Lab Series L5 Preamp +/-9V conversion: works great, but compressor LED always on

[aion]

I am working on a +/-9V conversion of the Lab Series L5 preamp's 2nd channel. It uses a LT1054 for the negative voltage and substitutes a LM13700 for the CA3080 and CA3094. Otherwise it's the stock circuit and identical to my earlier L5 boards that have been out in the wild for 4 years.

Sound-wise, it's great - everything works as it should.

But, one cosmetic issue: the compression indicator LED is always on no matter what I do. It does increase in brightness when the compressor kicks in, so it behaves correctly in that respect. It just doesn't turn completely off when it's not compressing or when the compressor switch is off.

Here's the schematic - the affected area is in the bottom right:



It's not just a faint glow, either - it's maybe the equivalent of using a 10-15k CLR with a normal diffused LED and then reducing it to 3k3 when the compressor is on. There is a noticeable change in the level of LED brightness, but in its resting state it's still bright enough to serve as a reasonable power indicator LED.

Here's what I've tried so far:

- Tweaked the limiting resistor (R50), using the stock 680R as well as 1.5k and 10k. The overall brightness of course changed, but nothing else, still not off when it should be off.

- Used different types of LEDs, in case it was related to efficiency/brightness

- Triple-checked the orientation of the 2N3906 and MPSA13 in the layout (I don't use either of them often, so I originally wondered if my Eagle libraries might have been in error)

I also have two different units doing it, one an adaptation of channel 1 and one an adaptation of channel 2, which have completely different PCB layouts but share the back half of the circuit including the compressor. So it's not related to the physical layout.

As I said, the two major differences were dropping the supply voltage and changing the OTA.

I didn't change any other values in the circuit, so maybe there are places elsewhere that need to be adjusted to compensate for the 40% drop in supply voltage, perhaps in the compressor's rectifier.

Could it be related to that? Where should I be looking?

[PRR]

> the two major differences were dropping the supply voltage and changing the OTA. I didn't change any other values in the circuit

You changed from a 0.6V OTA to a 1.2V OTA. *And* you changed from a dull 1979 LED to a hot 2019 LED.

A 100k B-E with the existing 100k will roughly compensate the added Vbe. But even the leakage of an "off" Darlington may light a modern low-leakage high-efficiency LED. A ~~1K across the LED will suck-out 1.6mA of current, which still responding linearly-enough to higher currents.

[aion]

> the two major differences were dropping the supply voltage and changing the OTA. I didn't change any other values in the circuit

You changed from a 0.6V OTA to a 1.2V OTA. *And* you changed from a dull 1979 LED to a hot 2019 LED.

A 100k B-E with the existing 100k will roughly compensate the added Vbe. But even the leakage of an "off" Darlington may light a modern low-leakage high-efficiency LED. A ~~1K across the LED will suck-out 1.6mA of current, which still responding linearly-enough to higher currents.

Thank you! I'll test this out tomorrow and report back.

[swamphorn]

This entire circuit is fascinating to me; the '13700 is one of my favorite chips to design with. In the first stage, it looks like it adds even harmonics by modulating the amplitude of the signal with a high-passed version of itself?

I did calculations for tweaking R_26,27 and C₁₂ to compensate for the additional V_BE of that stage as well, but the adjusted value of 29 kΩ is close enough that I don't think it would make an appreciable difference. I would be tempted to replace Q_1,2 and the surrounding resistors with one of the OTA buffers.

[aion]

Paul's fix did the trick!

I didn't need the 1k resistor across the LED because I was using a vintage-style diffused type, but I will tuck that one away for the future just in case.

Thank you very much!

[ljudsystem]

This is great. I've been thinking about getting the Aion L5 pcb from Musikding but if there's a new version on the way that can run of a standard 9v psu I will hold out for that one.