Help Debugging Ross Compressor Clone Build

[pork ruckus]

           Hey I'm not sure if I'm posting this in the right place but I'm looking for a little bit of guidance to get my recent Ross Comp clone up and running. I've spent quite a bit of time reading as much as I could on this pedal and similar issues other people have encountered in their builds but have yet to get this bad boy running.

So far I have:
- Triple-checked to make sure I have all the right components in all the right spots as well as no shorts between traces.
-Removed and tested all polarized capacitors, transistors and diodes - I confirmed that they were all working and put them back in the board.
-Confirmed all areas that should be getting 9V are and all components that should be connected to the ground bus are.
- Audio probed and found my audio signal seems to drop at the CA3080.
- Tried using 2 different CA3080 chips purchased at a local electronics store here in Canada - neither worked.
- Tried putting CA3080 in backwards, curiously in this orientation the readings are closer to the Fuzz Central V readings for this build...
- Read that there have been faulty CA3080 chips in the past but the thread I was reading this from was from 2007 -  I'm not sure if this is an ongoing thing?

I've posted my voltage readings and substitutions below, if someone would be able to lead me in the right direction I would greatly appreciate it. Thanks to all those that have contributed to this forum as well, it has been invaluable in my building education!

Comparous from Tonepad

http://www.tonepad.com/getFile.asp?id=9
https://fuzzcentral.ssguitar.com/ross.php

Etched the pcb myself, used the Fuzz Central mods for the Ross Comp

Substituted 50nf off of Q2 Emitter with 47nf cap and 2K trim pot for 5K trim pot otherwise all to Fuzzcentral spec
Q1-5: 2n5088
D1: 1n4002
D2-3: 1n914

Signal makes it to amp in bypass mode, when engaged signal makes it to P2 of CA3080, is very quiet at P3-Cc (Tonepad layout), no signal at and after P5 & P6.


Center negative Wall wart = 9.36V

9v pad on PCB =9.36V
IC:
1: 0.00V
2: 1.44V
3: 0.84V
4: 0.00V
5: 0.78V
6: 0.89V
7: 9.36V
8: 0.00V

Q1
C: 8.59V
B: 1.30V
E: 0.76V

Q2
C: 8.97V
B: 0.89V
E: 0.35V

Q3
C: 8.19V
B: 0.00V
E: 0.00V

Q4
C: 8.19V
B: 0.00V
E: 0.00V

Q5
C: 9.36V
B: 8.19V
E: 7.62V

D1
A: 0V
K: 9.36V

D2
A: 0V
K: 0V

D3
A: 0V
K: 0V

IC Backwards:
1:0V
2:4.83V
3:4.82V
4:
5:8.88
6:1.81
7:9.36V
8:0V

[italianguy63]

Yes.  Lots of faulty CA3080 chips out there...

Are you using the CAN (metal) style or the plastic DIP style?

I have had (much) better luck finding the DIP style that still work.

MC

[Slowpoke101]

Welcome to the forum. I hope that you will have fun here.

Having had a quick look at your voltages, I do agree that the 3080's voltages are odd but not that odd. But Q2's are very odd. The Emitter and Base appear low.
Take the 3080 out and retest the voltages. The Base should come up to about 2.9V..If it doesn't have a look at the 56K, 27K and 1uF cap the supply the Vb rail (Tonepad circuit ). Tonepad boards are very nicely designed but it is easy to accidentally short adjacent pads with a bridge of solder....and the bridge looks as if it meant to be there. Check your board very carefully.
If you can upload some pictures (as clear as possible ) of both sides of the board it would be very helpful.

Reversing none working chips  .....I cannot recommend that as a diagnostic procedure. Please don't do that. You may have a fault elsewhere that is stopping the chip from working correctly but having now having reversed the chip all you now know is that the chip is now very likely to be damaged - this is not progress  .
There are a lot of fake 3080s out there..I'm looking at about 10 right now and I'm suspicious of the other 6. There are some quick checks that you can do with a multimeter that can help determine if a chip is likely to be a real 3080..I'll have to look up my notes and post the procedure here.

Anyway, good luck and again, welcome 

[italianguy63]

There are some quick checks that you can do with a multimeter that can help determine if a chip is likely to be a real 3080..I'll have to look up my notes and post the procedure here.

OH!  Please do!  That would be awesome to have..

I have struggled with probably 100's of fake ones...

So many, I actually gave up with builds that have CA3080's... I have only a handful now that I know to be good.

MC

[Slowpoke101]

OH!  Please do!  That would be awesome to have..

I have struggled with probably 100's of fake ones...

So many, I actually gave up with builds that have CA3080's... I have only a handful now that I know to be good.

MC

No problem. I should be able to find them in the morning ( Its nearly 8:00PM here).
Keep an eye on this space....

[pork ruckus]

Hey thanks for the quick replies, it's nice to see there are people out there willing to help a stranger out!

In reply to italianguy63: It's a DIP style, I don't like the sound of 100s of faulty chips yikes! 

In reply to Slowpoke101: I proper set of testing instructions for this chip would be fantastic! I knew it wasn't kosher to flip the chip around but reading about faulty chips I thought the pin out could have been reversed, I'll definitely take your advice and not do that anymore. I took the chip out and got the following readings on Q2:
C: 8.17V
B: 1.81V
E: 1.25V

I also took the resistors and capacitor you mentioned out of the board to test and they all seem fine. Could the trimpot setting affect the voltage at Q2? Should I check components at all the Vb points as a starting point from here? I'll do a thorough check over the board again tomorrow, in the meantime photos! I hope these are clear enough, they looked alright on my phone but a little fuzzy on the computer screen. 

Thanks again!

[Slowpoke101]

I've had a quick look at your pictures and they are a bit too low in resolution - any way to change your phone's settings for a better resolution?
One thing that I do have to verify with you - the pictures show that the board is reversed from the Tonepad layout (mirrored ). Is the board actually like this or has the image been flipped?

[italianguy63]

Yeah-- I used to build Ross (MXR) compressors... and I liked the CA3080 in the CAN style for the "mojo" value.

It got to the point that I could not find any (copies) that were in spec, or even work anymore.  I was still able to find working DIP style without too much issue.  But, since, I have moved on and build a different style compressor now.

I have not tried buying and DIP 3080's recently (a couple years), so I don't know what the market climate is now.

MC

[duck_arse]

+1 hello welcome.
+1 "never do this" - rely on your datasheets, all else is wrong.
+1 bigger pictures, but -
+1 your board is mirrored left-right.

[pork ruckus]

I've had a quick look at your pictures and they are a bit too low in resolution - any way to change your phone's settings for a better resolution?
One thing that I do have to verify with you - the pictures show that the board is reversed from the Tonepad layout (mirrored ). Is the board actually like this or has the image been flipped?

Oh God it's definitely reversed.. 
I could put in the l/r oriented transistors in the opposite orientation and run wire from the board in a cross over fashion to connect to the correct pin out of the IC right?
Is this an OK idea considering the circumstances or should I start over with a new pcb?
Here's another attempt at photos though I imagine it's moot now.

[Slowpoke101]

The pictures are much better now. Very good detail.
The reversed pattern board does give rise to some difficulties as you have noted, particularly the orientation of the transistors and the IC. The transistors are not really an issue as you have noted. You simply rotate them 180 degrees (left to right). However the IC is a bit different. Yes, you could isolate its PCB pads and run some jumper wires to reconnect correctly..Or, just remove the IC socket and resolder it to the copper side of the board and then just plug the IC into the socket.
However, I do recommend making another board if it is not too much of a problem. That way your board will then match the available documentation.

[pork ruckus]

The pictures are much better now. Very good detail.
The reversed pattern board does give rise to some difficulties as you have noted, particularly the orientation of the transistors and the IC. The transistors are not really an issue as you have noted. You simply rotate them 180 degrees (left to right). However the IC is a bit different. Yes, you could isolate its PCB pads and run some jumper wires to reconnect correctly..Or, just remove the IC socket and resolder it to the copper side of the board and then just plug the IC into the socket.
However, I do recommend making another board if it is not too much of a problem. That way your board will then match the available documentation.

I ended up flipping around the transistors and soldering the IC socket on the copper side and IT WORKS!! Definitely won't be making that mistake again.

Also not to push but, I don't suppose you managed to find the info for testing CA3080s did you?

Here's a pic of the final product!

Thanks for all the help!

[bluebunny]

Also not to push but, I don't suppose you managed to find the info for testing CA3080s did you?

Use a known working Dynacomp/Ross compressor as a test bed.   

And buy known good CA3080s.  Smallbear (US) and Thonk (UK) sell the Rochester reissues, so no need to throw money away at eBay fakes.

[Slowpoke101]

That is good news that you got it to work  Well done and it looks good too.

I fully agree with bluebunny's suggestion of having one of these compressors on hand as to be able to test any 3080s that you may have now or later in the future. I have a compressor board just for this purpose and it does come in handy when I really need to check a 3080.

The multimeter test is simply to sort out what chips MAY be genuine and what may be fakes.
Observe static discharge precautions. There is no point in damaging an IC that may be a genuine device.

The first step is to look at the IC and see if it has a notch at one end (helps to identify where pin 1 should be). If there is no notch put the IC aside and check what other ICs you have. At the moment there are quite a lot of fake 3080s that do not have any notches.

Next step is to set your multimeter to diode test and test across pins 1 and 8 (red to pin 1, black to pin 8 ) of the ICs. There should be no reading at all. Now reverse your leads and check across pins 1 and 8 (black to pin 1, red to pin 8 ), again there should be no reading. Next, leave the meter's black lead on pin 1 and check each of the other pins (2,3, 4, 5, 6 & 7) with the red lead - no reading should show. Now place the meter's red lead on pin 1 and  check each of the other pins (2,3, 4, 5, 6 & 7) with the black lead - no reading should show.

Any ICs that do show any readings put them aside in the reject pile. A genuine 3080 has no physical connection to pins 1 and 8 at all. In fact there are no connections points on the chip die for connection to pins 1 and 8 of the lead frame.

This is a very quick and simple way to sort what MAY be a 3080 and what is not. Other ICs can test the same way so be aware of this. Hopefully you now have some ICs that may be OK but are now worth testing in a working unit.

There is another test which I do not recommend as it is possible to damage a working 3080 and it depends on how much voltage your meter applies to the test leads when in high resistance test - the lower the voltage the better. The internal construction of the 3080 is quite simple but it does have a very interesting circuit layout regarding pins 4 and 5. If you place the meter's read lead on pin 4 and the black to pin 5 (resistance check) you should measure about 4.6M, reversing the leads the reading should now be about 5.68M If you do a diode test (again, not recommended), red to pin 4 and black to pin 5 a reading of about 0.66 should be show. Reverse the leads and you should see about 0.7

Hopefully this helps to sort out some of the fakes out there. Also bear in mind that even though an IC passes all tests and appears to work, it may have been obtained from reject stock - so it may have noise or reliability issues that will give you grief later.When you consider the amount of time and frustration that you may invest in dealing with fakes it is a far better investment to buy from reputable suppliers. Bluebunny noted a couple of very good suppliers and I recommend them myself.

[italianguy63]

Thanks for that Slowpoke!

When I started finding "fakes" the pedal would be NOISY AF.  Basically sounding very over-driven. 

Others would have no or low output....

I got tired of screwing with them.  Again, I did not experience much trouble with the Chinese "DIP" style... only the CANs.

MC

[duck_arse]

I ended up flipping around the transistors and soldering the IC socket on the copper side and IT WORKS!! Definitely won't be making that mistake again.

buy a lottery ticket - this doesn't happen often. did any of the IC's get fried, do they all work, did you chance to test them? good work, nice box.

If there is no notch put the IC aside and check what other ICs you have. At the moment there are quite a lot of fake 3080s that do not have any notches.

Slowpoke - can you elaborate your why's on this point please?

[Slowpoke101]

If there is no notch put the IC aside and check what other ICs you have. At the moment there are quite a lot of fake 3080s that do not have any notches.

Slowpoke - can you elaborate your why's on this point please?

Good question. My point is based on experience over the past year or so with some eBay suppliers. I have found some suppliers that are pretty good at supplying genuine ICs but recently have had issues. My last foray into the eBay abyss resulted in 3080s which were supplied in packages that had no notch nor indent to identify pin 1. Also the printing quality of the chip label was highly questionable. Other people have also noted these oddities. Needless to say none of these chips worked as 3080s.

[duck_arse]

so not "notch" in the strictest sense, but pin 1 identifier mark, in general, notch or dot or dimple. or "1", like RCA used to put.

[pork ruckus]

I ended up flipping around the transistors and soldering the IC socket on the copper side and IT WORKS!! Definitely won't be making that mistake again.

buy a lottery ticket - this doesn't happen often. did any of the IC's get fried, do they all work, did you chance to test them? good work, nice box.

If there is no notch put the IC aside and check what other ICs you have. At the moment there are quite a lot of fake 3080s that do not have any notches.

Slowpoke - can you elaborate your why's on this point please?

From the precarious way I initially treated these chips I figured any harm I had done to them had already been tenfold to plugging them in with the proper orientation so I just crossed my fingers and adjusted the trimpot; it was definitely compressing and sounded very clean. I was definitely lucky this time around.

[garcho]

^ For future builds, find the equivalent versions that use a more "modern" OTA like the LM13700. Don't buy more CA3080s and don't try to build more circuits utilizing those, it's not worth it. Outside of the 3080 being all around noisier, there is such little sonic difference (that's the most diplomatic way i could put it) between 3080 designs and 13700 designs that you will never miss the mojo and in the same amount of time and money you spend hunting for parts, testing, debugging, etc. you could be playing guitar with all your new LM13700 pedals. Same goes for OTA filter/mod effects