Sound Skulptor forums
Build help and troubleshooting => LA502T Tube Optical Compressor => Topic started by: Perihelion on September 30, 2026, 10:34:40 PM
-
Heya,
Just assembled and calibrated my 502T and have a couple questions about things that may or may not be an issue. All test points checked out good, and there were no meter calibration issues. I did the D1-D4 shorting mod and the r16 cap mod.
Quirky thing 1:
The release sounds fast, really fast, like all-buttons-in 1176 fast. The GR meter also shows this. It doesn't exhibit a logarithmic release like you'd think an opto comp would. I can link a video in a follow-up post.
Quirky thing 2:
Switching to 5ms attack from 10ms or 20ms adds about 3-4 dB of gain reduction on a constant 1kHz sine wave input. I understand a faster attack would do more GR on a music signal, but I've never seen that effect on a sine wave.
Quirky thing 3:
A 1.000v 1kHz sine wave at the input comes out at 0.924v on the output, compressor engaged with no GR, no output gain. Bypassing the compressor restores the output to 1.000v.
Quirky thing 4:
There is an audible 14kHz whine coming from the unit itself when powered on. With the cover off I can hear it about a half meter away. I didn't want to get my ear too close but it sounds like it's coming from the center of the board around the tube or opamps.
The unit otherwise operates like it should. All of the pots and switches do what they're supposed to do.
All of this kinda adds up to something shorted in the side chain or maybe a bad U7 or U8, but if you have any advice as to where to look first I'd definitely appreciate it. Thanks!
-
Here is the 502 on a snare channel:
https://youtube.com/shorts/raAIkmTgPdc?is=uK2wjD3vC5Dufp2W
-
1. The decay is set by the cell itself, it is not reflected in the LEDs which are driven by the side chain. It is not very fast, about 200-300ms.
2. Not normal. It should only change the gain by 0.5dB.
Check the solder joints and components value of R44, R46, C21, C22.
3. Normal. There is a minimal attenuation caused by the input transformer and the attenuation network.
4. Not normal. Can you identify the source of the noise by pressing a pen or something against the components, most probably one coil.
Check if any component is getting abnormally hot.
-
1. The decay is set by the cell itself, it is not reflected in the LEDs which are driven by the side chain. It is not very fast, about 200-300ms.
Ok!
2. Not normal. It should only change the gain by 0.5dB.
Check the solder joints and components value of R44, R46, C21, C22.
R44 metered as 4.67kohm and R46 metered right at 100kohm. C21 is a Wima 0.1 and C22 is a Wima 0.22. The solder joints look pretty good all around.
3. Normal. There is a minimal attenuation caused by the input transformer and the attenuation network.
Ok!
4. Not normal. Can you identify the source of the noise by pressing a pen or something against the components, most probably one coil.
Ah yes now that you mention it it's obviously inductor whine. It sounds exactly like my GPU, just not as obnoxious.
Check if any component is getting abnormally hot.
Nothing is abnormally hot. The ICs and U9 have a little bit of warmth to them, but nothing is anywhere close to the heat of the tube.
-
2. Do you measure the gain reduction from the LEDs or the actual gain reduction on an external meter?
It is true that the LED meter is not accurate when the decay time is increased. This is because it is driven by the side chain, not by the signal itself.
4. We haven't encountered this problem yet. Is the inductor whine annoying? If it is the case, we can exchange the module.
-
https://imgur.com/a/dLkYro1 (https://imgur.com/a/dLkYro1)
Here's an album with a few pictures of the build, if it helps.
-
Ok I had the time to do some more probing and here's what I have found:
Firstly, I re-calibrated the meter with the attack switch on 10ms.
Now with a 1kHz 1.000v input signal and 0.711v on the output, the -3dB LED on the meter is just lit (perfect!)
With this configuration and switching the attack to 5ms, the output drops a little bit to 0.675v, which is about a 0.5dB drop, much less than I initially thought. However on the meter the -6dB LED is now firmly lit.
(This is true on the entire meter range. For example if I'm doing 9dB of gain reduction in 10ms mode, switching to 5ms will show 12dB of reduction, even though the actual output change is about -9.5dB).
The other thing I have noticed is that in 5ms mode the GR LEDs are slightly dimmer than in 10 or 20 mode.
I've already re-flowed everything in the side chain circuit, so I'll take a closer look at the meter PCB and the 10-pin header.
If the meter is not accurate at higher speeds, that's perfectly fine! Like I said all of the switches seem to do what they're supposed to.
In regards to the coil whine, I might take you up on a replacement sktps2. The whine stops if the board is removed.