Help with Virtual earth mixer feeding a Cathodyne PI

[walkman]

Hi ,

I'd like to setup up a virtual earth mixer to feed a cathodyne PI . How can i ensure that i don't have too much gain so the blocking distortion is caused in the PI. The normal input is from a typical 12ax7 input with simple tone control and another would be for a line in or additionally a reverb return?

Currently i have the virtual earth circuit set up as shown in Merlins book.

Chuck H you might be able to help? I've already read the following posts

http://music-electronics-forum.com/t11779/ ,
http://music-electronics-forum.com/t10422/

and see that your a n advocate of this approach

Cheers

Walkman

[J M Fahey]

I don't much see the point of using them on tube amps.
Why?
To approximate what it's supposed to do, you must kill the gain of one triode stage, by lowering it (through feedback) from, say, 50X, to 1 or 2X.
It costs you a triode, which besides the cost, takes physical space, heater current, etc.
Dunno, the added "enhancement" of the mixing function doesn't pay (for me).
Now, in an SS preamp, it's very different, an extra op amp comes 2 per dip 8 for 60 cents, and eats extra 2mA, too little cost for a much better performance.
The typical op amp has from 100 to 300X (open loop) at 10KHz, increasing below it at 6dB/Oct; now *that* does provide for a good virtual earth mixer.
Just my 30 cents (1/2 TL072)

[walkman]

There is a spare unused triode in the amp i'm working on and it already has a cathodyne PI. I'd like to retain the original character of the amp but add a line in / speaker driven reverb return

[Steve Conner]

Well, you can't really adjust the gain of a cathodyne PI as it has a gain of 1.

So I'm assuming the question you're asking is, how do I set the gain of the virtual earth mixer? Answer: same way as you set the gain of an inverting op-amp. If the input and feedback resistors are equal, the gain will be 1. If the feedback is twice the input, the gain will be 2, etc.

JM's points are valid (as always ) but I've also used a virtual earth mixer with a 12AX7 for the same reason. I had a spare triode, and I found the usual reverb/FX loop return circuit annoyingly noisy, since it essentially throws away signal and amplifies it up again.

I think mine had a gain of 2, 3, 5 or something.

[kg]

i use a VE mixer as an effects return buffer.

by varying the feedback resistance it is also a (rather transparent) master volume.

[walkman]

Thanks for your replies.

So I'm assuming the question you're asking is, how do I set the gain of the virtual earth mixer?

I was wondering how much gain would be appropriate. Ive seen examples from Aiken , Merlin and O'Connor with different input and feedback resistors. Should I go with a 1M FB resistor and adjust the input resistors on the different inputs accordingly say 100k or 200k for different gains.

There is no need for an add grid leak as the inputs see the FB resistor as the grid leak from the virtual earth right?

i use a VE mixer as an effects return buffer. by varying the feedback resistance it is also a (rather transparent) master volume.

thats a great idea to try..

[pdf64]

'Should I go with a 1M FB resistor and adjust the input resistors on the different inputs accordingly say 100k or 200k for different gains.'
Depends what you want the input impedance to be; 100k may be rather low?

'There is no need for an add grid leak as the inputs see the FB resistor as the grid leak from the virtual earth right?'
Not sure what you mean? The grid needs a real dc path to ground; the virtual earth is virtual for ac only. The feedback resistor (between grid and plate cap) doesn't itself provide a dc path to ground.
However, you may be able to find a path to ground via the input resistors; if not a grid leak is needed in order to reference the grid to ground. Pete.

[walkman]

Its been a while since I mucked around with this amp. Here is the schematic for the amp i'm working on. The bolder lines are what i've added to the amp

The virtual earth mixer works well but doesn't really give enough gain, though i may try a 220k or 270k anode load and or reduce the cathode from 1.5 k

The power section is push-pull ecl84. With the NFB in place i get a gain reduction but also a very useful bass cut , like the effect of changing from a closed back 12" to an openback 10" cab. This removes much of the muddyness that was in the amp. Not sure how the bass cut is achieved as NFB is supposed to flatten response and give the amp a wider range. I guess the preamp section must have a bass emphasis that hs now been flattened. Is that what may be happening here?

ImageShack® - Online Photo and Video Hosting

[walkman]

Its been a while since I mucked around with this amp. Here is the schematic for the amp i'm working on. The bolder lines are what i've added to the amp

The virtual earth mixer works well but doesn't really give enough gain, though i may try a 220k or 270k anode load and or reduce the cathode from 1.5 k

The power section is push-pull ecl84. With the NFB in place i get a gain reduction but also a very useful bass cut , like the effect of changing from a closed back 12" to an openback 10" cab. This removes much of the muddyness that was in the amp. Not sure how the bass cut is achieved as NFB is supposed to flatten response and give the amp a wider range. I guess the preamp section must have a bass emphasis that hs now been flattened. Is that what may be happening here?

[J M Fahey]

Sorry but I fail to see the virtual earth mixer.

[walkman]

Sorry but I fail to see the virtual earth mixer.

your right.

At the moment the extra triode is only connected with a virtual earth. Originally I tried also returning the feedback loop and then main preamp, though the mixer but didn't seem to work very well. I'd be happy to use the extra triode as a line in and it works quite well that well.

What I'm interested in is why with the NFB from the power amp why is the bass reduced quite a bit?

[J M Fahey]

You are missing a triode stage at the power amp.
You do not have enough loop gain for feedback.
You are simply loading the 4n7 cap with a 100K resistor to ground, hence the bass cut.