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

[fripholm]

That’s strange. Turning GRSCALE should definitely result in a change, also at the QBIAS voltage you are getting.

Yes, you were measuring the right spot.

Are your FETs matched? The QBIAS setting puts the compression FET into slight conduction and is different for each FET. For the GR meter circuit to show the correct amount of compression, the second FET needs to be matched to the first one. When idle, the QBIAS voltage gets added through the second FET & the GRSCALE trimmer to an offset provided by GRNULL.

Can you post a high resolution photo of the complete right hand side of the stuffed board (the one with the comparators and U10)? Maybe I can spot something that’s not quite right...

 

[absrec]

Is this clear enough? I feel so vulnerable. It's like sending someone a close up of your own face.

 

[absrec]

Are your FETs matched?

If I did it correctly. I followed your FET matching scheme and picked the closest two. They seemed to vary more in voltage than current from device to device. I only had 10 to get a match but I actually found 3 pairs that seemed within the 10% tolerance stated in the build guide. I'd say more like 1-2%. Basically the readings were about identical on the ones I ultimately considered pairs. I think the Vp was around 3.86-3.87V and Idss was .46-.48mA on the pair that made it to the board. I'm just going by memory so if those readings seem atypical, I may have to go do the test again and report back.

 

[fripholm]

Thanks, the picture is clear enough. I’m not in the studio at the moment - will check it in the morning.

 

[absrec]

UPDATE - It was apparently a problem with the LM339ANs. They are the TI brand. I ordered more LM339s a few days back and they showed up a few hours ago. I swapped them out and now it works. This just confirms what others said in earlier posts. I have the NSC (?) version I believe that's who makes it. They say SM8824 as well as LM339AN on the ICs.

Thanks Thomas! Great sounding project. Thanks for your time & guidance.

 

[fripholm]

Awesome!

This is the exact opposite of what I recommend in the guide after TillM ran into something similar with the type of comparators he was using. To double check I've just replaced the ST LM339N in my mono unit with TI LM339AN (presumably the ones that gave you trouble) and they work just fine. I haven't found a hint for any differences in any of their datasheets either...

In other words, there's no general recommendation what to use and what to avoid when it comes to quad comparators. When in doubt, just try a different brand.

Anyway, I'm glad you've got it sorted!

 

[scott2000]

Would be neat to know what is going on with these comparator gremlins.... There must be some reason I'd imagine. Maybe just some goofy chips or , in this case, the desoldering maybe....
Is there something before being calibrated that could cause damage?


Of course it's moot now.....

 

[absrec]

Awesome!

This is the exact opposite of what I recommend in the guide after TillM ran into something similar with the type of comparators he was using. To double check I've just replaced the ST LM339N in my mono unit with TI LM339AN (presumably the ones that gave you trouble) and they work just fine. I haven't found a hint for any differences in any of their datasheets either...

In other words, there's no general recommendation what to use and what to avoid when it comes to quad comparators. When in doubt, just try a different brand.

Anyway, I'm glad you've got it sorted!

The ones I’m using are marked LM339AN. They are just a different brand.

 

[absrec]

Would be neat to know what is going on with these comparator gremlins.... There must be some reason I'd imagine. Maybe just some goofy chips or , in this case, the desoldering maybe....
Is there something before being calibrated that could cause damage?


Of course it's moot now.....

Could be exposure to excess heat. Maybe they are just sensitive devices. Not moot if anyone else decides to build one of these units. I would just be sure to use IC sockets so if there is a problem, you don’t have to risk damage to the chip and PCB. Lesson learned.

 

[kante1603]

Would be neat to know what is going on with these comparator gremlins.... There must be some reason I'd imagine. Maybe just some goofy chips or , in this case, the desoldering maybe....
Is there something before being calibrated that could cause damage?


Of course it's moot now.....

I agree with Scott2000,something is strange here.
Damaged by excess heat while desoldering,o.k.......but something must have happened before.
Now I seriously wonder:

a)are those chips sensitive for ESD
b) what happens if they were mounted the wrong way around?

Best regards,

Udo.

 

[absrec]

I agree with Scott2000,something is strange here.
Damaged by excess heat while desoldering,o.k.......but something must have happened before.
Now I seriously wonder:

a)are those chips sensitive for ESD
b) what happens if they were mounted the wrong way around?

Best regards,

Udo.

I kept checking the component layout over & over for errors like incorrect values, incorrect polarity, etc. but never managed to find anything. Once I changed the comparator ICs, the circuit began to function as described.

I never applied power to the board when the chips were installed upside-down. I found the error and corrected it prior to the initial smoke test.

 

[grid_stopper]

Curious if anyone has found a mouser part or digikey(or-if anyone knows where else I can get them in canada!) part # for 2x13 header and socket? I haven't had any luck searching, but its entirely possible I'm just using their search function wrong.

 

[absrec]

Curious if anyone has found a mouser part or digikey(or-if anyone knows where else I can get them in canada!) part # for 2x13 header and socket? I haven't had any luck searching, but its entirely possible I'm just using their search function wrong.

Not from Mouser but… This is what I bought.  It’s ab assortment of different headers, all with a 2.5 mm spacing. Just snap or cut down to size depending on what you need 

Glarks 112Pcs 2.54mm Male and... https://www.amazon.com/dp/B07CWSXY7P?ref=ppx_pop_mob_ap_share

 

[grid_stopper]

Not from Mouser but… This is what I bought.  It’s ab assortment of different headers, all with a 2.5 mm spacing. Just snap or cut down to size depending on what you need 

Glarks 112Pcs 2.54mm Male and... https://www.amazon.com/dp/B07CWSXY7P?ref=ppx_pop_mob_ap_share

this is perfect, thank you!

 

[fripholm]

https://www.mouser.com/ProductDetail/TE-Connectivity-AMP/1-215309-3?qs=sGAEpiMZZMs%252bGHln7q6pm8SOCK6aAoLgbIkv1jVEs%252bM%3D

Yes, this one should be fine.

 

[Catrunes]

If i wanted to use the PSU board to power say a Mixbuzz 500, any problems? thinking about boxing up that at the same time i build this, itd make mixing on an M32 that much more fun.

 

[fripholm]

I'm not familiar with the MixBuzz500, but as long as it just needs bipolar +/- 15 or 16 volts, the PSU of the OP76 should be fine. If it also needs +/- 12 volts rails (like the SSL bus comp this is based on), these must be provided elsewhere.

 

[Catrunes]

fripholm, happen to have a to scale pdf of the pcbs? im being ocd af about boxing this up

 

[Ricardus]

This is intriguing. What's the status of this and availability?

 

[fripholm]

fripholm, happen to have a to scale pdf of the pcbs? im being ocd af about boxing this up

Yes, attached 

This is intriguing. What's the status of this and availability?

The boards are still available. See the first post in this thread or email me at fripholm [AT] gmail [DOT] com for details on how to order.