An unusual VAS CDOM variation ideas from a web page

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello

I did found an unusual ideas for a VAS CDOM variation , at this web page;

http://www.tnt-audio.com/feedback/may03.html

I include the text, any opinions about that ideas ?

Yet another opinion
IN MY OPINION.............
Many inexpensive solid state amplifiers use a large amount of global negative feedback, and they usually have a second feedback loop which rolls off the open loop gain at higher frequencies.
The roll off in open loop gain from this second feedback circuit is necessary to keep the amplifier stable, and it is nearly always accomplished with one or two simple feedback capacitors (Cdom) located in the voltage amplification stage (VAS). The VAS section of the circuit drives the output stage transistors. Often, the output stage transistors constitute a voltage-sensitive capacitive shunt from Cdom to the power rails of the amplifier.
Since the power rails are bypassed to ground, the feedback which the Cdom capacitors provide is modulated by the variable base-to-collector capacitance of the output transistors. In some amplifiers the VAS stage is buffered, so that Cdom is shunted by a only relatively small transistor, and therefore a relatively benign capacitance. But in other amplifiers, Cdom is shunted by the parallel capacitances of several large power transistors, and the modulation effect is more pronounced.
I have begun to test my hypothesis that part of the "transistor" sound of many solid state amplifiers is due to a non-linear capacitive shunt which affects the operation of Cdom, rather than being simply an effect of global feedback. My current test method is relatively crude. I simply remove Cdom, and instead use a small inductor in the emitter circuit of the VAS stage.
The inductor is a small shielded unit usually costing USD $2.00 or less, and it provides frequency-dependent degenerative feedback at the VAS stage. Significantly, the roll-off slope provided by the inductor is isolated, and hence not dependent on the variable capacitance of the output transistors. Based on my results to date with a few samples, this change could be a worthwhile upgrade for several popular amplifiers. So far the improvements have been noticeable but not profound. What do you expect for $4.00 in components?
Probably by now a little voice in your head is screaming something like, never use an inductor because they are large, expensive, generate interfering magnetic fields, and have high amounts of distortion. Well, I do agree that the substitution of a $2.00 inductor for a $0.20 capacitor in theory causes a rise in harmonic distortion. But actually, using an emitter inductor instead of Cdom in the VAS places less of a load on the input stage because the VAS has a higher input impedance. So the distortion from the input stage may be reduced. It depends. And it is better if the VAS stage works into a constant current, so that the current swings in the inductor are not so large.
But I hypothesize that eliminating Cdom with the emitter inductor often eliminates a significant source of phase modulation. That is because I think that normally the transit time through the VAS is modulated by the varying capacitance at the base or gate of the output stage. As for the inductor, you don't have to use a shielded model for a quick trial if you can place it away from strong magnetic fields. Choosing the correct value can be tricky, but you can start by installing a 1 mH emitter inductor, and just reducing the value of Cdom by twenty percent.
Those with SPICE or other analytic skills may wish to try a more elegant approach, but they must keep in mind that the result can only be as good as their models. And in some cases, the self-resonant frequency of an inexpensive inductor could be too low to provide stability. It is wise to remember that if an amplifier oscillates under any condition, expensive damage can easily occur. So this approach requires an experienced hand, a bit of luck, and an oscilloscope. Maybe someday a specialist company will offer a line of audiophile grade emitter inductors with silver windings and mumetal shields, but so far the $2.00 variety have seemed adequate.
.A related factor to keep in mind is that the percentage variation in base-to-collector capacitance of the output transistors is a smaller percentage when the output voltage swing is smaller. Also, the absolute value of the capacitance is smaller when the supply voltages are higher. So I still think the emitter inductor idea may have possibility for several moderately-priced amplifiers.
All of the above are my own UNPROVEN IDEAS, and by the above disclosure I imply absolutely no responsibility what so ever, to anyone whom ever, for anything what ever.
Cheers
Charlie Miller
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.