seperate channels sharing cathode follower

[EETStudent]

Im about done with my AC30 build with 6AQ5s(cause im a rebel) and it sounds really sweet. Unlike anything Ive owned. Anyway, ever since my last marshall 2203 build(modded for high gain. you know, the typical 5uf cap across V3 cathode resistor) Ive still been craving some searing solo distortion, something this ac30 isnt quite capable of. Ive got 4 free stages. I was wondering if I could link these stages with the cathode follower in the vox circuit. basically take the plate of V4 in whatever circuit i go with, and run in straight into the grid of the follower, and set the whole thing up as seperate channels. Part of me feels like with the two circuits joined at the grid, the unused circuit will still generate noise/have some negative on the circuit in use. Tell me im wrong...

[Chuck H]

I wish I could tell you that your wrong. But IME I it will hiss. I've set up amps like this before (so has Mesa incidentally). The high gain channel will audibly hiss in the other channel. Thats why, if you look at any schems for "channel switching" amps, they usually have a switched ground placed somewhere near the mix point for the channel that will not be in use. But if you can tolerate the hiss, it's a neat way to get channel switching without the use of LDRs or relays. Plus you can use an ABY pedal or a split adaptor to use both preamps simultaneously (as long as they're in phase).

Chuck

[mooreamps]

I see no reason why a cathode follwer circuit not be able to drive two separate loads, or in engineering terms, have a "fan-out" of two.

-g

[Russ]

What if the ABY pedal grounds the input of the unused channel - does the remaining channel still contribute hiss?

[Chuck H]

Yes, some. Mesa found this to be a problem when customers complained about too much hiss on the clean channel of one of their "channel switching" amps. Since there was no ground on the "lead" channel nearer the end of it's cascade chain (and the mixing resistors) the residual hiss was VERY audible on the clean channel. A high gain channel with no signal on it is just residual noise with a buttload of headroom. If you want a quiet "clean" channel in a "channel switching" amp, you really do have to place a ground somewhere near the end of the unused channels signal chain. Often at the beginning as well to avoid signal "bleed" via the voltage rail or shared grounds.

Chuck

[Chuck H]

Gary,

FWIW I follow your posts. Your insight is very experience based and practical. In my book that kicks engineering propers a$$ every time. It's the same reason I like to read Bruce and Enzo's posts.

But I think you misinterpeted the Q. It reads to me like he wants to sum the two channels at a cathode follower. I answered re the noise part of the Q, but to address summing the two channels...My vote is for a virtual earth mixer. Very low interaction between the channels due to the low input impedance and similarly low output impedance like a cathode follower.

Chuck

[Russ]

I've never heard of a "virtual earth" mixer, what is it?

[mooreamps]

My vote is for a virtual earth mixer. Very low interaction between the channels due to the low input impedance and similarly low output impedance like a cathode follower.

Chuck

I did misunderstand what his question was, and thank you for the kind words.
I would certainly agree running two signals into a low impedance input of a mixer would offer better noise immunity. Other points I would agree are valid too, the use of proper grounding, and using shielded cable for going to high impedance inputs.

-g

[Chuck H]

Russ,

A virtual earth mixer is a summing amplifier with unity gain. It is a basic shunt feedback circuit with multiple signals at the input each running through the same size resistance. For example:

12AX7 triode-100k plate load-.047 decoupling cap-1.5k cathode w/22uf bypass cap-1M grid to ground resistor

After the decoupling cap shunt a 100k resistor to that triodes input grid. Then feed that grid through a 100k resistor. input as many signals at that grid as you like, as long as each one goes through a 100k resistor. There will be no noticable interaction between the signals.

100k is just a generic R value for the shunt and input resistance but it works just fine for most applications.

You can also adjust the input resistances to alter the balance between the inputs. But at the compromise of some channel interaction.

After the decoupling cap and the shunt resistor you can series a second decoupling cap to roll off some bottom end if you like. But the shunt resistor should follow a largish first decoupling cap so that you actually "shunt" most of the signal. Similar logic on the large cathode bypass cap for this amp.

This circuit can be very useful in guitar amps for things other than just summing. I design with it alot and never understood why I don't see it often. I think one of the AA??? Bassmans used a similar circuit.

I started designing with them about 5 years ago on a tip from Ken Gilbert here on Ampage. I wanted a very low interaction summing amp with low output impedance to use as a reverb mixer. It worked great.

Enjoy

Chuck