Simple 300B Test?

[Paul Birkeland]

PJ will certainly weight in on this as well.  The impedance of an electrolytic capacitor is not well enough behaved to function well in the high frequency compensation network for the regulator. 

I haven't done any analysis on how critical this is, but since PJ has not specified a ceramic cap for that position, the situation is unlikely to be all that dire.

[Gerry E.]

@Gerry - I noticed from your photo that you've installed a V-Cap on what looks like a shunt regulator board, presumably in place of a stock Dayton Audio or similar. Did you notice any sonic change as a result? 

many thanks, Derek

Hi Derek:

I purchased my Paramounts from the son of the original owner/builder.  Sadly, his father had passed away.  Both amps already had the large, upgraded Mundorf (coupling?) cap installed.  Also, one of the amps already had the two smaller (.01uF?) V-Caps installed and the other amp had one.  It appeared that the father had installed that second V-Cap in one of the amps but passed away before he got around to doing the other. 

I didn't know the technical reason for upgrading those two smaller caps but in order to make the amps identical, I purchased another V-Cap and had it installed at the same time I had the v1.1 soft-start upgrade installed.  Now that you know the background story, the answer to your question is that I don't know, I always heard the amps this way.

Gerry                 

[Paul Joppa]

The cap is part of an RC network that maintains the low impedance of the shunt reg at high frequencies; without that network the impedance rises and becomes inductive. The values depend mostly on the voltage; at around 300 volts the corner frequency is around 10kHz so the cap is important above that frequency.

Therefor, in theory it should be well-behaved (i.e. with low ESR and ESL) at high frequencies. Initially, I did not expect it to be audible because it's dominated by the resistor except above 10kHz, and because the regulator itself is isolated from the tube by a very high impedance current source plate load. I included it because it should improve the stability margin of the circuit.

Of course, theory is limited in its ability to predict subtle effects that millions of years of evolution have made quite audible. Initially I doubted that it would be audible, but we tested this by substituting a Russian mil-spec teflon capacitor and were startled to find that it did make a small but audible improvement. So I spec a polypropylene capacitor, same as we use for coupling caps.

[Deke609]

Many thanks all.

@PJ: your comments are doubly helpful. Your observation that teflon caps produced a small but audible changes gives me something to play with without having to learn any new math  - I have a couple of 0.1 uF russian teflon caps kicking around.  Whereas your points about the functioning of the rc network will require some reading and new math. I quickly looked up rc series networks and was met with discussion of phasors and complex numbers - I haven't yet ventured into this area in my learning.  But leaving out phase, I think I mostly get how the inductance and capacitance of a component varies with frequency - so that mostly makes sense. But I've yet to make sense of how the shunt reg circuit works - I need to read up on what a LM/LT431 does and how it does it.  More late reading!  Many thanks.

cheers, Derek