bxt403
Well-known member
Gustav said:
It turned out to be the adapter cable I was using (mini-plug from my Macbook Pro 1/8" audio out to stereo XLR): I had one for input to the GSSL and one for output to my Objective 2 headphone amplifier).
The output cable was working fine (and was wired correctly with a 150 Ohm resistor connected between pins 2 & 3). The input cable going into the GSSL was wired in reverse (in that it was wired for an XLR-M but had pins 1 & 3 connected on an XLR-F connector to an XLR-M to M barrel adapter). I didn't notice the barrel connector compensates for the pin reversal until I replaced the XLR-F with a proper XLR-M.
Prior to testing with a properly wired test cable, I changed the Signetics NE5534As and NE5532As that were installed instead of the stock T.I. NE553xPs- with Signetics NE553xNs; along with changing the TL072 from a new one to an older one I have.
Also- I noticed that a Rev. 7 or Rev. 8 board seems to need 127 K Ohm resistor installed where it's labelled as 120K* on Rev. 9 or later (as it's listed on Gyraf GSSL page):
Knowing 120K is for the THAT 2181, according to the table (at http://diy.fischerworks.com/gssl_vca.shtml). My research (thus far) indicates (verifying from the PCB photos at audiokitchen) is that the 100K Ohm value (and DBX 2150) is at least up to the Rev. 7 PCB and 127K Ohm is for the THAT 2181 & 2180 (from the picture from the GSSL page on Gyraf's website but is contradictory to the Gyraf Rev. 9 schematic & "Matt's DIY Pages").
The table at "Matt's DIY Page" indicates the proper value is 120 KOhm (and the 127 K Ohm value listed on the Gyraf page implies that 120 K Ohm value listed for the 2181 on the Rev.9 (along with the Rev. 11 & Rev. 12) schematic also applies to the 2180 (but it isn't mentioned on the schematic). It's safest to assume that the 127K Ohm value might need to be double checked; given I'd rather not waste any more time trying to figure out if the 127K Ohms is correct or not on Rev. 8 or earlier versions of the GSSL: where the DBX 2150 is labelled at the 3 VCA slots that are labelled as THAT 2181 on the Rev. 11 boards I have (that's consistent with the Rev. 11/12 schematic that's posted at PCB Grinder at the time of this post).
Furthermore- it seems like you're better off finding an older Rev. circuit board (Rev. 8 or earlier) if you want to implement the DBX 2150; conversely- use a Rev.9 (or later) board if you want to implement the THAT 2181 or 2180 (knowing that the silkscreening on the Rev. 11 boards I was provided were labelled consistently to accommodate the THAT 2180- despite having "THAT 2181" silkscreened).
Also- it looks like Rev. 11 incorporates the CRC & HP filter so it's a safe assumption that adding an external CRC filter (like the one from EXPAT Audio) is redundant (and a waste of space, time and money for any people who are using the Rev. 11 or later circuit board). The compression seems to be more noticabe with the 553xP/N that the 553xAP/AN installed at the input and output but I've only used Signetics (N & AN) ICs (since they're true to the SSL design) so am not sure of how the newer Texas Instruments P & AP versions of the 553x op amps sound (w.r.t. to the original SSL circuit).
Finally: I'm thinking the SSL X-Logic Bus Compressor (or any current SSL Bus Compressor variation) is going to sound closer to Duende Native (in that you're going to hear the authentic SSL sound- albeit Duende's Bus Compressor is based around the 4000 G Series) while I'm inferring the GSSL is an optimized and/or improved version of the 4000 E. I can only compare my build to SSL's Duende Native's Bus Compressor (and my GSSL doesn't sound as "SSL-y" but suspect it's because Duende is meant to sound more like an optimized version of the 4000 G).
