Flatline Compressor is oscillating... component changes?

[hooya]

I used the Vero layout on this page:
http://tagboardeffects.blogspot.com/2012/02/john-hollis-flatline-compressor.html
because it was the physically smallest board and fit in the box I wanted to use.

I used a VTL5C3 for the LED/LDR pair. This is basically a VTL5C2 with a slightly faster response curve. I used 1N4148 silicon diodes because that's what I had.

The circuit works. Compresses as I would expect. The problem is that if I turn either knob, Sustain or Volume, up too high, I get chirping and oscillation. The chirping goes away somewhat if I play. If I turn the volume knob on the guitar all the way down it will quiet itself. Volume up all the way on the guitar provides the most oscillation until I play. It seems to be related to the Vactrol I used, or the resistor going into the vactrol. Maybe. Not sure.

I noticed that the layout I used varies in a couple of component values related to the Sustain knob. There is a 47u cap and a 100k resistor instead of 10u and 10k respectively from the originally hand drawn schematic. That seems to be the only difference in what I put on as far as component values. I followed the layout, not the original schematic.

Any idea what might be causing the oscillation? Is it the vactrol? If so, should I increase the 300ohm resistor to a higher value? I have some other VTL5C3 to measure, but I don't know exactly how to measure them on my multimeter. Just resistance values? The LDR side of the 5C3 doesn't do anything... do I need to power the LED side? Looking at the charts of the 5C3 it should work just like a 5C2 just faster...

I also notice some schematics put a 1n4001 diode between +9 and ground. Would this do anything?

I can use the pedal as it is, but I just can't put the Sustain up as high as I want without chirping... A very mild compression. I can't get any significant "pop" effect if I wanted that. I could also note that I have tried different opamp chips to rule out a bad chip.

[hooya]

OK, I just checked the datasheet again. I read it wrong.

The 5C3 has a significantly higher (10x) off resistance. It also has a much higher on resistance than the 5C2 on the light side, but that doesn't seem to be my problem as much as when it's off. The data sheet is weird because the graphs aren't to the same scale...

So I'm sure there must be a resistor to change somewhere to better suit the 5C3, but I'm not sure where...

http://dl.dropbox.com/u/967492/My%20Flat%20Liner%20PCB.gif
Should the cap values fix my problem?

[hooya]

I changed the cap and resistor values to those in the second schematic. Also added the 1n4001 diode between +9 and ground. No difference at all in performance of the pedal.

It's like it's feeding back on itself at low or no volume levels. The pedal will not oscillate if the input signal is shorted though, if that means anything...

Anyone have any ideas?

[R O Tiree]

VTL5C3 has a very short recovery time (35ms), compared with that of the -5C2 (500ms).  Thus, with small signal/noise, the gain is ramping up rapidly, which lights the LED, which cuts back the gain, which turns off the LED, which ramps up the gain, etc, etc, etc... Possibly...  Does it chirp when you give it a large input signal?  Or does it do it all the time? 

At first look, I think I would try putting a 1k resistor in series with the LDR element of the Vactrol.  After that, I might try a 22k resistor in place of the 47k... if that works OK, maybe go to a 33k?  I'm looking to try to reduce the gain in the sidechain with those last 2 suggestions.  Lastly, I might look at a current limiting resistor in series with the Vactrol LED element, say a 1k pot to see how it behaves.

Or go buy a -5C2?

[hooya]

I'll try some of those values. I'm going to have to tear this thing apart since I built it on perf....
The thing is that the "Afterlife Compressor" from madbean pedals uses a VTL5C3, as does the other schematic I posted. Both of those use the normal resistor values around the Vactrol. So It's weird.

A 5C2 is pretty expensive as a one-off, and I had a bunch of 5C3s left over from when I was building Meatball clones, so I really wanted to use that. Better recovery and faster response should be better in the end anyway right?

Anyway, to answer your questions, it chirps when I don't give it a signal. The worst is when I have the volume on the guitar all the way up but am not playing. When I play, the stronger signal I give it, the more the signal takes the chirping out of the sound. Also, the way you describe the gain/LED cycle, the actual sound is very consistent with that description. Depending on the Sustain pot setting I can adjust the speed of that cycle. Basically.
So first step is increasing the resistance on the LDR side. OK.... no room on this perfboard... 

[midwayfair]

I'll try some of those values. I'm going to have to tear this thing apart since I built it on perf....
The thing is that the "Afterlife Compressor" from madbean pedals uses a VTL5C3, as does the other schematic I posted. Both of those use the normal resistor values around the Vactrol. So It's weird.

I've built two Afterlifes, and the second has the VTL5C3. Neither chirps. I would get something else is wrong, but you can occasionally get a bad vactrol. Check out some of the weirdness people have gotten in, say, the Tremulus Lune when they had a bad vactrol -- I think similar noises were discussed.

[R O Tiree]

So first step is increasing the resistance on the LDR side. OK.... no room on this perfboard... 

Just de-solder one leg of the LDR element and tack in a 1k resistor or a small pot between the free leg and the hole it should go into, rather than rip up and start again... Mind you, there may be a whisker of solder you haven't noticed, so starting afresh could be the final option...

[hooya]

I think I might start over. I could build using the same layout (it's been verified) and just use the values present in the Afterlife.

Funny, because I just built a Tremulus Lune and it's awesome. :-) And I've built about a dozen meatballs, each of which have two Vactrols. None have been bad before.

I'll try adding that one resistor and if it still chirps I'll just start fresh.

[hooya]

OK. Problem solved.

The Vactrol was probably fine. I took the board out of the enclosure and started solder pumping all over the place. Remember, I used a Vero layout on perf board, so I had a lot of solder bridges or bent leads to make many of the connections. I replaced the Vactrol and had the same problem. I was sure that since I was using values that were in verified circuits based around the VTC5C3 that it couldn't be that. There are already 10k resistors on either side of the Vactrol anyway, so another 10% on one side couldn't be the issue.

Then I started to just trim the fat all over the place. Any leads that didn't need to be there because I had put them in to make it easier to complete the Vero layout went away. A single one, I think bridging the 100k resistor between 9v and Sustain 2 with something on the input side maybe?, had a lead going where it didn't need to. I swear it wasn't making contact with anything else in the box. I looked over that thing with a microscope. But I clipped that extraneous lead that was just on the board underneath some other components, not even any solder around there. All of a sudden the pedal works. Perfectly.

I'm such a noob.

So if anyone is curious, reading this later, use the values in the "4-MTA Inhibitor" schematic I posted earlier (just the cap between Sustain 2 and ground should be 47uf, otherwise as Hollis recommends) with the VTL5C3 and it'll work fine. If you don't pull a hooya and leave flying leads all over the place.

[midwayfair]

Glad it's working! It's a cool little comp.

[R O Tiree]

Whisker of solder, small piece of wire...  There's always bloody something, isn't there?  Glad it's working, now

How are you finding the compression on it, though?

[hooya]

Whisker of solder, small piece of wire...  There's always bloody something, isn't there?  Glad it's working, now

How are you finding the compression on it, though?

I'm not really a guitar player much, although I'm trying to get back into it after a decade long hiatus. I never used compression before, so I have very little basis for comparison. It's also hard to test it against bypass right now as I don't have a working bypass switch (I am using a broken Xwing just as a placeholder).

So, with that being said.

It's a nice effect. The sustain knob definitely boost the overall volume, so the volume is more of a "balancing" knob than anything it seems. Set the compression where you want then balance against the bypass or leave it up to be a bit of a boost too. It's quite subtle up to about 3:00 on the sustain knob, which is actually nice. I plan on using this as an "always on" effect when doing chording and rhythm playing, so I don't think having it be too overbearing would be good for me. It does add a nice clarity to the front end of the note without being a "pop" too early in the turn of the knob. After about 3:00 on the knob you can get that big "pop" or "snap" that is usually too easy to dial in with a lot of the modeling compressors on the Line 6 units, which is my only real point of comparison.

So overall, perfect for what I need. It's actually pretty quiet and clean, although again it's tough for me to test against the bypass with my broken switch (new 3PDT is on the way). I had built a transistor based comp a couple years ago for fun. Hated how it sounded. Waayy to much squash and it started that really big pop sound at like, 10:00 on the compression knob and wasn't nearly as clean.

I may swap out the TL072 opamp for something more high end at some time. Suggestions?
Also looking for a suggestion for a more high end version of a 4558 for my Tremulus Lune that I also built recently.

My guitar board now has a Meatball, germanium NPN fuzz face, this compressor, and a Tremulus Lune. A total of 4 VTL5C3s. 

[midwayfair]

I may swap out the TL072 opamp for something more high end at some time. Suggestions?
Also looking for a suggestion for a more high end version of a 4558 for my Tremulus Lune that I also built recently.

There is absolutely no reason to change out the op amps in these effects. They're quiet and they amplify signal -- that's their only job in these circuits. They aren't used for distortion or overdrive. You're exceptionally unlikely to hear ANY difference, let alone a tonal difference.

... This is without even giving my opinion as to the benefits of changing out or obsessing over operational amplifiers in general.

[hooya]

I may swap out the TL072 opamp for something more high end at some time. Suggestions?
Also looking for a suggestion for a more high end version of a 4558 for my Tremulus Lune that I also built recently.

There is absolutely no reason to change out the op amps in these effects. They're quiet and they amplify signal -- that's their only job in these circuits. They aren't used for distortion or overdrive. You're exceptionally unlikely to hear ANY difference, let alone a tonal difference.

... This is without even giving my opinion as to the benefits of changing out or obsessing over operational amplifiers in general.

Good to know. Glad I said something.