Help with tube optical compressor (vactrol)

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

figuringstuffout

Well-known member
Joined
Dec 10, 2011
Messages
45
hey all...I'm in need of more help (so what else is new?), but this time with a tube compressor along the lines of an la2a.

So I got some Jensen transformers and wanted to play around with some compression circuits, and while it has largely been a huge success, I am having a pesky problem with quiet distortion that increases with the amount of compression.  I have attached a schematic of the circuit in hopes of getting some quick help for I imagine my ignorance of everything is contributing to the problem.

I am using a Vactrol (VTL5C2) for the compression and have also tried a full wave rectifier for the LED voltage as an alternate method (in place of what the schematic shows), but to no avail.

While the compression is awesome and really sounds good, I cannot figure out how to get rid of the quiet distortion behind the signal that gets louder as more compression is applied.  Anyone have any ideas how to fix this issue?

*Side note - the same kind of distortion but worse happens when I make another sidechain circuit to light up a blue LED as the volume increases....related?

Thanks!
 

Attachments

  • Compressor Test.jpg
    Compressor Test.jpg
    173.8 KB · Views: 210
that would make a lot of sense except for the fact that when I disconnect the entire sidechain and Vactrol (run just the 12ax7 as a stand alone preamp with the trafos) the signal is perfectly clean and sounds really good.


 
A good test might be to hook up a fixed voltage to the threshold pot (250k) and disconnect the SC driver tube (or ground the grid input). Then vary the compression with the pot and see if you hear the distortion problem. If you do hear the distortion, it would seem to be a problem with the line amp on it's own. If you don't hear the distortion, it would point to be a problem with the sc driver stage causing distortion on it's own, or interacting with the line amp.

I'd guess the problem might be in your sidechain design-  since the vactrol only works to compress for half the phase of your audio signal since the led only conducts one way. And there is no capacitance to smooth out the attack / release. I can't recall how fast that vactrol is, but you may want to experiment with adding an attack release circuit (R-C) to slow down the vactrol action and even out the compression.See if that gets rid of the distortion. You might want to read about the DAOC project by [Silent-Arts] since that had a similar sidechain to yours.
 
dmp said:
A good test might be to hook up a fixed voltage to the threshold pot (250k) and disconnect the SC driver tube (or ground the grid input). Then vary the compression with the pot and see if you hear the distortion problem. If you do hear the distortion, it would seem to be a problem with the line amp on it's own. If you don't hear the distortion, it would point to be a problem with the sc driver stage causing distortion on it's own, or interacting with the line amp.

I'd guess the problem might be in your sidechain design-  since the vactrol only works to compress for half the phase of your audio signal since the led only conducts one way. And there is no capacitance to smooth out the attack / release. I can't recall how fast that vactrol is, but you may want to experiment with adding an attack release circuit (R-C) to slow down the vactrol action and even out the compression.See if that gets rid of the distortion. You might want to read about the DAOC project by [Silent-Arts] since that had a similar sidechain to yours.



so i fixed the voltage from a completely different power supply (grounded together) and sure enough the distortion went away so indeed it is something with the SC.  I disconnected the resistor side of the VACTROL from the circuit but left the SC LED connection and the distortion remained.  I tried a 12au7 tube and the distortion dropped a little but about proportionally to the lower gain avaliable in a 12au7 than a 12ax7.  My next thought is maybe it is crosstalk or something?  It seems to be the LED lightning up that causes the distortion issue and so my next task will be to do my best to isolate the B+ for the sidechain tube as much as possible (bigger resistor off the first filter cap instead of a 10k after the signal 12ax7 stage)


any more ideas?
 
I would put the sidechain tube supply on it's own RC filter tap. You're probably dumping a bunch of crap into the b+ and it's making your line amp tubes sad...
 
gemini86 said:
I would put the sidechain tube supply on it's own RC filter tap. You're probably dumping a bunch of crap into the b+ and it's making your line amp tubes sad...


yeah it has been on its own RC tap but I just increased the R to 47k and nothing changed.  The only way to stop the distortion is to unplug the tube running the SC or disconnect the VACTROL LED side.....uuuggghhhh
 
I disconnected the resistor side of the VACTROL from the circuit but left the SC LED connection and the distortion remained.

This tells me that the distortion is coming from the sidechain tube - not the LED lighting up, since with the resistor disconnected the LED wasn't lighting up.
So some type of unwanted crosstalk/communication/feedback seems likely between your components.
There are several ways the sidechain tube could affect the audio path. It could through the ground path, the B+ path...
I'm not sure if this would be a problem, but I think the sidechain is set up with positive feedback on it's own, since the output of the SC gain stage is talking to the vactrol, which is connected back to the sc tube input.
Because your tube stage is inverting, a little negative voltage wiggle on the input of the sc grid causes an upward wiggle on the vactrol LED, which causes the vactrol resistance to wiggle down, which causes a bigger dropping voltage wiggle on the input of the grid. Unless I am wrong, it is set up with positive feedback. And you could just have some noise / oscillations going that distort the audio path from this design issue.
 
Just looked at the La2A schematic - it has two inverting stages (i.e. thus non-inverting) for the sidechain...
 
dmp said:
I disconnected the resistor side of the VACTROL from the circuit but left the SC LED connection and the distortion remained.

This tells me that the distortion is coming from the sidechain tube - not the LED lighting up, since with the resistor disconnected the LED wasn't lighting up.
So some type of unwanted crosstalk/communication/feedback seems likely between your components.
There are several ways the sidechain tube could affect the audio path. It could through the ground path, the B+ path...
I'm not sure if this would be a problem, but I think the sidechain is set up with positive feedback on it's own, since the output of the SC gain stage is talking to the vactrol, which is connected back to the sc tube input.
Because your tube stage is inverting, a little negative voltage wiggle on the input of the sc grid causes an upward wiggle on the vactrol LED, which causes the vactrol resistance to wiggle down, which causes a bigger dropping voltage wiggle on the input of the grid. Unless I am wrong, it is set up with positive feedback. And you could just have some noise / oscillations going that distort the audio path from this design issue.

not that i am disagreeing with you (i'm actually quite clueless), but why does this indicate to you it is a problem with the sidechain tube?  I disconnected the tube, installed another sidechain circuit to simply light up an LED and the distortion was again there increasing with the light.  When I disconnect the LED the distortion goes away.

I also tried adding another gain stage for the gain makeup in order to reverse the phase from the sidechain and the same problem remains.

I also rerouted all of the sidechain ground multiple times to different star locations and the result is the same.  All of my stars are routed to a single connection to the chassis.  My next step is actually to rip the whole thing out and try again.


Can it be determined for certain that my schematic is sound and shouldn't be producing background distortion?  If the schematic is good then I'm going to completely re-install the blasted thing.

Thanks!
 
I disconnected the tube, installed another sidechain circuit to simply light up an LED and the distortion was again there increasing with the light.  When I disconnect the LED the distortion goes away.
I didn't realize this. This test narrows it down a lot. Seems like the only way you would be getting distortion in this situation is from the LED ground being poor. If the LED side of the vactrol is grounded to the star ground, perhaps try moving it back to ground close to the input transformer ground [secondary (-)]. The input transformer secondary (-) should be grounded, by the way - it isn't in your schematic.




 
dmp said:
I disconnected the tube, installed another sidechain circuit to simply light up an LED and the distortion was again there increasing with the light.  When I disconnect the LED the distortion goes away.
I didn't realize this. This test narrows it down a lot. Seems like the only way you would be getting distortion in this situation is from the LED ground being poor. If the LED side of the vactrol is grounded to the star ground, perhaps try moving it back to ground close to the input transformer ground [secondary (-)]. The input transformer secondary (-) should be grounded, by the way - it isn't in your schematic.

Sounds good...i'll give it a try

you are right that the trafos are not grounded in the pic....they are indeed grounded in the unit.

 
as a follow up to this problem...I ended up ripping the whole thing apart and putting it all back together with proper grounding being my main focus and sure enough the problem with quiet distortion went away.  Man if the grounding isn't perfect things start to suck.....

thanks for the help everyone!
 

Latest posts

Back
Top