240091
The O/P trafo in the Schade configuration is outside the NFB loop. Therefore, the NFB can't correct any aberrations in the performance of the "iron". However, that's not the whole story. Short loop NFB does reduce the impedance driving the O/P trafo and that definitely has a positive impact on the performance of the "iron".
Eli,
As the load on the output tube plate changes, the amount of error that the Schade feedback sees also changes . . . Right?
The Schade feedback does not care whether the error is from tube non-linearity, or whether the error is from the varying load line versus frequency that the plate sees.
Schade feedback can not discriminate the cause(s) of the error . . . Right?
What changes the plate load?
The primary inductance of the output transformer
The load on the secondary that is reflected to the primary
The distributed capacitance of the primary
The leakage reactance between the primary and the secondary, and how that affects
the amount of secondary load that gets reflected back to the primary (versus frequency)
The change in inductance when the laminations saturate
I am wondering if perhaps we are just talking about semantics here.
As the load on the output tube plate changes, the amount of error that the Schade feedback sees also changes . . . Right?
The Schade feedback does not care whether the error is from tube non-linearity, or whether the error is from the varying load line versus frequency that the plate sees.
Schade feedback can not discriminate the cause(s) of the error . . . Right?
What changes the plate load?
The primary inductance of the output transformer
The load on the secondary that is reflected to the primary
The distributed capacitance of the primary
The leakage reactance between the primary and the secondary, and how that affects
the amount of secondary load that gets reflected back to the primary (versus frequency)
The change in inductance when the laminations saturate
I am wondering if perhaps we are just talking about semantics here.
It seems like Eli is simply stating that the primary and secondary waveforms don't necessarily match (which my midband measurements indicate is true, and I would expect frequency extremes to be even worse) and that a feedback loop from the secondary will attempt to correct for those additional errors and will drive the primary even harder and result in a situation where the global feedback can't correct the problems and just sounds terrible.
The shorter loop might have sounded much better, but only because it was completely oblivious to some errors in the output transformer that it would have stepped all over system-level performance trying to correct.
The shorter loop might have sounded much better, but only because it was completely oblivious to some errors in the output transformer that it would have stepped all over system-level performance trying to correct.
My point was that at the beginning point of saturation, it can be caused to be worse, even if the only negative feedback is Schade feedback.
It is not as bad as Global negative feedback, but it still has at least a mild effect to make saturation worse.
The lesson is:
Without enough laminations, (or considerable pre-filtering of low frequencies) you may have core saturation.
True for single ended, true for push pull, true with negative feedback, true without negative feedback.
To some extent, you get what you pay for.
It is not as bad as Global negative feedback, but it still has at least a mild effect to make saturation worse.
The lesson is:
Without enough laminations, (or considerable pre-filtering of low frequencies) you may have core saturation.
True for single ended, true for push pull, true with negative feedback, true without negative feedback.
To some extent, you get what you pay for.
Output transformer has 4.48 Hy of primary inductance (static measurement).
You realize that you can not expect hifi sound, no matter the tube circuit?
You realize that you can not expect hifi sound, no matter the tube circuit?
dannve,
4.48 Hy is 2,814 Ohms of inductive reactance at 100Hz.
That is roughly equal to the 2.5k connection of the 125DSE.
That gives a bandwidth of about 100Hz.
The acoustic guitar low note is about 100Hz (a little higher than an electric guitar's 80Hz).
One man's poor sounding amplifier, is OK for another man (acoustic guitar, flute, etc.).
Hi Fi is one thing, application is another.
"All Generalizations Have Exceptions" - Me
4.48 Hy is 2,814 Ohms of inductive reactance at 100Hz.
That is roughly equal to the 2.5k connection of the 125DSE.
That gives a bandwidth of about 100Hz.
The acoustic guitar low note is about 100Hz (a little higher than an electric guitar's 80Hz).
One man's poor sounding amplifier, is OK for another man (acoustic guitar, flute, etc.).
Hi Fi is one thing, application is another.
"All Generalizations Have Exceptions" - Me
6A3bummer (yes difficult to spell names right...),
Then I am curious what application the TO has in mind as there is another forum for guitar amps.
Then I am curious what application the TO has in mind as there is another forum for guitar amps.
The Norwegian OP asked for circuits of the reproduction kind. The 125DSE "iron" had already been acquired. Collectively, we will do our best to turn the lemons into lemonade.
FWIW, Hammond's 125GSE has the core mass needed for a 5 W. setup employing GNFB to attempt HIFI reproduction. Unfortunately, the entire Hammond 125xSE line seems to be a poor value. It's a shame that Edcor's stuff is not easily accessed in Europe.
my take, do not ask your irons to perform beyond its capabilities for which they were designed for...
as long as you understand the limitations in each case, you should be fine...
output irons are only as good up to 400 hz and beyond that, Bud Purvine calls it an antenna event..
Patrick Turner in his blogs dealt with core saturation and has a nice explanation....
opt's in an amplifier tested with 1khz sine and 10 watts, will probably saturate at 5 watts and 30 hz...
as long as you understand the limitations in each case, you should be fine...
output irons are only as good up to 400 hz and beyond that, Bud Purvine calls it an antenna event..
Patrick Turner in his blogs dealt with core saturation and has a nice explanation....
opt's in an amplifier tested with 1khz sine and 10 watts, will probably saturate at 5 watts and 30 hz...
Last edited:
Ok, it just seemed like you were disagreeing with Eli who also seemed to be saying this exact thing. Carry on.
Let's not confuse apples and oranges.
At HIGH power the power bandwidth will fall at 100Hz. NFB may make the transition more abrupt but does not change the general trend. The tube can't make the extra current.
At LOW power some NFB can flatten the curve to far below the L/R product. I have seen 20Hz out of $1 transformers which did not aspire much below 150Hz (cheap table radio), when measured at a tiny fraction of nominal Power (like 0.1W on a 1.1W core).
PRR,
Absolutely, you are correct.
there is Bandwidth, and there is Power Bandwidth.
I remember the old Lafayette 60 Watt per channel vacuum tube stereo power amp.
They had a spec of bandwidth at 1 watt, something like single digit-Hz to 100kHz, +0 - 1dB.
Of course at 60Watts the power bandwidth was much much less.
Absolutely, you are correct.
there is Bandwidth, and there is Power Bandwidth.
I remember the old Lafayette 60 Watt per channel vacuum tube stereo power amp.
They had a spec of bandwidth at 1 watt, something like single digit-Hz to 100kHz, +0 - 1dB.
Of course at 60Watts the power bandwidth was much much less.
daanve,
You said:
"Then I am curious what application the TO has in mind as there is another forum for guitar amps."
I agree, in post # 2, the original poster mentioned his Klipsch loudspeakers.
The fact that he is using those speakers is our only indication that this is for HiFi.
But as you may have observed (and I certainly have), there are a lot of questions about guitar amplifiers that are posted here on the Hi Fi part of the forum.
And I believe that instead they belong on the instruments and amps part of the forum.
Sometimes the original poster has us giving several posts of questions and answers before revealing that his question is for an electric guitar amp, even though it is on the tube amps part of the forum.
That should not be done that way.
What is wrong with keeping all guitar amps on the instruments and amps part of the forum?
You said:
"Then I am curious what application the TO has in mind as there is another forum for guitar amps."
I agree, in post # 2, the original poster mentioned his Klipsch loudspeakers.
The fact that he is using those speakers is our only indication that this is for HiFi.
But as you may have observed (and I certainly have), there are a lot of questions about guitar amplifiers that are posted here on the Hi Fi part of the forum.
And I believe that instead they belong on the instruments and amps part of the forum.
Sometimes the original poster has us giving several posts of questions and answers before revealing that his question is for an electric guitar amp, even though it is on the tube amps part of the forum.
That should not be done that way.
What is wrong with keeping all guitar amps on the instruments and amps part of the forum?
There seem to be foul information in Hammond 125 DSE datas: https://www.hammfg.com/files/parts/pdf/125DSE.pdf
Given secondary inductances produce about 50% smaller impedance transforming ratio as should be.
Example: Full secondary should produce 10k to 32 ohms, which is 312.5 as impedance transforming ratio.
However, inductance ratio (pri. to sec.) is 4.48 H / 27.4 mH = 163.5.
This gives 5.23 k to 32 ohms, not 10k to 32 ohms as specified.
Given secondary inductances produce about 50% smaller impedance transforming ratio as should be.
Example: Full secondary should produce 10k to 32 ohms, which is 312.5 as impedance transforming ratio.
However, inductance ratio (pri. to sec.) is 4.48 H / 27.4 mH = 163.5.
This gives 5.23 k to 32 ohms, not 10k to 32 ohms as specified.
let us take 2.5k to 4 ohms, that is an impedance ratio of 625, a turns ratio of 25....
same turns ratio as 10k:16ohms.....
with 32 ohms, turns ratio is 25/2...
same turns ratio as 10k:16ohms.....
with 32 ohms, turns ratio is 25/2...
Last edited: