Reducing SE amp noise by injecting noise

While digging through tubecad.com I came accross an article about reducing power supply noise in SE amps by interjecting a small amount of power power supply noise thru a capacitive divider into the cathode so the plate can track the noise.

http://www.tubecad.com/april99/page2.html

It all seems simple enough, and seems like a good idea, but one question I just can not answer. If you can reduce noise of the output stage so easily with just the addition of one cap, why have I never seen this used before. Is there some sort of downside not addressed here that I am unable to see. It seems like a useful trick, but it seems like there must be some sort of downside. Also the additional cap is in series with the power supply cap, so it decresses the value of this cap, and decresses filtering. Which is not an issue in the power stage since we are going to cancel it out anyways, but not in the drive stages. That may or may not be an issue, and can always be remedied through one more stage of filtering on the power supply, but that seems to negate the advantage of doing this in the first place.

adam

The technique and related ones work, but the drawbacks include effects of the initial tolerances of the parts and their drift. Electrolytics in particular have poor tolerance to begin with, and tube parameters are also pretty loosely distributed. So you will end up tweaking the system for best results. Then the nulling will shift away with time and temperature.

It's a bit more practical to use similar techniques with sand state, and I've picked up 15 or 20 dB better power supply rejection sometimes in circuits that are reasonably reproducible. I tend to work it down to d.c., particularly for current sources where I want some nearly free additional line regulation. But with silicon you can usually get there other ways with a few more parts, for circuits that are more consistent.

At high frequencies things will tend to fall apart, but there you ought to be able to bypass out the noise with capacitors directly on the power rail.

BTW the cap in series is not reducing the effective power supply capacitance (the main rail bypass itself is not shown), but is adding a bit, namely about the series value of it and the cathode cap---for the example given, about 7.5uF. I wouldn't go as far as to use that as the only PS capacitance though, other than maybe when tweaking the injection cap value to begin with.

http://www.groupdiy.com/index.php?topic=9318&start=0&postdays=0&postorder=asc&highlight=

> interjecting a small amount of power power supply noise

About half the mass-produced ten-buck 5-tube radios took the power supply through a few-turn tap on the output transformer to cancel some of the buzz left from too-cheap power caps.

If you have to replace the transformer with a generic, you have to up-size the filter cap too.

Before that: field-coil speakers tend to cancel some amp buzz, when both field coil and plate are fed the same current.

Broskie has taken the idea much further, and the Aikido Amp has become a Big Fad.

Is an interesting idea, but to my eye it often ends up twice as complicated as a simple amp with a monster rail cap.

I think I will have to play around with this one, if nothing else it seems worth the time to figure out and stow away in memory. Thanks for the replies.

adam