Modifications on my GSSL

  • Thread starter reanimatorstudio
  • Start date
GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
feedback mode for sidechain? please elaborate! sounds very interesting.

As for opamp upgrades, I am using OPA2604 and OPA604 for the input and output. It's a definite improvement. I preferred them over OPA132 and OPA2132 (and definitely over any NE553X type). Then, if you want to alter the actual compressor action, use OPA404 and OPA2604 in place of the TL074 and TL072. It will be more of a mastering unit this way, and have less of the famous SSL "vibe", but has much more stable compression (more stable image, less distorted attack) on very high ratios.

Other upgrades to consider:

Fix the PSU. Skip the -+15VDC area of the PCB completely and use something higher quality. You will have less hum and more stable rails. Something like this instead: http://www.geocities.com/m_natsume/files/MNATS_PSU.pdf You could even run the unit at -+17V for minor sonic improvement (mostly impacts the VCAs).

minor details that have surprisingly big impact:

GSSL input and output are filled with 100nF bypass caps for the electrolytics. They actually degrade audio signal quality, especially if you have even halfway decent electrolytics. They should be removed.

In the schematic there is a 100pF stabilisation cap on the VCA opamp buffer. Keep it this way. In some places it is suggested to use 100nF here, but it makes the compressor a bit sloppy due to slight integration of the VCA control signal. 100pF will keep the VCA attack intact.
 
Kingston said:
GSSL input and output are filled with 100nF bypass caps for the electrolytics. They actually degrade audio signal quality, especially if you have even halfway decent electrolytics. They should be removed.

Interesting, do you think the Panasonic fc series enter ''the halfway decent'' category?

Thanks

Hubert.
 
Yes of course they are.

With modern manufacturing methods, any known brand is decent. They're all cheap already (say, panasonic, elna, nichicon) so it's pointless to start gambling with the chinese stuff.
 
Talking about mods ;D...mods that i have done on mine:

-Turbo mod, never use the normal mode since then.

-Hi-pass on the sidechain vca's, 60hz and 120hz done with two different pair of caps just before the sidechain vca's. Sometime,, it's just what it need to make the ssl very transparent.

-No more 1.2s release time, never used it, replaced it with a ultra fast release. So instead of 180k,270k,560k,1.2m,auto on the release switch i now have 82k,180k,270k,560k,auto. nice when you want to use the compressor to really squash things.

Very interested on your feedback mod. Does it really change the caracter of the comp?

Hubert



 
hi, i'm also very interested in the feedback in the sidechain  :)

mod done on mines :
- turbo card
- super sidechain board W T,H*R-U,S*T
- a swith w/ a resistor to get two different treshold range
- power supply enhancement (a filering board between diode bridge & 1000uF caps that i also increased to 2000uF)
all power supply mounted on a separated board > a must to do

i'm thinking about replacing the resistor "web" for attack & release by pots for the one i use for live PA
any advices on that ?

& i'm thinking about adding a switch to change the well known 100K/127K resistor near the TLO74 on the sidechain to get a different set of ratio... but maybe that's not the better place to do that.

i've got a board from livingnote to do the clean/comp blend just got to do it...





 
nice one.. will try the opa's and poly caps on mine.

sounds strange but what is really cool to have is a lo pass filter, yes.. low pass.. so that only bass comes through in the sidechain signal.
In combination with the wet/dry it's really cool, you can make it pump but with subtle dry signal mixed in it's great!
I have a 150 and 250hz (approx) setting on mine.
 
great idea (low pass filter) + mix'n'blend  ;D

can you send any more info on that (schemo / photo) ?
it could be cool if we can implement that on our existing SuperSC boards.
 
How about bipolar electrolytics in signal path?
As far I understand they are much more long living creatures in that position  ???

-Paavo
 
reanimatorstudio said:
Maybe a new GSSL pcb with all the mods like the turbo mod, sidechain mod, ... would be cool

At the moment I'm working something I call SuperGSSL. My plan is to use two separate pcb's. One for audio stuff and another for sidechain.
Right now I'm working with pcb design for audioboard. There's integrated basic GSSL circuit for NE5534/2180 vca emulation, crush'n'blend and PSUs based LM317/337 regs.
In sidechainboard I want to integrate turbo mod, couple of famous filters and tuneable HPF. Tuneable LPF would be nice add more. And changeable slopes. And feedback in sidechain. And...and...and...and...and...and...

Schematic for audioboard http://www.5by5.fi/diy/supergssl/SuperGSSL_audioboard_schematic.pdf
It's not absolutly perfect yet propably not even close. I need some time for this due I'm newbie diyer and don't have too much knowledge but with this project I've learned and I will a lot :)
I hope I have working proto of audioboard and schematic for sidechain board during few next weeks.

edit: There is in fact three pcb's. Third one is relay-bypass and in-out connectors  :D

-Paavo
 
reanimatorstudio said:
Quote from: Kingston on September 01, 2009, 05:47:48 pm
GSSL input and output are filled with 100nF bypass caps for the electrolytics. They actually degrade audio signal quality, especially if you have even halfway decent electrolytics. They should be removed.

Can you tell me how they degrade the signal, is there a theoretical reason for?

LM317 & LM337 are better for noise issues, I don't have any humm problem. Good idea I'm goinig to try that!!

as for how they degrade the signal, they screw up transients slightly, which I did not expect. Some (not much) discussion on theory: http://www.groupdiy.com/index.php?topic=35066.0

Are you still using the original PSU on board? you might think you have no hum but did you measure it? There isn't stock GSSL board in the world capable of better than about -80dB (RMAA scale). That's was decent in the forties, but today you'd expect about -100-110dB, and that's possible with moving the PSU outboard and with LM317. You can still keep the +-12V regulation onboard.
 
I remember Matthias did a sort of "crush mod" on one of his builds. Don't know which one he eventually implemented, but...

here's one of the tests:
http://www.groupdiy.com/index.php?t...ll -mod and then wet-dry... just for drums.
 
HI,


  I have just done a turbo mod to one of mine. I MUST do the others ASAP! I have just scored some 202c's, so will be replacing 2151's in one of mine soon. For my trusty dbx2001 vca version, I am thinking of building a separate board for discrete op-amp input and output, still deciding what op-amps to use, and whether to use just one on the output and use a transformer to get the extra 6dB, or use two . . . I kinda like the idea of an all discrete G-SSL compressor! in the mean time I have got some spare LME49410 and LME!4562 to try. I already swapped NE5532 for OP275, and like it a lot. Very subtle difference, but a tad less harsh. I am going to try OPA404 and OPA2604 in the side-chain, but my instinct remains sceptical on this one! My mains transformers are always built into a separate box, so hum doesn't seem to bother much. I have not measured any levels. Anyhow, I am not known for quiet mixes . . . . .

 I have another completed board with dbx2001's, and will possibly try an "ultimate" version just for shits n' giggles . . .



 I am thinking about how to implement a "hold" function for the G-SSL, to overcome the big peak after a quiet passage when compressing hard, especially in auto mode. Any suggestions? Kieth has strongly recommended using reed relays or fets. I was thinking of using an opto-isolator, but Kieth has suggested that it's off value might cause some leakage.



  Kindest regards,


    ANdyP


  Kindest regards,


      ANdyP
 
on the thread that EEMO1 points, the feedforward mod is :

Unsolder the 100kΩ resistor right from the Threshold pot. It should have the same effect as before, but you will have the threshold pot working.

i assumed that's the one between 620K & 220K near the sidechain VCA on the PCB.

could someone confirm that ?


GSSL reanimator (thanks for the pic ;) ):

the sidechain feedback is only the 330K and a switch, or do we have to change something else ?


with my little theorical background, it seems that the two mods doesn't do the same thing...

a "theorycal" advice on that would help choosing which one to do...
 
EEMO1 said:
i'd like to build one with crush the hell -mod and then wet-dry... just for drums.

Fast attack, deep threshold and high ratio will have any SSL sounds like shit. All these variants of That and DBX VCA's will crap out (distort) before anything resembling good or even useful squash.

You need a completely different topology for that.
 
there is a thread from Luny Tune about this, with a drawing.
it's about output trafo only.

if I remember well you just need too change some resitors.

ahhh bingo, found it...
http://www.groupdiy.com/index.php?topic=17527

cheers.
Jr.
 
Yes, you have to implement the little schematic with the 4 resistors and 2 caps on the other half of the 2-polar switch too

sorry i was too quick...

thanks.

i'll build it.

do you have any advice on the other "crush" mod ?

what do lifting that 100K do ?
 
Back
Top