Multiple chokes in power supply?

[JayCobie]

I've been trying to read up on tube amp power supplies, I'm still very much a novice in these matters, but I have been trying to find the answer to the following thought I had:

Every filtering section I've seen in a tube amp power supply so far only has one choke to smooth ripple voltage. The rest of the filter stages in said power supplies always have a resistor-capacitor filter.
My question is: In a "size-and-cost-are-no-issue" amplifier, what would happen if the filter section would only have appropriately sized chokes instead of resistors in the filter stages?

Of course you would design the rest of the amp to reflect whatever higher or lower HT voltage would be present as a result. I'm not well versed enough to realize what negatives (besides size, practicality and cost) there would be.

[jjasilli]

A choke is an inductor.  An inductor is a coil of wire which has static, or DC resistance.  It also impedes the flow of AC on a frequency dependent basis; it does this by having the affected AC frequencies dissipate their energy by causing them to generate a magnetic field.  One unwanted is result is that those magnetic fields may induce noise in the amp's circuitry. 


Anyway, full wave rectification of the B+ supply leaves some ripple (AC) @ 120 Hz which causes a noticeable hum.  This is reduced by the B+ filter circuitry.  Each node reduces ripple by a small amount, so a number of filter nodes are used.  This is done with a series of resistors & caps -- an R-C filter -- or inductor(s) & caps -- an L-C filter.


Operationally, resistors can be substituted for inductors, and vice-versa, anywhere in an amp.  Pragmatically, there is no reason to use an inductor if you have achieved the desired result with a resistor. 


But inductors do have features that resistors don't.  Inductors filter AC (such as ripple) while presenting very little DC resistance; hence little DC voltage drop.  However, excessive B+ filtering can make a tube amp sound sterile.


Also, inductors have hysteresis -- the property of persistence a given state.  Until AC flows through an inductor, there is no magnetic field.  When AC does flow, the non-existence of the magnetic field persists for a while.  I.e., it takes time for the magnetic field to arise.  And vice-versa.  When AC stops, the magnetic field continues to persist for a while.  This gives the amp a different feel.

[Williamblake]

You could also add a common mode choke one half on the B+ Input and one half on the ground bus ... Merlin talks about that i think and i had that choke lying around so i tested it and it didn`t hum.

[PRR]

> In a "size-and-cost-are-no-issue" amplifier

That only happens in some world where money falls from the sky. Or the local steel and copper industry is subsidized to where chokes are cheaper than resistors.

Another issue is that an all L-C filter tends to ring. You need resistors. Often the circuit stages damp the ring enough, but you have to check it. Otherwise it motorboats.

A 30mA 3H screen choke is a thousand turns of moderately small wire. As you get to the 2mA preamp stages you need much higher choke impedance which leads to tens of thousands of turns of wire smaller than unicorn belly hair. Winding costs go up (you say you don't care) and reliability goes down because the ultra-fine wire corrodes-through much faster.

But yeah, it has been done, and you can do it too.

[JayCobie]

I appreciate the answers!

I wasn't planning on making an amplifier with multiple chokes, I already knew that it would be far beyond practical execution (hence the caveat). I was simply entertaining a thought experiment since I couldn't find anyone talking about what negatives would be involved in such an undertaking.