GSSL HELP THREAD!!!

[Harpo]

Ill take a look again with fresh eyes.

...and using a magnifying glas. Tightest spots for this -12V rail to 0V reference voltage seem to be at the connector between main pcb AND control pcb.

I take it you're not passing audio as well?

With pulled out TL074, TL072 and sidechain-VCA and unit switched to bypass, audio should pass clean.
With these chips pulled out and possible shorts removed, does the -12V rail come back to life? (might read a lot higher without load connected).
The 22R and 79L12 seem alife, but the vreg is in self protection mode, caused by a short/heavy load downstream this rail.

 

[Everland]

Hi Harpo,

Yep, I've used a magnifying glass and multimeter also to check for shorts downstream from the 79L12 and I can't find anything.

With these chips pulled out and possible shorts removed, does the -12V rail come back to life? (might read a lot higher without load connected).

Not for me - I get a reading of -0.54V at all the pins where -12V should be expected. The 79L12 is still really hot too.

I don't know if this presents a clue at all, but at initial power up the reading is -1.5 volts which then falls steadily to the -0.54 reading.

 

[frazzman]

Ill take a look again with fresh eyes.

...and using a magnifying glas. Tightest spots for this -12V rail to 0V reference voltage seem to be at the connector between main pcb AND control pcb.

I take it you're not passing audio as well?

With pulled out TL074, TL072 and sidechain-VCA and unit switched to bypass, audio should pass clean.
With these chips pulled out and possible shorts removed, does the -12V rail come back to life? (might read a lot higher without load connected).
The 22R and 79L12 seem alife, but the vreg is in self protection mode, caused by a short/heavy load downstream this rail.

Thank you Harpo - with your help I've found and rectified the fault.

I had a short on TL074 between pins 3 and 4. As soon as it was removed the 79L12 stopped heating up and everything worked.

Everland - check out the picture i've attached. I used this to go through and identify the possible shorts that could cause our -12V to malfunction

RED are all paths leading from pins 1, 2 and 3 of the regulator. The green box is where my short was.

No doubt you will have a short somewhere in that red highlighted area.

I also used some circuit board cleaner and a tooth brush to remove all the flux and any small solder traces, once I did that I could see the short as plain as day. I find sometimes with flux around the place its harder to tell if something has shorted, especially when the pins are as tight as these are.

Thank you both for your help

 

[Everland]

Thanks Frazzman, but no joy.

I cleaned the PCB and used your diagram (thanks for that) to trace around that entire area looking for shorts but I could not find any! Just to be safe I even used a razor blade in some of those tight areas in case it was something microscopic I couldn't see with the magnifying glass but still the negligible voltage on the -12V side.

Very frustrating!

What would be the next step in finding this short? Is there a way to use the multimeter to find it? Should I start removing components?

Thanks for your help so far!

 

[Everland]

Using the multimeter I find I have continuity between pins 11 and 12 on the socket where the TL074 would sit. This seems like it shouldn't be the case right?
If so, does that help me to trace the short at all?

 

[Harpo]

Using the multimeter I find I have continuity between pins 11 and 12 on the socket where the TL074 would sit. This seems like it shouldn't be the case right?
If so, does that help me to trace the short at all?

Shouldn't be connected. Maybe use a cutting knife to seperate these pins from a pcb copper bridge, if there is a bridge in this spot (hence use a magnifying glas). The short between the -12V rail and 0V reference voltage can be in a different spot as well, FI between TL072-pins 3/4 or 4/5 or any other spot on pcb between these rails will measure the same short. From my "AND control pcb" you checked the control board pcb for shorts as well?

 

[Everland]

Thanks Harpo. I just realised that if I remove 79L12 the short between pin 11 and 12 on TL074 goes away. Does this point to a dead 79L12?

 

[Harpo]

Most likely.

 

[frazzman]

Hi Everland, sounds like you're on the right track. I use a DMM in resistance mode to help identify any shorts. Replace your regulator and let us know how you go. It could have got nuked in the process but I'd still be inclined to think there might be a short somewhere but hopefully not

 

[Fuccimain]

Hi I have a couple questions about the design/ circuit, perhaps if Jakob or anyone familiar with the schematics could help me out I'd be most grateful,..

1, the 5534 on the channel inputs, are they running for unity gain? if not how much gain are they supplying?
2, the 5532(x2) on each channels output, sort of the same, are they boosting the level, curious why a single opamp like      the 5534 isn't enough.
3, the makeup gain control, I'm having trouble deciphering it on the schematic, i see "makeup" next to the bypass switch, but I'm unsure where the pot is in the drawing, and I'd like to know is the makeup gain boosting the 5532s ? or the 1/4 of TL074 to the left of where it says make-up? (  i feel like the 5534/2s are the audio path and the TL074/2s are side chain path, but I'm not certain)

point to all of my questions is I'm thinking about replacing the IC opamps in the audio path with DOAs like the 2520 with a 2503 tranny on the outputs.. kind of a cross btwn the SSL and API 2500 is my desire.

You think I should replace each half of the 5532s on the outputs with a 2520 for each half or just simply go out from the DBX VCA into a circuit like the API 325 for the output..?

thanks for any possible insight.

s

 

[Fuccimain]

actually after reading the circuit description on the grapf site, I've pretty much answered questions 1 and 2 myself.

guess question 3 is still where i'm still confused, but I believe that makeup gain is coming from the main VCAs (controlled by TL074s) ...

 

[gyraf]

Makeup gain is just a DC offset applied to both the audio VCA's, controlled from the makeup gain pot.

The threshold pot does the same, just to the sidechain VCA.

Jakob E.

 

[MeToo2]

point to all of my questions is I'm thinking about replacing the IC opamps in the audio path with DOAs like the 2520 with a 2503 tranny on the outputs.. kind of a cross btwn the SSL and API 2500 is my desire.

You think I should replace each half of the 5532s on the outputs with a 2520 for each half or just simply go out from the DBX VCA into a circuit like the API 325 for the output..?

thanks for any possible insight.

s

The point of using two op amps on the output in the gssl schematic is to provide + and - phase relative to 0V ground to provide electronic balancing of the output, whilst still having sufficient voltage swing of the output with enough headroom under high signal without clipping.

If you're going to use a single 2520 discrete op amp and an output transformer you have galvanic balancing anyway, and plenty of current drive, so you don't need to do any phase splitting with 2 discrete op amps. You will need to wire up your transformer to provide some voltage gain (like an API312) because ±15V supplies on a single discrete op amp won't be enough to give you +24dBm of output (assuming you want +4dBm level as 0dB and an additional 20dB of headroom)

Same on the input: if you've got galvanic isolation via an input transformer, you don't need to replicate a complex electronic de-balancing circuit with dual op amps or isolation caps: just a simple inverting op amp circuit + input transformer should do the trick.

Humbly suggest you start a separate design thread if you want to discuss further.

 

[Aleguitarpro]

I built two GSSL with the new revision (the one with the Sidechain filter insert point in the pcb).
I noticed immediately that the s.c. insert point is indicated in the schematic before the 22uF cap but in the pcb the connector is placed after the 22uF, before going into the two 22k resistors.

The first "new" GSSL I made is a normal GSSL with only the main board and the control board.
I decided to put an EXTERNAL sidechain insert point using the connector in the new pcb (the point at the negative pole of the 22uF capacitor goes to the "s.c. send"-tip part of the sidechain insert jack and the "s.c. return"-ring goes in the two 22k resistors.
I wired directly a DPDT switch from the connector point in the pcb:
1. normal position shorts the wires
2. second position the signal goes into an external s.c. TRS jack
Everything works as expected.

I decided to make the second "new" GSSL with a second sidechain (turbo) and copying the S.C. insert point in the main pcb in my homemade TURBO (second sidechain) pcb.
This time I decided to put also a HPF with a rotary switch with 4 positions:
1. shorted (no HPF)
2. 220nF cap - 80hz HPF
3. 100nF cap - 160hz HPF
4. EXTERNAL sidechain

Obiouvsly I copied this thing for the mainboard and the turbo pcb.
Shorted it's ok. External sidechain it's ok but the filters (the 2 caps) absolutely don't act as HPF.
I tried to test them with a SINE generator but nothing happens with the HPF on...

I attached the schematic to show the point in which the sidechain filters are.

Any idea?
Anyone has made a new-GSSL with the caps in the S.C. insert point to obtain a HPF?

Thanks.

 

[Harpo]

I decided to put an EXTERNAL sidechain insert point using the connector in the new pcb (the point at the negative pole of the 22uF capacitor goes to the "s.c. send"-tip part of the sidechain insert jack and the "s.c. return"-ring goes in the two 22k resistors.
I wired directly a DPDT switch from the connector point in the pcb:
1. normal position shorts the wires
2. second position the signal goes into an external s.c. TRS jack
Everything works as expected.

Yes for external insert, No for external only input (IE key input).

I decided to make the second "new" GSSL with a second sidechain (turbo) and copying the S.C. insert point in the main pcb in my homemade TURBO (second sidechain) pcb.
This time I decided to put also a HPF with a rotary switch with 4 positions:
1. shorted (no HPF)
2. 220nF cap - 80hz HPF
3. 100nF cap - 160hz HPF
4. EXTERNAL sidechain

Obiouvsly I copied this thing for the mainboard and the turbo pcb.
Shorted it's ok. External sidechain it's ok but the filters (the 2 caps) absolutely don't act as HPF.
I tried to test them with a SINE generator but nothing happens with the HPF on...

The HPF caps are in series to the 22uF cap. The 220nF set -3dB cutoff @ 73Hz, the 100nF set -3dB cutoff @ 160Hz with following 10K (20K||20K=10K).
If you most likely have these switched in parallel to the 22uF cap, HPF wouldn't do much (0.723Hz without, 0.716Hz with paralleled 220nF cap and 0.720Hz with 100nF cap in parallel.)

 

[Aleguitarpro]

The HPF caps are in series to the 22uF cap. The 220nF set -3dB cutoff @ 73Hz, the 100nF set -3dB cutoff @ 160Hz with following 10K (20K||20K=10K).
If you most likely have these switched in parallel to the 22uF cap, HPF wouldn't do much (0.723Hz without, 0.716Hz with paralleled 220nF cap and 0.720Hz with 100nF cap in parallel.)

Of course... In series to the 22uF caps (the one in the main board and the one on the turbo board) but nothing noticeable happens...
I'd like to know if someone tried to build an HPF with a turbo GSSL in this way.

 

[gyraf]

I did this - and it worked like expected. Try sweeping through 20Hz-10KHZ, and watch compression rise above cutoff.

The effect is not dramatic, but great for preventing your bass drum from pumping the comp.

Jakob E.

 

[Aleguitarpro]

I did this - and it worked like expected. Try sweeping through 20Hz-10KHZ, and watch compression rise above cutoff.

The effect is not dramatic, but great for preventing your bass drum from pumping the comp.

Jakob E.

I've already tried with a sine sweep without success...
If you confirm me that the HPF works also with a "turbo" Gssl, I've to find where is the problem in mine!
Thanks

 

[gyraf]

I've had no part in the "turbo", so I  can't answer for that...

Jakob E.

 

[Harpo]

I've already tried with a sine sweep without success...
If you confirm me that the HPF works also with a "turbo" Gssl, I've to find where is the problem in mine!

If you don't want to cut a pcb trace on the turbo board and make it behave as on the latest GSSL version, the HPF switching would go between the lifted negative leg of the 22uF cap C2 and the solder lug where this caps leg was pulled from pcb.
I still would prefer a switchable HPF/external in front of the sidechain VCA(s) current limiting 47K input resistors on GSSL main pcb, so both an external insert and external key input would work as well. As always YMMV.