Fault in SSL compressor design

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

bpatural

Member
Joined
Jul 3, 2005
Messages
12
Location
hollywood
Hello everyone,

I just wanted to say that me and my friend have recently both built SSL compressors from the Gyraf design and have used this user name together so we are actually representing 2 people. In testing our projects it seems that we have come across what appears to be an apparant design flaw in this DIY project. Mind you it is not something devastating but seems to be a big enough deal to look into. Let me first say that we are not people with anything more than remedial electronics knowledge. With that being said I encourage fellow builders to try this test.

First off, this is only apparent when the bypass jumper is OUT!!!!! Having makeup gain active while the compressor is in bypass just seems useless and that's why this seems like such a big deal. With the jumper out on the bypass switch the compressor bypasses the makeup gain while the compressor is out. First put a 1 Khz sine wave into the compressor, I recommend -18 but any level will work. Make sure the switch is in bypass mode. Put the threshold all the way up to start. YOu should hear a perfect sine wave if your compressor is working properly. Start to sweep the threshold down while still in bypass mode. What we would hope to hear would be no signal changed but low and behold the more we get to the middle of the threshold, you should start to hear a square wave being added into the signal. I f your ratio is set to 10 to 1 it will be most apparant. In fact, half way through the threshold control you will almost hear a perfect square wave as harmonics are being leaked form the sidechain into the audio signal around 60 db down. This only happens if the bypass jumper is removed but seems to apparant on All units. So much for our bypass!!!!!!! As you move the threshold farther from the middle the distortion starts to go away and is eventually eliminated as it is turned all the way down. THIS ONLY HAPPENS WHILE THE UNIT IS BYPASSED! It doesn't happen when the compressor is engaged. Big deal then you say except if I'm gonna throw program materila through this compressor it better be freakin working. Well it is working, and yet this is a consistant artifact with the design. The guys that built hard bypasses are in heaven right now as they won't have this problem. I personally would like to know how to fix it on the unit without having the makeup gain engaged and without hulding a hard bypass which seems like the only real option right now. I have very little electronics knowledge but know that this would be unacceptable on any piece of pro audio gear. THE SIDECHAIN IS LEAKING INTO THE AUDIO PATH!!!! Let's figure this out! DO the experiment, so far it seems like fact. Remember, NO JUMPER AND IN BYPASS MODE. TURN THE THRESHOLD CONTROL. WHat the !#$%!#$% is happening??? Other than this problem this is the most useful compressor in the world.

Peace!!
 
Yes, I do hear this as well but only at the 10:1 setting an in the ccw area of the threshold knob though.

I guess hard relay bypass is the only proper way to go then right?....
 
Reverse the switch on "E" input and connect them ("E" line) to the ground in the bypass position.

Regards,
Milan
 
[quote author="moamps"]Reverse the switch on "E" input and connect them ("E" line) to the ground in the bypass position.

Regards,
Milan[/quote]

Elaborate please
 
Hi,

620k resistor on 72's input must be grounded in the bypass mode. (It's high impedance input).

Regards,
Milan
 
moamp, it seems to me that you are super experienced in this matter and can solve it but please give the beginners a dumbed down real world answer on how to solve the problem. Thanks,
 
OK. I figured out that my bypass pot was a little different than the schematic. I have done the mod you described and it is excellent except at the very low end of the threshold there is still a little bleed around 70 db down. Anyone else tried it?
 
Hey Brian,

I don't want to go OTT, but just wanted to say I liked your site, nice name for a studio, mine here in SA is called 'The Orange Room'... freaky....

I've yet to test out this bypass issue on my SSL clone, but will try check it out sometime, though I did suffer from the 'Threshold' issue some others did, I.E the Threshold when the comp is active is kicking in way too early, I did the suggested Mod, not sure if it is an issue for you but you can check out this thread: http://www.groupdiy.com/index.php?topic=2418

Cheers

Matt
 
[quote author="bpatural"].... a little bleed around 70 db down......[/quote]

further mod
(bypassing the threshold pot in the bypass mode ;)):

1. disconnect threshold pot's wiper from "F" point on the control pcb
2. connect wiper terminal to "B2" terminal on the bypass switch
3. connect "F" point on control pcb to "B" terminal on the bypass switch
4. place a jumper between "B1" and "C1" terminals on the bypass switch

(I suppose that the bypass switch is "brake before make" type.)

Regards,
Milan
 
Back
Top