GSSL: SC resistors with turbo mod

Hi,

I'm posting here because I found an old topic (here: http://www.groupdiy.com/index.php?topic=29386.0)  about installing the turbo on GSSL and, if previously removed, stuffing again the 47R and 10K resistors on the main board, without removing the corrisponding resistors on the Turbo board.

In My build, since I've used THAT 2180, I've not stuffed those 10k and 47R near the sidechain VCA and changed that 3k9 resistor to 4k7 (I also had a jumper linking the unstuffed holes of the 10k resistor, but I removed it as a guy told me it was wrong). For simmetry I did the same on my turbo board, removing 47R and 10k and changing 3k9 to 4k7, as someone told me so.

Now the compressor has a strange behavior, it seems like it compressess too much and randomly it totally reduces the incoming signal, with the meter suddenly jumping to the maximum GR level.

should I just put back those resistors on the main or turbo board?

This problem happened since I changed the 3k9 to 4k7 on the turbo board: before that (having only removed the 47R and 10k) it seemed as it was acting right in normal mode, but drammatically reducing the GR when put in turbo. I did some acurrate tests and seen that the right channel (in solo), was acting good, giving the same GR value between normal and turbo, while the left (when in solo) did reduce the GR in normal mode. 
That's why I tried to change 3k9 to 4k7 like in the main board.

What should I do now: put again the Sidehain VCA's 10K and 47R resistors on both the main and turbo boards, changing back from 4K7 to 3K9 on the turbo? Or what else would you suggest? 

I hope you can help, Thank you!

ok, I finally tried putting back those resistors on both turbo and main boards... and now everything seems working!  ;D

I can't explain why though..

that strange overcompressing behavior was caused by a miscontact.. but I leaved the 4k7 instead of 3k9 on the turbo board, and it works fine.

the only problem is that my right channel is compressing slightly more then the left (about 1 db), and I don't know why...
... and I have a drastic reduction of gain when putting in bypass (the same I have with makeup to minimum) and don't know if it's normal or caused by that 15k to 27k mod for unity gain,

some help?

angelo2979 said:

ok, I finally tried putting back those resistors on both turbo and main boards... and now everything seems working!  ;D

I can't explain why though..

that strange overcompressing behavior was caused by a miscontact.. but I leaved the 4k7 instead of 3k9 on the turbo board, and it works fine.

Click to expand...

3k9 would still be the better value for the recommended 2.4mA bias current of the THAT2180 VCAs.
(4* -0.65V diode forward drop - -12V)/0.0024A=3916 ohms
The 4k7 would be the choice when using the NEC uPC1252-H2 for its recommended 2mA Iset.
(4* -0.65V diode forward drop - -12V)/0.002A=4700 ohms

IMHO this "My builds use both 2180LB and dbx C1252H2 VCAs" was one of the reasons for misbehaviour from your linked thread. This NEC type VCA, built in license for DBX noise reduction units, was closer related to DBX2150 VCAs and these, contrary to the recommended pin4 left floating THAT2180, require pin4 connected. There are more differences between these chips, IE max.i/o-current, control port law -5.9mV/dB vs. -6.1mV/dB, ... For usual there is a reason different chips come with a different label/number.

the only problem is that my right channel is compressing slightly more then the left (about 1 db), and I don't know why...

Click to expand...

most likely resistor parts tolerances around the 'THAT2180 to DBX202 emulation circuit', especially at the voltage divider 1k/120R, connecting to non-inverting input of control voltage buffers NE5534-pin3.

... and I have a drastic reduction of gain when putting in bypass (the same I have with makeup to minimum) and don't know if it's normal or caused by that 15k to 27k mod for unity gain,

Click to expand...

I stopped counting my answers to this subject in the official GSSL help thread. Wild guessing, all of your gssl related questions have been answered in this thread numerous times.
Good luck

Thank you for your detailed answer!

just to make it clearer:

would you recommend using 3k9 instead of 4k9 with THAT2180 VCAs?

IMHO this "My builds use both 2180LB and dbx C1252H2 VCAs" was one of the reasons for misbehaviour from your linked thread.

Click to expand...

Yeah, I thought the same, but I've put back those resistors and it's working... do you suggest removing them?

most likely resistor parts tolerances around the 'THAT2180 to DBX202 emulation circuit', especially at the voltage divider 1k/120R, connecting to non-inverting input of control voltage buffers NE5534-pin3.

Click to expand...

how would you fix this problem?

about the bypass..I did a search through  the forum before writing here and didn't find any solutions, but I'll search again and more accurately in the GSSL thread for my specific problem...

Thank you again!

I found that my bypass problem was caused by the use of unbalanced cables on the output..eheh..  :

but still have the left channel higher because the right channel is compressing slightly more..I would like to fix this..

angelo2979 said:

just to make it clearer:

would you recommend using 3k9 instead of 4k9 with THAT2180 VCAs?

Click to expand...

Dunno what's so unclear with '3k9 would still be the better value for the recommended 2.4mA bias current of the THAT2180 VCAs.'
With THAT2180 VCAs  3K9 for the sidechain (between pin-5 and -12V supply) and 5K1 for the audio-VCAs (between pin-5 and -15V supply).

> IMHO this "My builds use both 2180LB and dbx C1252H2 VCAs" was one of the reasons for misbehaviour from your linked thread.
< Yeah, I thought the same, but I've put back those resistors and it's working... do you suggest removing them?

Click to expand...

These resistors were ment for a different chip to help noise modulation. Not needed/contraproductive for THAT218x VCAs.

> most likely resistor parts tolerances around the 'THAT2180 to DBX202 emulation circuit', especially at the voltage divider 1k/120R, connecting to non-inverting input of control voltage buffers NE5534-pin3.
< how would you fix this problem?

Click to expand...

Easiest spot for a fix seems to be the lowest 1K resistor right side next to the 'RIGHT VCA' outline. Substitute with a 2K trimmer/rheostat, initial value about centered. Adjust for closest R/L compression match with a constant signal feed that causes compression.

ok, perfect now, you couldn't be clearer  ;D
thank you for your patience and support, I hope this could help someone else!