GSSL add-on help thread

[fritz]

Double check for steady +/-12VDC again with makeup-gain in both CW and CCW settings in order to exclude a wrong hooked up bypass switch.

- both TL074 show a constant voltage of 12.27 VDC (pin4) and -11.96 VDC (pin 11) in both (CW & CCW) MakeUp-Gain settings

Measuring voltage at the VCA pins 1 or 8 doesn't make much sense, as the VCA is a current-in/current-out device, not a voltage-in/voltage-out device. Compare VAC input voltages again at the other side of the input resistors (opposite of the resistor leg that connects to VCA-pin1).

- If you mean this point (see figure), then I measure 000.2 VAC on both boards

Did you also measure the diodes on the Turbo board? No need to desolder them. Can do in place. -- This should reveal a dead diode. Do the same with all diodes on the main board

Until then, I didn't measure the diodes, which I should have done because, in my opinion, there are inconsistencies here. I have entered the values of the measurements in the following sketch. I measured the diodes in both directions and two diodes seem to be defective. So in total there would be 4 defective diodes. The two defects on the main board are probably the corresponding ones of the turbo board, right? if i have understood correctly, my multimeter should only display one value if the diodes are not defective.

Could also measure all resistors on the Turbo board and compare to values of corresponding ones on the main board. Whatever those values are (cos resistors are interacting with surrounding components), they should read the same, as the Turbo board is an exact copy of the main board sidechain, it only looks different -- This should reveal faulty solder joints.

- the resistors are identical on both boards and also the values.

I'd say your readings at 4:1 look somewhat off. I'd check the voltages of all ratios (with threshold CW and CCW) and check the diode in the ratio network in particular.

- here again the current values in CW and CCW:

Mainboard Ratio (TL074 Pin 14):

• 2 = 0,780 CCW & 0,172 CW
• 4 = 0,301 CCW & 0,338 CW
• 10 = 0,138 CCW & 0,372 CW

Turbo Ratio (TL074 Pin 14):

• 2 = 0,777 CCW & 0,224 CW
• 4 = 0,193 CCW & 0,345 CW
• 10 = 0,206 CCW & 0,380 CW

the different values are probably due to the defective diodes, right? the diode in the ratio network works well. could not find any errors here.

The release timing caps. The position shouldn't make your needle jump. I'd check for orientation of the caps. Those tantalums have a plus and minus, but at least on older boards the silkscreen was hard to read (as are the plus signs on those caps).

- the tantalums are all correctly oriented

so now first replace the diodes and see what happens?

 

[Script]

While at it, could match diodes. Helps get closer control voltages

Also check for solder joints near those diodes.

 

[babaorum]

Identifying input pin 1 of THAT VCA.

Ha sorry I thought you mean test the support of the THAT not the THAT itself
I just tested it seems that I putted the THAT in the wrong way ! because I have voltage in my 'supposed 7-8' so this is my 1-2 in fact !!
test in context this evening (france)

 

[babaorum]

So it was not that if I reverse the THAT : nothing run well, I think it was in the good way.
But now , I know that in turbo my turbo board seems to not compress the sound I test with a 1000 Hz If I pan left compressor doesn't work (in turbo mode) , in standard mode all is right compress left right (-3dB of comp gain instead of the middle) and middle. I'll do measurements between main and turbo board to compare.

 

[Gustav]

Ha sorry I thought you mean test the support of the THAT not the THAT itself
I just tested it seems that I putted the THAT in the wrong way ! because I have voltage in my 'supposed 7-8' so this is my 1-2 in fact !!
test in context this evening (france)

If you reversed the VCA - its dead.

Gustav

 

[fritz]

While at it, could match diodes. Helps get closer control voltages

Also check for solder joints near those diodes.

have now desoldered the two diodes on the main board to check whether they are really broken. False alarm, now the diodes are showing correct values. I don't understand the world anymore. solder joints near those diodes all look good.

 

[Harpo]

- If YouTube mean this point (see figure), then I measure 000.2 VAC on both boards
View attachment 81725

No (and measuring 200mV in this close to virtual ground spot doesn't sound sane). Input resistors for the sidechain-VCA's are the two 47k on main board, one with probably a leg lifted, going to the turbo board. Measure AC voltage on the not lifted side of these resistors with same maybe 100Hz audio signal applied to both l/r gssl inputs.

 

[fritz]

ok, understand, but what do i do with the supposedly "defective" diodes? simply solder it in again and measure again?

one with probably a leg lifted, going to the turbo board.

the two 47 k resistors were completely removed because the two sidechain filters are attached to them. only as information so that you have an idea. I also uploaded a sketch of the wiring in my first post. if interested, here again

 

[Script]

Diodes. Looks like it was a symptom, not the diseases. Better replaced with matched anyway. Interesting that it happens on both boards.

//
Check for connectivity between points B and C in both modes. There should be no connection at any time !! Also shake the unit a bit (loose wire connection? solder debris ?). Test again.

Do you run LEDs or anything else-- apart from the Turbo board -- of the 12V rails ? If so, disconnect. Problem persists?

Take your Turbo select switch out of the equation by hardwiring the unit for permanent Turbo mode. Problem persists ?

Your sure your 12V is really stable when compressing music with the comp and your needle jumps? Are the regulators running hot ? Replace. Problem persists?

 

[fritz]

Take your Turbo select switch out of the equation by hardwiring the unit for permanent Turbo mode. Problem persists ?

before i get to work i would be interested to know what happens to the red cable that goes from the 47k section to the turbo button. just not connect?

 

[fritz]

ok, I have now re-soldered and repositioned for permanent turbo.

Diodes. Looks like it was a symptom, not the diseases. Better replaced with matched anyway. Interesting that it happens on both boards.

yes, they are only symptoms because the new diodes show the same values as the old ones.

Check for connectivity between points B and C in both modes. There should be no connection at any time !! Also shake the unit a bit (loose wire connection? solder debris ?). Test again.

No connection between Point C and Ponit B

Do you run LEDs or anything else-- apart from the Turbo board -- of the 12V rails ? If so, disconnect. Problem persists?

no, only those that are built into the push buttons. These are connected as in the sketch.

Take your Turbo select switch out of the equation by hardwiring the unit for permanent Turbo mode. Problem persists ?

yes

Your sure your 12V is really stable when compressing music with the comp and your needle jumps? Are the regulators running hot ? Replace. Problem persists?

12V are stable at main and Turbo board yes. Needle is jumping and the compression can be heard clearly. have now also swapped the voltage regulator (only 7815 & 7915). These get warm, but not so hot that I can no longer touch them. Think that's normal, or shouldn't they get warm at all?

I also noticed that the needle not only deflects to the right when I switch between the ratio, but also when I turn the compression back again (with no sound), but then does not deflect quite as strongly as with the ratio. All other switches do what they should do, no problem here

 

[Script]

Orientation of 22u on Turbo ? Let me guess : all good.

You have voltage on your ground ? Doh, your soundcard also acted weird before cutting the pins 1 on the output XLRs.

Maybe through the ground of those two switches? Or elsewhere.

 

[babaorum]

After several measurements and tests I have the explanation, it's just the type of 2181C which provided not from the same source, 1 is an 'Ali' chinese. Now I have only 2dB differnce between right and left panning. If I inverse the THAT i Have the same diff but in the other side So I just need to order a pair of THAZT which will be identical and my GSSL will be very OK.

 

[Script]

And I was about to ask where one can get an ICs without orientation mark ; )

 

[babaorum]

Actually, I had a second THAT identical than the Turbo one, I put it on the main board , now all works fine because the same level in turbo mode !
My GSSL is finished and it sounds very very good

 

[fritz]

Orientation of 22u on Turbo ? Let me guess : all good.

yep.

You have voltage on your ground ? Doh, your soundcard also acted weird before cutting the pins 1 on the output XLRs. Maybe through the ground of those two switches? Or elsewhere.

checked everything. no voltages on the ground. I noticed, however, that the small voltage regulators (78L12 & 79L12) are getting quite hot. now do not know if this is normal, but do not think If I exchange these, the new ones will be just as hot, certainly just a symptom. I think it's hopeless. a professional has to look at it. My half-knowledge is simply no longer sufficient here.

 

[Script]

Frustrating. I know what you mean. This also is DIY.

12V and 15V regulators both running a bit warm here too (about the same). Still possible to touch.

If it was me, before sending unit off for repair (time/cost ratio), I'd try to isolate the problem by disconnecting everything in steps. -- I always take closeup pictures (!) for easy reassembly when building, troubleshooting or repairing.

Several steps (random order). In your case, checking diodes after each step and after switching unit back on.

(1) Unplug all audio cables on XLRs.
(2) Disconnect input and output crimp connectors .
(3) Disconnect 10 pin crimp connector.
(4) Disconnect all LEDs and lights. That is really all of them..
(5) Disconnect all remaining wires on both switches (ON/OFF and Turbo). --That is really all wires. (If power and ground connections are fixed on both ends, I usually just cut the wires -- and later hook them up and solder them back together again [shrink tube]. Otherwise I tape loose wire ends for protection.)
(6) Disconnect both SSC boards (you use crimp connectors?)
(7) Disconnect turbo board (crimp connectors?)
(8) Disconnect metre (and add-on VU board).

Could also take out all ICs at some point. You use sockets? ---Check diodes.---

This leaves you with a bare main board connected only to the transformer for checking correct connectivity and voltages at empty socket pins. (The very first step in almost any DIY project).

If the problem at some point disappears, you know where to look.If it doesn't, there is not much left where it could be .

I assume your transformer is generally up to the task and connected correctly.

 

[Harpo]

the two 47 k resistors were completely removed because the two sidechain filters are attached to them. only as information so that you have an idea. I also uploaded a sketch of the wiring in my first post. if interested, here again
View attachment 81765

So measure AC voltage at the 47k resistors before the sidechainfilter-outputs with audio signal applied, maybe 100Hz, so your multimeter can handle it. Should measure the same, else further measurements don't make much sense. Maybe the level discrepancies are already in front of the filterboards.
If you hooked up the sockets for the pushbutton-switches as shown in the drawing (orange wire connecting to -15VDC instead of +15VDC) and the LED's light up, you have connected the switch upside down, causing erratic behaviour of the following circuit.
Connecting two sidechainfilterboards, you'd probably link the L+R inputs per scf-board, so both 20k input resistors are in parallel, giving 10k for unity gain response, else you'd lose 6dB control signal.
Voltage regulators running hot might be caused by too high transformer secondary voltages.
Else continue as Script already suggested.
Good luck

 

[fritz]

If you hooked up the sockets for the pushbutton-switches as shown in the drawing (orange wire connecting to -15VDC instead of +15VDC) and the LED's light up, you have connected the switch upside down, causing erratic behaviour of the following circuit.

thought that was all right. if it's wrong with me, then it must be wrong here too, right? GSSL Turbo Sidechain Wiring
I basically stuck to the sketch by don-audio

Connecting two sidechainfilterboards, you'd probably link the L+R inputs per scf-board, so both 20k input resistors are in parallel, giving 10k for unity gain response, else you'd lose 6dB control signal.

I do not really understand this. should the points R also be connected?

So measure AC voltage at the 47k resistors before the sidechainfilter-outputs with audio signal applied, maybe 100Hz, so your multimeter can handle it. Should measure the same, else further measurements don't make much sense.

I'll check that again this evening and then I will let you know

 

[craigmorris74]

I recently completed a GSSL build with CRC and Turbo boards and 202x VCA's. I researched the SSL/DBX documents and made all the appropriate component changes to accommodate the 202x's.

I'm in the process of calibrating the unit, and have noticed a level difference of about .3-.4 dB between the left and right channel (in and out of bypass, Turbo or no Turbo). Levels are the same all the way up to the 202x's, and the level difference appears after their output. I've trimmed them both for lowest possible distortion.

I don't see any component differences between channels or soldering errors on the boards, so what would be the best solution to match the channels?

Thanks,
Craig