I guess it depends on how one defines "poor." From the perspective of a bean counter, I suppose one could say that it is rather "poor." But from a more practical and meaningful perspective, i.e. actually listening to the result, the experience belies the "numbers."
Well, grab yourself a pair of 10C's and find out for yourself.
(Perhaps somone should invent a transformer with built in negative feedback)
Or you could drive it from a negative impedance equal to the primary's winding resistance. That'll keep the bean counter's happy.
se
Generally what we're more tolerant of is low order distortion versus high order distortion. With transformers, their distortion is pretty much just third harmonic.
Also, transformers exhibit nowhere near the amount of IM distortion that one would typically get from an active circuit with similar levels of THD.
Their distortion characteristics are rather unique and more benign than the raw numbers would lead you to believe.
se
Yes, I agree with your conclusion about lower order distortion.
I have read at several places, although I have no references at the moment, that the human ear is not able to distinguish distortion in the nether regions easily. Conventional cone speakers start distorting much earlier in the bass frequencies before distorting higher up. A harsh treble is much more annoying that a "harsh" bass, and I haven't ever heard a harsh bass.
I have read at several places, although I have no references at the moment, that the human ear is not able to distinguish distortion in the nether regions easily. Conventional cone speakers start distorting much earlier in the bass frequencies before distorting higher up. A harsh treble is much more annoying that a "harsh" bass, and I haven't ever heard a harsh bass.
Gripracer,
You asked how much gain is enough. The voltage amplifier should remain linear, while being capable of driving the "finals" into clipping. The FET PSU starts with 18 VRMS. The I/P is 2 VRMS. A voltage gain of 10 should prove adequate.
You asked about using a 6Н2П (6n2p) or 6Н6П (6n6p). For all practical purposes, the 6Н2П is equivalent to the 12AX7. That type has an amplification factor (μ) of 100. That's way too big. OTOH, the 6Н6П data sheet indicates that it is quite suitable. A quick glance suggests a 160 VDC plate to cathode differential and -6 V. as the grid bias.
You asked how much gain is enough. The voltage amplifier should remain linear, while being capable of driving the "finals" into clipping. The FET PSU starts with 18 VRMS. The I/P is 2 VRMS. A voltage gain of 10 should prove adequate.
You asked about using a 6Н2П (6n2p) or 6Н6П (6n6p). For all practical purposes, the 6Н2П is equivalent to the 12AX7. That type has an amplification factor (μ) of 100. That's way too big. OTOH, the 6Н6П data sheet indicates that it is quite suitable. A quick glance suggests a 160 VDC plate to cathode differential and -6 V. as the grid bias.
I once made an amplifier using a step up audio transformer as the VAS. It worked OK but lacked a certain "definition". I gave up on the idea because I was searching for something much better. It just seemed too simple. There must be a reason why everyone is not doing it.
But for certain, the quest is entirely possible using only solid state.(no transfo of course
) As for relative simplicity?....to be a realist, unfortunately 
There is requirement for multiple stages and multiple layers of feedback.
That's rather like saying you once made an amplifier using transistors as the VAS and you didn't like it so you gave up on transistors (or tubes as the case may be).
Just as it's possible to create a solid state or tube amp that doesn't sound good, it's just as possible to use a transformer as a VAS that doesn't sound good.
And without knowing exactly what transformer you used and under what conditions it was used, it's impossible to say why it may not have sounded so good. For all I know, you might have used a $6 Edcor transformer or something.
There must be a reason why everyone is not doing it.
Well, given that some of the finest recordings ever made involved using transformers for voltage gain, I don't suspect it has anything inherently to do with transformers in and of themselves.
se
So 20kHz don't matter then?
In this case, those transformer distortion plots really don't mean a thing. With the highly variable and non-linear input impedance (both with ampitude and frequency) of that MOSFET power output stage, using transformer as a VAS, driven with anything but an ultra low impedance source, is a recipe for a monumental distortion generator.
In this case, those transformer distortion plots really don't mean a thing. With the highly variable and non-linear input impedance (both with ampitude and frequency) of that MOSFET power output stage, using transformer as a VAS, driven with anything but an ultra low impedance source, is a recipe for a monumental distortion generator.
Point taken.
I thought I'd mentioned it here, but it turns out it was in a reply to a visitor message, that the circuit you hitch it to should be able to perform well when fed from the transformer's relatively high output impedance.
So if the MOSFETs in his power buffer will perform that bad as a consequence, perhaps put a B1 in between the two.
se
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.