GSSL HELP THREAD!!!

[Barbera Media]

Another observation. I just performed the same test that Keith did to prove his Turbo board works, and found something interesting. In regular GSSL mode with one channel showing 4db down, I unmute the second channel and the meter actually moves DOWN about 1.5db instead of doubling like it should. What could this mean? When enabling turbo mode, the GR meter jumps to -14db and doesn't move when the second channel is added like it should. What could this mean about the regular GSSL mode? What could cause something like this?

 

[redmojosteve]

Hi,
I got my GSSL working and connected up to my home studio and it works really well. Better than I expected to be honest as this is the first project like this I have completed for a few years. Thanks for the help along the way! Next is a dual G1176 unit and then another GSSL with turbo board so I am reading the latest thread with great interest.

 

[Harpo]

  I'm getting :
       Pin 1 = +13.51 vdc - should be 0.0
       Pin 2 = 0.0 - o.k.
       Pin 3 = 0.0 - o.k.
       Pin 4 = -14.96 - o.k.
       Pin 5 = 0.009   - should be 0.0
       Pin 6 = 0.611  - should be 0.0
       Pin 7 = -13.51 - should be 0.0
       Pin 8 = +14.98

Check for all pins of the 5532 in their socket position. Could happen that a pin bends inwards when pushing the IC in. Doublecheck resistor values, especially the 15K (or 27K for unbalanced out) feedback resistor between pin 6 and 7 of the 5532 [pin 1/2 in schematic] (also check for a broken/lifted trace here) , the 27K in front of the VCA and all values in the DBX202 emulation circuit when using a THAT21xx VCA.
Pin 1 = +13.51 vdc - should be 0.0  [o.k. - should be same voltage as pin 7 but of opposite polarity]
Pin 6 = 0.611  - should be 0.0 is current input coming from the VCA. The chip tries to keep it at same voltage as pin 5 (virtual ground), but fails to get there for maybe previous mentioned reasons.
Good luck.

 

[audiophreak]

  Thanks Harpo... 

  I'll start there first !

 

[Harpo]

At first it seemed that all sound stopped, but upon closer inspection, the GR meter was pinned, and turning up the makeup gain all the way revealed a faint overly-compressed sound. Seems like the board is overly sensitive or something, makes the unit WAAAAY to sensitive. any faint sound above the sound floor activates it and the resulting sound is very faint. I've inserted several different TL074's and 2181's and also tried some 2180's. This didn't help at all. Checked all orientation of all diode's and other polarized components and everything checks out. Any ideas? I'm game. Thanks.

You got the right side of the 47K resistor lifted for switching between gssl and turbo mode?

 

[audiophreak]

Hi Harpo,

    I checked all the pins on the Opamps- none bent, and checked the resister values in the area and the ones you indicated and all are correct. Re-checked for solder bridges in the area and again none found.  :'(

    I purchased two sets of PCBs and populated both at the same time, putting in one group of same components at a time on each board, then checking off the BOM as I went down the list. So I swapped out the main PCB, and  ... this one the Left channel works but not the Right    .. but this one is different in that NO audio comes out at all, but the meter shows signal.
  also, all the voltages that you listed with empty sockets and Trancedentals list w/populated sockets is spot on with the second main PCB  ???

    I'd like to think I have decent soldering skills, but have resigned to the idea that its some solder bridge (s)  somewhere and just have to go over the board pad by pad til I find it....

    If you or anyone else have any other ideas it would greatly be appreciated,  I'm talking the weekend off this project...maybe read some more of this thread...maybe something will jump out at me and smack me in the head ..  ;D  and then start going over pads on Monday.. as I said in my G9 build... the bigest thing I learned was that I have a lot to learn.. ;D

  Thanks again,
              Chip


 

 

 

[intellijel]

I am just about to drill a hole for bolting the transformer on but it seems weird to me to have such a large bolt head sticking out of the bottom of my case.
Is this normal or is there are trick I should be doing? (like using a counter sunk bolt instead of the one that came with my transformer)?

 

[Harpo]

..NO audio comes out at all, but the meter shows signal.

The meter shows the amount of gain reduction coming from the sidechain for the summed L+R audio and the control voltage coming from the sidechain is obviously working. If the meter still moves when you remove the left side audio feed, the input stage of the right ch. must also be working. (or check if both 5534 pin 6 measure the same when fed with same input signal).

Next check is around the 5532 op-amp. You already have swapped the left and right ch.op-amps, so the right ch.op-amp alone should be working.
Bypass the right ch.audio-VCA by pulling it out of its socket and temporary plug in a piece of wire between VCA socket pin 1 and 8.
Audio now should pass the right ch (without compression of cause).
- Case not and you still have this -13.5V at 5532 pin 7, replace the broken 15k (or 27k) resistor or mend the broken trace between pin 5 and 6 causing this open loop.
- Case voltage at pin 7 is the same/half as audio in but still no sound, check for lifted traces, especially at the 100uF output caps. (don't ask me how I know  :)
With a little luck the from schematic upper one has survived the -13.5V reverse biasing, else replace this 100uF cap connected to the +/in phase output.

If now audio passes, either the right ch.audio VCA could be broken or the 5534 in the DBX202 emulation circuit is broken or has a wrong resistor value connected (especially the 120 ohm in the 1k/120R voltage divider to scale the control voltage down by factor 0.1).
Remove the temporary wire and swap L+R audio-VCA. If the fault now swaps from left to right side audio, replace the broken VCA, else replace the 5534 of the right ch.emulation circuit.

good luck

 

[tongenerator]

hello...

i have just finished soldering the gssl pcb
i have connectet one audio in channel only to see if it works...
the meter is not moving an inch and also there is no output signal

i have not connectet the 2 47 k pots and the other thing switch that is not on the controll surface pcb?

 

[redmojosteve]

Re fitting the transformer. I ended up with head of the bolt on the bottom of the case. There isn't much choice as the steel sheet of the case is too thin to effectively countersink. You could press a countersunk profile into the case to allow a countersunk head to be flush with the base but that isn't easy. I left the bolt head but put small feet on the case to prevent it fouling other equipment. If you are using a 1U case you may find the transformer and bolt only just fit. It's been said here before but be don't let one end of the bolt touch the top plate as the bolt and case in effect become a winding. You may need to grind a few threads of the end of the bolt to maintain clearance.

 

[audiophreak]

Harpo,
        Thanks so much for taking the time, I'll start checking all that out on Monday,
                                                                                                                Chip

 

[Barbera Media]

Harpo,

Yes, the right side of the 47k resistor is lifted.

Here's something even weirder. I bought 2 turbo boards for both of my GSSL's. I took out the board in question, compared it to the other board I had, there were no differences. All resistor values were the same in the same places, I even swapped the chips. Finally I swapped the boards out and this new one worked like it should.

So now i'm going to install the board in question into the other GSSL I have and see what happens. This is really blowing my mind.

 

[Barbera Media]

Well, I found out what was going on. When checking continuity between pins 4 and 11 on both TL074's i found that the positive buss wasn't checking properly. Upon closer inspection, I found that the bottom solder pad for the +12 input had fallen off. I guess it's time for a new board huh? Thanks for humoring me.

 

[tongenerator]

hello...

i have just finished soldering the gssl pcb
i have connectet one audio in channel only to see if it works...
the meter is not moving an inch and also there is no output signal

i have not connectet the 2 47 k pots and the other thing switch that is not on the controll surface pcb?

shame on me  :

u need to have the other lorly switch with that bypass thing connectet to get signal inside
u need to hae the make up gain thing connected to get any signal out the compressor
and i also forget to mountr one 15K resistor  ;D

well it seems to work now... will test it properly when it is in case...

it sounds like a little distorted... wich is probably something that has to get tuned
the only realy anoing thing so far ist the mad hum noise  

 

[tongenerator]

i have another question...

how do i get big holes into the metall for the vu meter
wich tools can i use for it?

 

[Harpo]

i have another question...

how do i get big holes into the metall for the vu meter
wich tools can i use for it?

Wrong order.
You start out with a hole of the correct diameter.
For a perfect circular outlook you use just the hole, not a plastic tube or alike.
Place this hole at the correct position on to a fireresistant form of the desired shape, maybe 19" wide.
Now take a material block of your prefered metal, melt it and pour it into the form around this hole.
Wait until its cooled down before touching or painting.

..

..

..

..

..

..

: ;D

more serious: cover your front panel with masking tape for marking and to prevent scratches and use a dremel, file, stepped drill, nibbler, punch, circular drill, saw, .. whatever works best for you or is availiable.
Have a look in the meta for working with metal.

 

[tongenerator]

thanks for the hint

 

[audiophreak]

  Hi All,

    Well...I now have two fully functioning GSSL main PCBs, one is compleate, the other needs a case. 

  Main PCB N0.1  had audio out of the Left Channel but it was loud and distorted and had no GR on meter...nothin. With Harpos and Trancedentals posts with voltage readings of Opamp and VCA sockets and also populated readings, I found a problem with the left output 5532. The Right channel worked fine.  I took the Gyraf PDF of PCB and "Mirrored " it in Photoshop and made it a little bigger to make it easier to check each trace and pad for solder bridges and lifted traces, I checked them ALL, no problems found. Checked resistor values as suggested, all good, removed left VCA and jumpered pins 1 and 8 as suggested, audio passes good. I had purchased a couple extra VCAs, I got the 2181BL08-u, and replaced the left VCA and BINGO... all works fine.  ;D

  I removed PCB No.1 and installed main PCB N0.2, this one has no audio out of Right channel, no GR on meter, nuthin. and got similar readings on the right output 5532 as I did on the Left 5532 on the other PCB.  :'(    ...  Left channel works fine. I replaced the right VCA, but didnt get that lucky  :'(  removed the No.2 main PCB and started checking traces and solder pads and found a solder bridge between pins 5-6 of the right 5532 - Harpo , you zeroed right in on it..   ...  reinstalled and all working as it should  ;D

  Thanks to Gyraf and Jacob,and all who posted before, especially Harpo for taking the time.... and I still have a lot to learn. ;D
                                                                                                                                                                              Chip

 

 
 

 

[Harpo]

Well...I now have two fully functioning GSSL main PCBs
..and found a solder bridge between pins 5-6 of the right 5532 - Harpo , you zeroed right in on it..

well done. congrats.
(so my guess of an open loop between pins 5-6 was wrong and instead they were shorted.)

 

[intellijel]

My gssl is done but it is not working

I have a barclaycon super side chain board installed and followed the directions from here:
http://picasaweb.google.com/barclaycon/GSSLSuperSidechainPCB#5185516904372762722

I have installed this in a purusha case and I am using a farnell meter. I put 330Ohm across the meter and used a 1k resistor for the meter sensitivity.
I also have a 47k pot attached to leg of the Threshold pot. I am getting this weird thing happening where if I back off on the threshold the needle floats to one side really slowly. When I use higher resistor values in place (I tried a 1k) I get very small movement on the meter, even at lowest threshold it is always a deviation of just 1 or 2 meter marks.

For VCA's I am using THAT2181LA's
I did all the mods as described here:http://homepage.mac.com/marten.thielges/gssl/calibration.html
except for the unity gain.

The meter is showing gain reduction but there is basically no compression effect on the audio, even at extreme settings. The makeup gain works fine.
When I connected an external side chain source (e.g. kick drum from a drum machine) the gssl compresses in a more expected way and I can hear it slightly pump on extreme settings but it is not that strong and the front panel controls seem to have little affect.

Another odd symptom is when I flick the switch to engage the compressor and I have a very low threshold setting, it slams the meter to max and then goes to a normal signal. At the same time as this happens I also hear the audio compress very heavily... it almost seems like it is compressing for that small fraction of time and then goes back to non-functioning mode.

Any help is greatly appreciated!!! I have spent several days debugging and I am getting no where

UPDATE:
I have checked voltage at all ICs and they are as expected.
Checked in great detail all connections and values around the TL072
Verified that the Threshold and Makeup pots are changing voltages at pins 1 and 7 on the TL072
Checked all resistor values, solder joints and bridges (about a million times)
Checked every point to point interconnection between the SSC, control panel and main pcb.
Replaced the Sidechain VCA
Replaced the TL072
Removed the sidechain processing board (but left its relay control of bypass) to help isolate if it was the problem

I am getting ready to throw this thing out the window!!!  ???