GSSL HELP THREAD!!!

[Dissler89]

Double check all +/-15VDC sup.....

I checked everything again, replaced all the 5534s, 5532s and the 7815.
The unit now passes audio and all the Rails seem to be fine, including the 15v+
However...

The unit is not compressing. The only thing that seems to function at the moment is the bypass Switch and the Make up Gain.

I looked around the forum and thanks to this had noticed that I was missing the Jumper at the HPF*.
I went ahead and added that jumper, but the unit is still not compressing.

After double checking and resoldering some of the switches the unit is now succesfully compressing!
Evreything seems like it is working as it is supposed to.

Thank you Harpo for your support!

 

[beatnik]

hi guys, what value for resistor marked % in the sidechain section if i'm using 2181 VCA?

also, better putting a trim instead of resistor 120K* for correcting ratio calibration?

 

[Harpo]

hi guys, what value for resistor marked % in the sidechain section if i'm using 2181 VCA?

What type of 2181 VCA?
For a THAT2181-A, THAT2181-B or THAT2181-C in the sidechain section probably none in the % position.
For whatever reason trimming THD you could inject a compensating current into pin4 from a high impedance source as shown and with parts values from datasheet for your specific THAT2181-x VCA, but there is no provision/footprint on pcb for this option.
For sidechain-VCA type DBX2150, NEC uPC1252HA2, THAT2150A, THAT2151 or THAT2155 use a 51R resistor in the %marked position.

also, better putting a trim instead of resistor 120K* for correcting ratio calibration?

Yupp. Maybe use a 82K or 91K fixed resistor in series to a 50K trimmer.

 

[salomonander]

did anyone ever breadboard a turbo board? if yes, where can i find info? it is sold out at expat and no one answered my mails. cheers

 

[beatnik]

hi guys, what value for resistor marked % in the sidechain section if i'm using 2181 VCA?

What type of 2181 VCA?
For a THAT2181-A, THAT2181-B or THAT2181-C in the sidechain section probably none in the % position.
For whatever reason trimming THD you could inject a compensating current into pin4 from a high impedance source as shown and with parts values from datasheet for your specific THAT2181-x VCA, but there is no provision/footprint on pcb for this option.
For sidechain-VCA type DBX2150, NEC uPC1252HA2, THAT2150A, THAT2151 or THAT2155 use a 51R resistor in the %marked position.

also, better putting a trim instead of resistor 120K* for correcting ratio calibration?

Yupp. Maybe use a 82K or 91K fixed resistor in series to a 50K trimmer.

thank you very much harpo

 

[Dissler89]

At the moment I am in the process of making my unit look more 'appealing'.
I was wondering how other people have managed to attach the 'control board' to the front plate along with their buttons without having the potentiometers/switches jiggle around.

Ideally I'd like to make all the buttons look like the Sidechain switch at the far end.

Any advice/opinion welcome.

 

[beatnik]

just cut the switch the right amount

 

[blackartmixing]

how are you guys testing your compressors?  i have a problem with mine (one channel is really distorted).  i found myself a really old oscilloscope (iwatsu ss-5711) but it seems to be working.  so i guess i need a test signal.  can i just run a 1k sine wave out of my desk into the input of the compressor and measure it somehow?  or do i need a dedicated signal generator?  i have NEVER done anything like this before so any help would be great.

 

[beatnik]

you can hook up a simple audio probe

http://www.diystompboxes.com/pedals/debug.html

then send a 0db sine from daw and monitor the signal path of the compressor until you find the problem

if your build is compressing and only has the distortion problem it will be easy to find

 

[Greg S.]

Question:

By using a a Sine 0dB input Signal, what are the max out ratings?

And what is the output impedance of the Gssl?
(might depend on the VCA, right?)

 

[Gustav]

how are you guys testing your compressors?  i have a problem with mine (one channel is really distorted).  i found myself a really old oscilloscope (iwatsu ss-5711) but it seems to be working.  so i guess i need a test signal.  can i just run a 1k sine wave out of my desk into the input of the compressor and measure it somehow?  or do i need a dedicated signal generator?  i have NEVER done anything like this before so any help would be great.

If only one channel is distorting and everything else is fine, the possible errors are narrowed down significantly.

I would start off by visually inspecting the soldering on the distorted channel path from input, through the VCA to the output section, then swapping the Opamps and VCAs to see if the problem follows those.

Gustav

 

[Harpo]

By using a a Sine 0dB input Signal, what are the max out ratings?

And what is the output impedance of the Gssl?
(might depend on the VCA, right?)

Whatever level your 0dB input signal might be (dBu, dBV, dBFS, ...), the max.output rating (assuming voltage and not exceeding the chips current drive ability) is dependant on supply voltage and type of last opamp stage. A NE5532 with +/-15VDC supply might give out typical +/-13Vpk max.into >=600R load as this type of opamp doesn't swing rail to rail. This 13Vpk is 9.19V rms or 19.27dBV for unbalanced connection to the following piece of gear. The inverting buffer doubles this up to 18.38V rms or 25.29dBV for balanced out connection. With decreasing input impedance of the following connected piece of gear (lowZ <600R historical gear), max.output level will decrease accordingly.
Output impedance is typ.100.3R from opamp + last series resistor, 200.6R for balanced out operation.
The type of VCA has nothing to do with output impedance.

 

[Orchid0]

Hi,
My Gssl is working and through sifting this thread I've been able to narrow down/fix about 90% of my problems/distortion/buss issues...
The last One Im having is that my make up gain nob is not responding to anything.

-Gssl Power Voltages checked all are correct according to PDF posted on this thread.
-Is passing audio just fine with some DB loss which I assume is acceptable..
- Unit compresses, Threshold works, attack + release are responsive, Bypass works, Ratio works (Though 10:1 is usually pretty overkill)
- Meter is working

Looking on the schematic I thought it could have been my bypass wiring. It says on the schematic E(ribbon cable) is the makeup gain. I am showing low mv values on that pin. However I do have 12V on the outside pin and 7v on the middle pin at the 50K Pot.

Im using a PCB grinder kit. My switch was wired according to his guide. I also can't find the 120* resistor around the makeup section on the schematic on my PCB. Also, Im doing a barebones GSSL with no side chain/turbo. Is there anything I potentially could have forgot to jumper or bypass? HPF* is jumped first hole to third, 12v 0 v NC are left bare - not jumpered. I also tried replacing my 50K pot as I read somebodies was faulty.

I feel like I've tried everything and I can't find my answer on here yet. Any suggestions?

 

[steelduck]

Hi. I have a problem with my GSSL that i don´t quite understand. It is built from PCBGrinder´s kit. No mods.
Left channel seems to work as intended, but on the right channel only the negative part of the audiowaves get compressed.
The positive peaks come through uncompressed. In bypass it passes audio. Same when i raise the threshold.
I really wan´t to understand the problem, and I read up on how VCA´s work, but I can´t think of anything, that would make
only the negative peaks compressed. I changed the VCA in the right channel with the known good one from the left channel, and the
problem was still on the right channel. Also tried changing the opamps. Voltages for both channels match each other within 1-2%.
Test points for +/-15V reads +14,84V and -14,85V
Anyone have an idea what could cause such behaviour??
Thanks!!

 

[gyraf]

only the negative peaks compressed. I changed the VCA in the right channel with the known good one from the left channel, and the

Maybe the supply power for the vca - resistor going to V- ?

 

[steelduck]

only the negative peaks compressed. I changed the VCA in the right channel with the known good one from the left channel, and the

Maybe the supply power for the vca - resistor going to V- ?

Hi, thank you for responding! That was my first thought, but the
voltages on all pins of the vca where identical to those on the working
channel.

 

[sineoff]

Hi all - my first post here  ;D

I have a few questions about (what I think is…) the RMS conversion part of the Gyraf G SSL schematic.

1) Can anyone confirm that the second and third op amps following the Sidechain VCA is actually the RMS conversion part of the schem? I'm asking because, to me, it just looks like two full wave rectifier/absolute value circuits - I can't see the log/multiplier part of the circuit I've seen in other RMS to DC converters that is normally done with transistors or a diode array.

2) Has anyone got a different/clearer schematic of the SW1/Ratio part with the dual R pairs? I can't quite work out what's going on there from the schem online at Gyraf
 
3) Assuming this is the RMS conversion - can anyone explain how this circuit is working?

Link to the schem FYI: http://www.gyraf.dk/gy_pd/ssl/ssl_sch.gif.

Thanks in advance (and apologies if this is in the wrong place).

 

[Harpo]

Left channel seems to work as intended, but on the right channel only the negative part of the audiowaves get compressed.
The positive peaks come through uncompressed. In bypass it passes audio. Same when i raise the threshold.

Check/replace the right side audio-VCA control port buffer NE5534 and/or double check surrounding parts of this opamp (2x1K, 1x120R, 2x100pF) ?

 

[gyraf]

Sineoff,

There is no RMS-converter in the stereo bus compressor: it's peak rectified, and works in feedback-loop compression.

You may confuse it with the channel compressor, which uses a RMS-converter, and is a feed-forward design.

Jakob E.

 

[darkcharly]

I finished the GSSL and want to thank Jakob for this beautiful project. I have a question: my VU meter (Hairball 8027-B 1mA Meter) has 2 white LEDs 5.6-7.6V DC. It could replace the 7812 (light and relay) for 7806 to power the LEDs?
Thank you very much.