PCBs for 1176-flavoured FET compressor w/ opamps - (almost) no wiring

[Mike Havok]

It should work but the input signal would be 6 dB lower with the DOA (the website says it's a -6 dB receiver). You'll be getting slightly less compression/distortion that way.

I happen to have a spare PSU part  ;D

Thanks for the reply!

Their page confused me a bit to be honest, it says that if it's balanced with a transformer wired 1:2 that will provide unity gain signal. Then later on it says the same setup will provide "+12dB of gain in relation to the card's input signal" so I'm not sure if they can both be right or not 

Ah, thats not ideal then, I had thought that the 6dB drop was to account for the 2 signals being combined into one, and that was why people use the THAT1246 more often than the 1240 (I'm still learning!  ;D)

I would definitely prefer to have the usual amount of compression & distortion (cos that's what we love about this compressor!) So would I be able to change some component values to account for the voltage drop? Maybe changing the value of R10? (Sorry, I'm still pretty new to this)

That would be great if I could get a PSU board too thanks. How much would it be for 1 board and the PSU?

Thanks!

 

[fripholm]

I would definitely prefer to have the usual amount of compression & distortion (cos that's what we love about this compressor!) So would I be able to change some component values to account for the voltage drop? Maybe changing the value of R10? (Sorry, I'm still pretty new to this)

There's nothing you can do on the OP76 board to compensate for the level drop. This has to be done within the DOA circuit before entering R10. Do you happen to have a schematic for that?

That would be great if I could get a PSU board too thanks. How much would it be for 1 board and the PSU?

PM'd you.

 

[Mike Havok]

Thanks for the PM 

I don't have a schematic for the board but I will try and trace it out tonight and get it posted

 

[Mike Havok]

A little late but here is the schematic, apologies on the size but I had to create it on my phone!

 

[fripholm]

This circuit should give the same output level as the original input stage on the board, so you're good to go 

 

[Mike Havok]

This circuit should give the same output level as the original input stage on the board, so you're good to go 

Awesome! 

Thanks for checking it over

 

[grid_stopper]

Just sent an email to (hopefully) purchase some PCB's. very excited! ;D

 

[absrec]

I just built the first of two (2) OP76s. It passes signal and sounds about like any 1176 I've ever heard.

The problem I'm having is with the GR meter. Long story short... I put the LM339s in upside down. I caught it before I powered it up for the first time so I think I managed to save myself blown ICs. However, the weren't in sockets and through the desoldering process, a couple eyelets came out. I had to install a couple jumpers and have done a continuity test and everything seems to be connected to where it needs to go. Unfortunately, every light on the GR meter is on. I checked the polarity (with a 3V coin cell) and that seems to correspond to the schematic and layout diagram. Also, this is the case no matter whether I have a ratio selected or have the ratio switch in the OFF position.

After measuring the voltages on the outputs of U8 & U9... starting with U8 pin 2 - pin 1 - pin 14 - pin 13 each output is ~2VDC apart. Same for U9. When I measure all the way up the string of LEDs, I get the same readings - 2VDC apart. This makes perfect sense for a 1db/1V scale. Or thereabout.

I see that the string of resistors (R56-R64) seems to have the inputs of the eight (8) comparators biased at their + inputs. These appear to be separated by somewhere between -150 & -200 mVDC. Hard to tell if the actual voltages are correct but... the fact that they are in even increments makes me think they are.

What is hanging me up is the output of U10 is reading roughly -15.4 / -15.6 VDC. This is with no signal going through the unit. I feel like this should be close to 0 if no gain reduction is happening since this voltage is supposed to be representative of what's happening in the sidechain. If this is correct, the issue must be earlier in the detector part of the circuit. Which is puzzling because everything appears to be correct, layout-wise. I double checked a bunch of resistor values and diode polarities and I'm just stumped at this point.

The only other thing I can't think is that maybe I messed up my LM339s by de-soldering and re-installing them. I don't think that's right though because -

1) It is supplying the correct voltages to the various LEDs.

2) If I momentarily drop my probe across the - & + inputs of any of the channels, the respective LEDs turn off.

Also, the GRNULL & GRSCALE seems to be working when I probe and adjust but they don't seems to have any effect on the on/off state of the LEDs.

If anyone's got an idea of something to check or try, I'm all ears.

Thanks,
-Aaron

 

[fripholm]

What is hanging me up is the output of U10 is reading roughly -15.4 / -15.6 VDC. This is with no signal going through the unit. I feel like this should be close to 0 if no gain reduction is happening since this voltage is supposed to be representative of what's happening in the sidechain. If this is correct, the issue must be earlier in the detector part of the circuit. Which is puzzling because everything appears to be correct, layout-wise. I double checked a bunch of resistor values and diode polarities and I'm just stumped at this point.

The lower the voltage at the output of half-wave rectifier U10 (which is actually the anode of diode D7) the more LEDs will be turned on. At -15VDC it makes perfect sense that all LEDs are lit. On one of my units this is at about -40mV.

Trimmer GRNULL should control the offset on U10's output. Does it change when you turn it?

 

[absrec]

Not in the least. I tried both trimmers - GRNULL & GRSCALE. Like I said above, they seem to be working because I could measure the voltage across them and see it changing as I tweaked the pot.

 

[absrec]

The lower the voltage at the output of half-wave rectifier U10 (which is actually the anode of diode D7) the more LEDs will be turned on. At -15VDC it makes perfect sense that all LEDs are lit. On one of my units this is at about -40mV.

Trimmer GRNULL should control the offset on U10's output. Does it change when you turn it?

CORRECTION -

By turning GRNULL & GRSCALE fully CW, I managed to get the output (PIN1) of U10 down to around 377mV. All 8 LEDs are still fully lit up. PINs 4, 6, 8 & 10 of U8 & U9 are reading ~1.38V

 

[absrec]

After reading the last page of the build manual, I thought I had found my error. The paragraph about turning C20 around looked promising and I was reading a negative DC voltage on PIN 7. After flipping it over, it is still reading exactly the same. -1.7VDC with no input signal. That seems like a healthy bit of offset. Once again, I’m a little stumped. This isn’t my 1st (or 5th) build, I promise.

 

[scott2000]

Wasn't there an issue with certain 339 brands?? Not sure if it's relevant here.....

 

[TillM]

Yes, it was me who had issues with the 339.
Thomas write it in the Build manual, i don’t have it in mind if it was TI or Semi.
I think I had problems with TI LM339

 

[absrec]

Yes, it was me who had issues with the 339.
Thomas write it in the Build manual, i don’t have it in mind if it was TI or Semi.
I think I had problems with TI LM339

I think mine is the TI version. LM339AN I believe. I thought it was recommended in the build guide.

 

[TillM]

Yes, this worked.
I had trouble with the LM339N.

 

[fripholm]

After reading the last page of the build manual, I thought I had found my error. The paragraph about turning C20 around looked promising and I was reading a negative DC voltage on PIN 7. After flipping it over, it is still reading exactly the same. -1.7VDC with no input signal. That seems like a healthy bit of offset. Once again, I’m a little stumped. This isn’t my 1st (or 5th) build, I promise.

If everything is correct, this offset can't make it into the sidechain nor the GR meter section so this isn't likely to be your problem.

You wrote earlier:

All 8 LEDs are still fully lit up. PINs 4, 6, 8 & 10 of U8 & U9 are reading ~1.38V

If I'm reading this correctly, you were able to adjust the DC voltage on the inverting inputs of the comparators from around negative 15 to more than +1 VDC? Sure, it's positive 1 VDC?! It's just, that with the given circuit, the voltage shouldn't go above zero.

Anyway, that range seems to be just fine without any input signal. The first LED turns on, when PINs 4, 6, 8 & 10 (which is the output of the half wave rectifier) get more negative than -150mV. This is the lowest voltage the LM339 compares its input to. Between zero and -150mV all LEDs should turn off. Try to find this spot with the trimmers and go from there.

 

[absrec]

You wrote earlier:

If I'm reading this correctly, you were able to adjust the DC voltage on the inverting inputs of the comparators from around negative 15 to more than +1 VDC? Sure, it's positive 1 VDC?! It's just, that with the given circuit, the voltage shouldn't go above zero.

Sorry... negative -1.3VDC is what I should have written. That’s the lowest it’ll go. The LEDs still don’t turn off.

 

[fripholm]

Yeah, that's too far off the -150mV for them to turn off.

You mentioned turning both trimmers fully CW. What happens when you turn GRSCALE the other way? Does it lower the voltage (more negative) at U10 or does it get closer to zero (which we want)?

What did your QBIAS end up at (measured at the right side of R48)?

 

[absrec]

You mentioned turning both trimmers fully CW. What happens when you turn GRSCALE the other way? Does it lower the voltage (more negative) at U10 or does it get closer to zero (which we want)?

Measuring U10 PIN 1 to GND the voltage goes from +365mVDC (Fully CW) to +438mVDC (Fully CCW). GRSCALE is the recommended 100k Trimpot.

Additionally, measuring U8/U9 PIN 4,6,8,10 to GND - I'm getting a reading of negative (-)1.38VDC and turning GRSCALE doesn't have any affect on the reading.

What did your QBIAS end up at (measured at the right side of R48)?

From R48 to GND I got negative 3.347VDC.

Just to make sure we're talking about the same point, I included a screen shot. The other side was pretty close. Maybe 200mV difference. Not 100% sure I understand what the QBias even does. I did adjust it along with the distortion trim according to the build guide. Both of those adjustments worked exactly as you said they would.