LA-2A opto cells switching

[beatnik]

I am designing an LA-2A PCB with two switchable T4B opto cells, to select between a fast and a slower version.
I have attached a schematic of what I've come up with so far along with an original LA-2A schematic for reference.

Basically I am using a pair of DPDT relays to make the switching.
One pole is switching the 0V reference for both the audio LDR and the EL panel in the input section of the T4B. The other three poles are switching the metering section.

I am positive this should work well but a second opinion would be certainly welcome.

 

[rock soderstrom]

I am positive this should work well but a second opinion would be certainly welcome.

Interesting project! I would probably switch the sidechain connection too.

 

[beatnik]

I'm trying to avoid a third relay, cause I'm really short of space on the PCB.

The sidechain signal is always connected to both EL panels but switching the 0V reference should be keeping only one EL panel on at a time, this also saves consumption on the EL panel that is not used (they tend to degrade with use)

 

[beatnik]

I have been playing around with the above circuit, the switching works as expected, compression seem to work properly and I can zero the meter on both settings using the respective calibration.

There is a major issue though, when the switching is happening while there is signal in the sidechain, this triggers oscillation and distortion, which remains for a while.

This behavior seems to be related with the signal being in the recovery portion of the compression. If switching when the compression is fully recovered (it takes a while to go back to 0 on the meter particularly with a slow T4B) then this problem doesn't seem to happen.

I have no idea what is causing this oscillation, do you think I should consider adding an extra relay to switch the sidechain signal and/or the audio LDR ?

I am actually wondering whether the Omron G5V-2 I am using are make-before-break, or break-before-make.
I would assume they are non-shorting type, but in fact if they are shorting type that might be a problem.

 

[Script]

Switching the LDR to audio path out of circuit avoids both panels acting at the same time (two panels fighting each other?). However, when suddenly switching out the recovery time of one panel, I'd assume you will very likely get a loudness hiccup (briefly back to unity gain) that can be even more annoying.

Also, aren't the two LDRs in parallel? So switching one out affects the overall resistance?

 

[Bobby Baird]

Check this out. My Opto 6 does this with a simple switch no relays no random loud sounds are pops.

 

[beatnik]

Check this out. My Opto 6 does this with a simple switch no relays no random loud sounds are pops.

Thanks. Are the audio LDRs (pin7 on the T4Bs) just joined together ?
I guess the T4B that is dark has high resistance so when in parallel with the lighted up T4B which has low resistance then the overall resistance is dictated by the T4B that is giving the lower resistance ?

What is happening with the metering section ?
Looking at the Drip manual it's showing only one set of calibration controls so I assume the meter LDRs are also connected in parallel ?
Doesnt the meter change gr 0 position and tracking when switching between the two settings ?

 

[Bobby Baird]

It's in the rack now but from what I remember both pin 7 was connected together with a trace I believe. Concerning the metering is the caveat. I spoke with Bill from Kenetek about this in the white market thread. My question post #18 was about how to get the 2 different cells to match to have the same calibration on the meter. His reply is post # 19. https://groupdiy.com/threads/new-ke...or-your-la-2a-la-3a-and-similar-builds.72265/. Basically you would need 2 meters for the separate cells. My 2 cells are off by about 3 to 4 db. My thought was that is way to much trouble for a meter. You should rely on your ears to tell you which way to turn the knob. I do need to take it out of the rack to put in the de-ess passive band-pass circuit I drew up.

 

[Bobby Baird]

Here's a picture I found on the internet of the traces on the back of the opto 6 board compared to the silkscreen. you should be able to figure out the implementation of the switching. Looks like pin 3 goes to SPST T4B selector switch.

 

[beatnik]

It seems like the Drip PCB is following exactly the design in the schematic you posted above (without the "both" and "off" settings)
LDRs both in audio path and metering are paralleled and the switch is alternating the side chain signal between the two EL panels.

I have tried it this morning and while the metering barely drifts a half dB between a Kenetek "Classic" and "Good Fast"
when alternating these with a vintage Teletronix T4B or my homemade version, which both have a slower response, the meter tracking is indeed offset.
In fact I found the difference between the two Keneteks to be quite subtle sound wise. So that might explain why the metering aligned so well.

The information from Bill you are referring to confirms exactly what my initial thoughts were. To maintain meter tracking when using T4Bs with a completely different timing response requires two separate metering circuits.
I have hacked the PCB and this updated schematic seems to work well.

Now the question is whether it's a problem or not to have the dark LDRs with high resistance in parallel with the bright LDRs.
If I unplug the dark T4B I don't notice any level change and the compression amount stays the same, so in practice it seems to be fine.