What/ Weak Joe - re-designed balanced I/O schematic needs review

Hey guys,

as I wanted to have at least a couple of WHAT Compressors and coulnd't find any ready made PCBs to order, I started redrawing the schematic and added balanced input/output and a mix know.
Also I tried to get rid of all electrolytics in the signal path and replaced the 5532s with Opa2134s (which isn't really necessary of course).

As I wouldn't really consider myself an eingineer, I would be great if some of you guys could take a look at the input and output sections and the ratio and mixing pot bits. Also I'm totally sure about the poti orientation and if to use linear or audio ones.

Thank you, Sebastian

Hi... C16 and C17 should be the same, you could use matched resistors or a trimmer to improve CMRR, in a BBC paper I'd seen also a trim cap to improve HF/RF CMRR. Also I'd use lower resistors values. I don't know DRV134, I would go with a simgle op amp and 2 res/caps/res as recomended by jensen, you would be loosing 6dB output level, but is better about compatibility with unbalanced signals. It isn't a symetrical signal but it is balanced.

JS

I'd recommend the thatcorp 1646 for output. Great chip. You might also gain from A THAT chip for a line receiver also.

The THAT chip and DRV chip are interchangeable so you wouldn't have to change anything design wise. Sound and performance wise i've found them both great and much easier than trying to leave it unbalanced or build a balanced output with 5532s

Thanks so far! I already used the DRV134 in other design and I worked very well. Single-ended operation is also no problem, so you can use it un-balanced, too.
But again for the potis: I didn't touch this project in a long time and don't really remember how I came up with the coupling caps values and the "MIX"-circiut.
So if there is anything wrong with those, please let me know.
Sebastian

And another question. Does anybody know why the phase of the signal is inversed in the input and outputstages of the original weak joe design? Doesn't make much sense to me and I'd rather not do that to prevent caleation in my mix-circuit.

There must be something wrong with the gain amp, can't drive a current sensing input (virtual earth amp) directly from the output of a voltage output device... a lot of saturation and massive gain would occur there.

JS

IC3B Out can not connect direct to IC4A inverting input that way. (What JS is saying.)

Several pot values not noted?

> why the phase of the signal is inversed

I'm not going to go through and count flips and flops. Especially since it is balanced in and out. Tack it together, check phase. If not what you want, switch 2 wires, input or output your choice.

Thanks a lot for the hints - I just forgot to put a 6,81K in between IC3B and IC4A!

My phase problem question wasn't amaing at the input and output of the whole unit. But my compressed signal goes though a voltage follower and than an inverting gain stage. It is then mixed with the dry signal (Mix-Pot), so I would assume this cannot work, as the dry signal will be cancelled out (partly) by the inverted compressed signal.

So does it make more sense to have a non-inverting gain stage or just to put in another inverting opamp that doen't do anything except for the phease inversion?

Oh, an the pots would be:
Gain: 10K audio
Mix: 10K lin
Threshold: 10K audio
Attac: 10K lin(?)
Release: 1M lin(?)

Now, to avoid cancellation in the mixing pot, it shouldn't be a problem just to invert the input of IC3B and make it an inverter this way, right?