[BUILD] SSL 9K2

[gemini86]

In October 2011, a friend of mine and local studio owner had an interest in building some 9K pres and we went about pricing out the build, and then I remembered that he has an API 200 rack with only 4 spaces being used, so I offered to design some PCBS for him to fill the thing up and save building another case.

This meant fitting everything on-board, and adding the little go-between extras needed to complete the preamp. Some forum members expressed interest, so here we are, the SSL 9K2 for mass consumption:

Features input pad, phase and phantom on board, balanced output using THAT1646 IC. Runs on +/-16V

The BOM is as follows:

          +48   1 "SW1"
          100K   6 "R70, R78, R106, R114, R120, R123"
          100n   2 "C1, C45,C2, C3" ((2.5MM lead spacing))
          100p   4 "C56, C61, C70, C86"
          100R   4 "R5, R57, R72, R108"
            10K   1 "R112"
            10p   1 "C48"
        10u/25   1 "C72"
          120R   1 "R8" ((for 20dB pad-shunt))
          15uH   2 "L1, L2" ((the little ones that look like resistors))
      THAT1646   1 "U1"
            1K   4 "R73, R88, R109, R115"
          200R   2 "R75, R110"
        220u/25   3 "C34, C62, C92"
          22uF   2 "C4, C5" ((non-polar/bipolar electrolytics for 1646 sense circuit))
            27R   1 "Ra1"
      2K5 R/LOG   1 "VR3" ((from audiomaintenance.com))
        2U2/63   4 "C55, C68, C69, C84" ((10 or 15MM lead spacing))
          300R   2 "R71, R107"
          330p   4 "C57, C58, C71, C82"
            39R   1 "Rb1"
            3K9   2 "R69, R105"
            3p3   4 "C46, C47, C73, C74"
          470k   3 "R82, R118, R119"
          470n   6 "C38, C39, C64, C65, C93, C94" ((5MM lead spacing))
          470p   2 "C59, C83"
          470R   2 "R6, R7" ((for 20dB pad-series legs))
            47R   2 "R76, R77"
          4K22   1 "R113"
            50K   2 "R81, R117"         
            6K8   6 "R83, R84, R121, R122"
            6K81   2 "R1, R2" ((precision matched for phantom power))
        MAT-02   1 "Q5"  ((from forum member Electrochronic))
        NE5534N   5 "U22, U23, U24, U29, U30"
            PAD   1 "SW2"
            PH   1 "SW3"
          TL072   1 "U31"

All switches are ALPS SPUJ series DPDT or equivalent. I added another 100nF cap to the underside of the PCB directly under the PCB pins of each IC, but that is likely not needed. This will be added to future revs if the need arises.

 

[gemini86]

Another note:

To anyone building these for an actual API 200 rack (as I had intended originally), The resistors along the top are a bit too close to the edge, and I didn't realize the plastic guide rails in the rack are so deep. They fit fine, but aren't as smooth as I would like.

Just trim the leads flush before soldering, or file them down after. The board will drag on the bottom edge because it's only tight on the top rail. Irritating, I know.

 

[envelope]

Another note:
To anyone building these for an actual API 200 rack (as I had intended originally), The resistors along the top are a bit too close to the edge, and I didn't realize the plastic guide rails in the rack are so deep. They fit fine, but aren't as smooth as I would like.
Just trim the leads flush before soldering, or file them down after. The board will drag on the bottom edge because it's only tight on the top rail. Irritating, I know.

Hi Rodney, thanks for getting this out for the 'masses' I've PM'd with contact and PP info. thx

Will there be any resistor leads or other leads/pads, that are dissipating heat, in contact with the plastic rails?  I've been looking at my API 200 rack, and it seems that I can shave a part of the rail away, if resistor or other heat bearing leads/pads are in contact. If no heat, no prob.s

 

[gemini86]

I had it running all night in my rack, nothing got warm.

 

[gemini86]

Updated to revision 2.0

The documents have been altered with the needed trace edits, some slight nudging of components, and the addition of some PSU bypass caps(CX1,2,3,4) that are NOT on the first run of boards... The only one that seems to be needed according to my testing is the one under the tl072, I just soldered 100nF directly to pins 4 and 8.

All the rev1 boards that are sent out have the trace cuts already done, and jumpers marked. So it's still just paint by numbers.

 

[gemini86]

Links to the updated rev2 Layout is pretty much the same, but with corrected trace cuts at 2 ICs and at the gain pot, added bypassing at every IC, and components nudged away from the edges.

https://docs.google.com/open?id=0B6s_klsk9wMDNjNjYmFkOGEtOGVkOC00ZmRhLTlhNTctZWY3ZTA3ODBmNTMw
https://docs.google.com/open?id=0B6s_klsk9wMDM2Q1N2NlNGItYzY3Ni00Mzc2LWE5ZTQtODY1ZmQwZGM0YTUw
https://docs.google.com/open?id=0B6s_klsk9wMDNjk3MDNkNmUtN2QwOC00YWJkLTkzNDEtODg0ZGJkOTIxMGE5

 

[gemini86]

That part looks good to me.

 

[gemini86]

Okay guys, one last trace edit that needs to be done that I misses...

It;s right next to the pot by the others, see pic...

The yellow is the trace that needs to be cut, the blue is where to jumper...

Sorry to be such a pain in the arse... :S

https://lh6.googleusercontent.com/-RaKGN6PL-sU/Tz4ZpvQBUYI/AAAAAAAAAag/KEgD7zjEpvg/s800/2012-02-17_01-02-27_312.jpg

 

[gemini86]

Strange thing is, it did work... but I had missed that error, somehow.

Just cut the trace highlighted in yellow, jumper where it's marked in blue.

 

[r2d2]

very nice unit !
well done man !

peace
r2d2

 

[3nity]

Testing my first channel...
All i have is a high pitch sound....not sure where to look it all seems good...
However i havent performed the pot trace yet.

Ill do that right now!

 

[gemini86]

What are you using for DC servo? I had the same issue when using tl052.

 

[3nity]

Im using TL072.
Should i use something else?
Thanks

 

[gemini86]

No i'm currently using 072 with good results.

 

[3nity]

Still nothing.
I know whats giving me this problem is the TL072 because everytime imove the board and my hands get closer the IC31 noise disapear.
Pot trace was cutted and jumpered and still same issue...no sound at all.
Thanks

 

[gemini86]

Just to make sure, did you do the newest trace edit that I posted here?  The one that is not marked on the pcb. I'm at work now so I'll jump online to try to help later tonight.

 

[gemini86]

First thing first, let's get some static voltages from each IC. Then, probe to see where the signal gets lost.

Channel one oscillation; try adding a 100nF bypass from v- to v+ on the bottom of the PCB, like so:

Sorry if I can't respond instantly to help. I am but one man, but the forum is many.

 

[gemini86]

Here's a couple pics of blank pcbs.


FRONT
BACK

 

[gemini86]

If I had a nickle for every time I did something like that...well, you know.

Progress! Keep at it, you'll get it worked out! It's a complicated circuit, one I needed help debugging on my first build, so just give your brain a rest after staring at it too long.

Yes the pot mod is needed. As the circuit board is routed right now, the second stage gain pot is backward, meaning the gain will rise slowly until almost full rotation. This will result in a very low gain setting until the very end, making gain settings not very user friendly. Also, one of the trace edits is need to correct a feedback path on one of the IC's.

 

[3nity]

Mike do you have a front panel layout for this, you can share?
Thanks.