Engineer's thumb and JMK panner

[teknoman2]

Hi guys, long time to post here.
My good old friend Ettore_M gave me as a present his amazing engraved Engineer's thumb compressor, check this marvelous build.
He did an outstanding job.

http://www.madbeanpedals.com/forum/index.php?topic=11664.msg105300#msg105300

So I was thinking adding a blend knob to this great compressor, so I took JMK panner and put it with the compressor.
http://jmkpcbs.com/wp-content/uploads/2014/03/Panner.pdf
Of course it didn't work as expected... so I started reading posts in the forum.
Everyone said that compressor has phase inversion and obviously we can't mix-blend different phase signals.
The result was volume drop and tone sucking.
After that I put an inverting amplifier before the volume control of the compressor so I can flip the phase 180 degrees.
The result was the same as before, in the middle position where is 50% wet 50% dry I had again volume drop and tone sucking.
Since now the phase is the same I am thinking there is probably an impedance issue.

Do I have to omit the inverting op-amp and put an emitter follower buffer after the compressor volume control? Any thoughts on that?

[teknoman2]

I ve got second thoughts regarding the emitter follower buffer after the volume control of the compressor.
I think its useless since the compressor by itself has low output impedance.
Also I did some tests using multisim, and I found out that Engineer's thumb doesn't flip the phase.
The input sinewave doesn't have 180 degrees phase comparing to the output.
Is that correct or am I missing something here?

[Scruffie]

The Engineers thumb is indeed non-inverting and the panner is also non-inverting.

The Panner only has 100k input impedance so that explains the tone sucking, you could make R1 & 2 - 220k+ to improve it a bit although the larger the resistors the more thermal noise.

Yes I would put a non-inverting buffer after the thumbs volume control.

[teknoman2]

Thanx buddy for the answer really appreciate it.
You are totally correct regarding panner, it does have low input impedance, so I am thinking taking the high input impedance version of geofex to tame the tone loss.

http://www.geofex.com/Article_Folders/panner.pdf

Now regarding buffer after thumbs volume control, since thumbs output impedance is relative low does it really need a buffer?

[Scruffie]

Too lazy to actually check and think it through right now but resistors in series and parallel and whatnot interacting I should think rather than impedance.

[teknoman2]

Awesome!! I will make those changes and hope everything works great. 
I will post my final impressions

[teknoman2]

Finally the boards arrived and populated.
The blend is working as expected the thing is that I ve faced two issues so far.
First one which is solved, I used TL072CN for all the opamps and the decay of the compressor was a distortion rather than a compressor.
I changed all of them with TL072CP and everything is normal. All the volatges are good and the compressor is working as expected.
Now the second issue that I can't find a way to solve it is so far is that when the ratio is above half way up ( I use a 500kA for the ratio pot) there is always static noise when I don't play anything.
I saw an old thread of aballen who post a video also.
Here it is.
https://www.youtube.com/watch?v=FQfFmRRfLJ0&app=desktop

In my build I don't have distortion only the static noise.
The noise appears after the 10k resistor of pin3 of OTA.
Since this is a very quiet comp does anyone know does this happening?

[lars]

Hi guys, long time to post here.
My good old friend Ettore_M gave me as a present his amazing engraved Engineer's thumb compressor, check this marvelous build.
He did an outstanding job.

http://www.madbeanpedals.com/forum/index.php?topic=11664.msg105300#msg105300

"I would have liked to have seen Montana."
~ another post that has been pho-bucked ~

[jubal81]

I'm suspicious it might be radio signal. Opamp inputs are very sensitive and those bias net traces become antennas, so it's important to keep the traces as absolutely short as possible.

[teknoman2]

I'm suspicious it might be radio signal. Opamp inputs are very sensitive and those bias net traces become antennas, so it's important to keep the traces as absolutely short as possible.

thanx for the reply jubal81.
I know exactly what you mean, I was afraid of that..
Even when I put the backplate of the enclosure I had the same issue.
So you are saying that I must keep the bias net short as possible to eliminate this issue.

Since it's a dual layer pcb I was thinking, putting all the bias traces in first layer and the audio traces in the second.
Another thought is to make a dual pcb sandwich using pin headers, one for the mount knobs, power section and blend circuit
and the other one for the compressor section.

Or the simplest way is to keep the compressed signal as far as I can from the output jack.