Two tone stacks parallel after cathode follower

[dobrota]

I guess it's possible to connect parallel e.g. two classic Marshall tone stacks after cathode follower and use relay switching only on outputs of stacks (tone stacks for crunch and lead channel) and it will work, but:
- isn't it too high load for cathode follower ?
- won't affect changes first stack tone controls frequency response / sound on output second tone stack ?
- doesn't two stacks parallel change operational point of follower and so signal harmonics / shape of overdrive signal instead of usually only one tone stack ?

[DRH1958]

You could use a double pole relay to switch both the input and outputs to isolate the front and back end of the stacks from each other. This way there will be no interaction of the two. I hope I understood your intention when suggesting this. Also, the cathode follower will drive a wide variety of stacks, so I wouldn't worry too much about the difference in them unless they have a very different input impedance.

[d95err]

Checkout the Marshall JVM410. Series resistor between the treble viper and volume for each channel, then channel mute and a series mix resistor.

No need to switch on the input side of the tone stack.

EDIT: Schematics here: http://forum.metroamp.com/viewtopic.php?f=25&t=21932

[dobrota]

Of course, double pole relay is standard circuit to disable any interaction between two tone stacks, but I would like to know if possible to do it only with one pole relay, two stacks parallel, without some major difference in sound and circuit operation.

[dobrota]

Thanks, I have this schematic, it's clear.

[J M Fahey]

Short answer: the cathode follower *easily* drives 2 regular Marshall tone stacks, just switch between stack outputs or if you wish, between following volume controls if you have them there.

[bea]

But there may be still significant interaction between the two stacks. It would be wise (and not too difficult) to check and possibly optimize Your intended setup with LTSpice.

[Mike Sulzer]

But there may be still significant interaction between the two stacks. It would be wise (and not too difficult) to check and possibly optimize Your intended setup with LTSpice.

I have to second this. Cathode followers have low impedance when they operate linearly, but halving the impedance of the load decreases the linear voltage swing by a factor of two. (You only have so much current available)

If a stage of gain operating into a cathode follower is operating linearly, then the low impedance of the CF (order of 1/gm) keeps interaction low when driving two stacks. But if you drive two stacks instead of one when one is intended, it would reach its linear limit sooner, and then you have interaction since you lose the low impedance.

Also, if you over drive the stage of gain driving the cathode follower, then the follower does not necessarily operate linearly, and then it does not have a low impedance output and you lose the isolation between the stacks.

I think it is better to switch at the input unless you can be sure neither of these things happens.

[bea]

And each tone stack sees the impedances of the other one. The system is hardly overseeable.

An alternative to a switch in front of the stacks might be one or two JFET source followers after the cathode follower. With a common input resistance somehow mimicking the load of one single tone stack. Then You should be able to safely switch with a single switch.

[nickb]

Another strategy is to swap the CF cathode resistor for a transistor or MOSFET based constant current source of around 2 to 3mA. Very inexpensive and drives capacitive loads very well.

[Mike Sulzer]

And each tone stack sees the impedances of the other one. The system is hardly overseeable.

An alternative to a switch in front of the stacks might be one or two JFET source followers after the cathode follower. With a common input resistance somehow mimicking the load of one single tone stack. Then You should be able to safely switch with a single switch.

The two tone stacks do not see each other as long as the impedance of the cathode flower is low. You do not want to use solid state with such a large voltage swing

[Mike Sulzer]

Another strategy is to swap the CF cathode resistor for a transistor or MOSFET based constant current source of around 2 to 3mA. Very inexpensive and drives capacitive loads very well.

The current source is a high impedance; it cannot help here.

[J M Fahey]

Agree with last two and not much with earlier ones, but learnt not to argue against Deeply Established Beliefs

[Mike Sulzer]

Agree with last two and not much with earlier ones, but learnt not to argue against Deeply Established Beliefs

OK, but I would argue for the difference between a low dynamic impedance and the ability to drive a low impedance load to near the power supply rail. The extreme example is an op amp follower, which might have a few micro ohms output impedance at 100 Hz, but lack the ability to drive a 500 ohm resistor to near the power supply rails. This is not a DEB, but rather a DET.

[loudthud]

If you watch the output of the common 12AX7 cathode follower when the signal is a big square wave you can see where positive to negative transitions are slewing as fast as the 100K can pull down. The tube cuts off until the 100K can pull the tone stack low enough to catch up with the cathode follower's input. This changes as you adjust Bass and Mid pots so there will be some interaction between two tone stacks if signals are large. One way around this is to use a cathode follower with at stronger current source. Perhaps a 12AU7 with 22K load.