Finally got to build this sucker, haven't boxed yet and found a possible issue.
When I turn the Comp knob past half way it lowers the audio to mute.
What could be the likely culprit? I doubt this is designed and I will look for micro bridging.
R4, R6, R11? J-FETs? Controll or replace.
I think I found the issue, one of my leads isn't clipped short enough and appears to be touching the metal on the 9mm.
Generally it is better not to bend leads at the other side, but hold detail in position and clip leads cca 1mm upon pcb piont, and then solder. minimal chance for brodges or dirt around. And using rosin - it makes nice rounded drops, helping to good contact too.
It was just not clipped short enough, I am pretty good about not getting bridges, but this board has some tight tolerances and I did have to clear a few while building.
I am going to try it out and see if that was the issue.
That couple of J-FETs is very sensitive for setting. Maybe you was "happen" to get some ones, that are just too opened. Here is approach how to mend it. So, if issue is still here, I can describe it. Generally, alle FET trans are sensitive for static electricity while handling, so here would be problem too.
What is happening from what I can hear is when the Comp is turned all the way up it sends 98% of the signal to ground, when rolling it back there are some dead spots that go back to behaving normally after a half second.
These parts have been in a anti-static bag for maybe a month and when handling I kept my self in ESD.
The parts made these days can generally take a good static shock easily though, it's possible I've gotten a bad one, but I didn't build it for compression so for now I can live with it until I have more time and can find my tool to probe the board.
Maybe pot is wrong - did you measure it? But seems to me that issue is about those j-fets.
Pot should be correct, I sorted them on type and as far as measure I didn't see a reason to.
I guess I could remove the fet and check it, I am wondering if I have a micro bridge hidden under the 16mm pots, I put a little cardboard from the tape reel some caps came on under them and a dot of hot glue to keep them in place to avoid shorting on the pots.
When I have a chance I'll lift them and look/check.
OK, this will be useful. Here is another approach I would try:
See picture, I attached. Temporarly disconnect R10 (red X). Set Comp pot at maximum, that end which is on ground. Use trimmer 10k, I assigned with red point. With audioprobe at input (blue pointer) set now trimmer at position, that fet transistor starts to limit signal. Then connect again R10 and try funcionality. If OK, then fets are set properly. If not, fets are wrong. Controll them, or replace with another.
That is just how it works. The comp knob in the Triplet is a Dan Armstrong Orange Squeezer. I believe there is a note about it in the build doc
Quote from: Aentons on July 02, 2021, 12:36:06 PM
That is just how it works. The comp knob in the Triplet is a Dan Armstrong Orange Squeezer. I believe there is a note about it in the build doc
You are correct, I missed that entirely.
COMP: Sets the threshold point where the compressor kicks in. Turn this up to lower the threshold and increase compression. Due to the variability of the FETs in this circuit, this control intentionally has too much range...which means if you crank it all the way, the signal will cut out completely.
Any way to bypass it when the knob is at max and disable the Comp and keep signal throughput?
Quote from: Zerro on July 02, 2021, 09:56:40 AM
OK, this will be useful. Here is another approach I would try:
See picture, I attached. Temporarly disconnect R10 (red X). Set Comp pot at maximum, that end which is on ground. Use trimmer 10k, I assigned with red point. With audioprobe at input (blue pointer) set now trimmer at position, that fet transistor starts to limit signal. Then connect again R10 and try funcionality. If OK, then fets are set properly. If not, fets are wrong. Controll them, or replace with another.
Thanks for the help, but it seems it is supposed to behave like this.
Clearly, comp will limit dynamic of signal, if some peak level will be touched. It is only question of setting that j-fet, when it will start to do it - that threshold. That's why I recomended to set it individually with trimmer.
I don't have one to use, guess I'll have to come up with a physical limiter, maybe I could 3D print something to stop the knob at about half way.
I could probably incorporate it into the box and get away with it.
Anyone else found the snipped leads from a 1N4001 etc. work great in place of bus wire?
I had twisted together a few smaller resistor leads previously for the Triumvirate, but the 1N4001 leads work a lot better to hold the switch in place.
The double sided foam tape I am using under it is almost redundant. ;D