GSSL HELP THREAD!!!

[frazzman]

Very confused, whats the purpose of the 'c' connection anyhow? Havent got my head around the turbo board entirely...

The Turbo board is essentially a second side chain amp/ detector circuit copy of the one on the main board.

Point 'C' is where the output of the Turbo side chain amp/detector circuit and the original main board side chain amp/detector circuit connect back into the original side chain timing filter on the front panel daughter board.

There's a 1n4148 diode on the main board at the output of the original side chain detector, and also one 1n4148 on the output of the Turbo side chain detector board, meaning that point 'C' should be the most negative of the two control voltages = compress both channels by the max of (-CV_L , -CV_R).

Perhaps check for the correct orientation of the diode on the Turbo board connected to pin 14 of the TL074, and also that the output of the Turbo board isn't permanently stuck at Vcc or Vee, before connecting the Turbo board back to point C.

Thanks for your reply...! I'm still wrestling with this one...

Checked diode orientations - all good, checked for any shorts - all good, replaced VCA and TL074 - made no difference.
I noted that changing attack and release settings causes a popping noise and makes meter spring to 0 then straight back to pegging far right.

In my eagerness I noticed that I hadn't soldered the top of the board (newb error??? I haven't done any double sided pcbs) .... I guess this could be a contributing factor? I just went over all the diodes on the top of the pcb, but haven't done the resistors....

 

[MeToo2]

Very confused, whats the purpose of the 'c' connection anyhow? Havent got my head around the turbo board entirely...

The Turbo board is essentially a second side chain amp/ detector circuit copy of the one on the main board.

Point 'C' is where the output of the Turbo side chain amp/detector circuit and the original main board side chain amp/detector circuit connect back into the original side chain timing filter on the front panel daughter board.

There's a 1n4148 diode on the main board at the output of the original side chain detector, and also one 1n4148 on the output of the Turbo side chain detector board, meaning that point 'C' should be the most negative of the two control voltages = compress both channels by the max of (-CV_L , -CV_R).

Perhaps check for the correct orientation of the diode on the Turbo board connected to pin 14 of the TL074, and also that the output of the Turbo board isn't permanently stuck at Vcc or Vee, before connecting the Turbo board back to point C.

Thanks for your reply...! I'm still wrestling with this one...

Checked diode orientations - all good, checked for any shorts - all good, replaced VCA and TL074 - made no difference.
I noted that changing attack and release settings causes a popping noise and makes meter spring to 0 then straight back to pegging far right.

In my eagerness I noticed that I hadn't soldered the top of the board (newb error??? I haven't done any double sided pcbs) .... I guess this could be a contributing factor? I just went over all the diodes on the top of the pcb, but haven't done the resistors....

I can't think of anything that would make the GR meter go hard right in normal operations ....

So maybe check that your ±12V power supplies are holding up
Check pins 4 & 11 of the TL074 on the Turbo board for Vcc & Vss.

Double sided board shouldn't be the problem: the holes should be plated through and solder will flow in automatically anyway .....

Comparing the Turbo with the original detector is your friend. They are in essence electrically identical (expect of course for the physical lay out)

Try temporarily unhooking the connection to point C from the output of your Turbo board.

Does the unit compress OK (triggered by a signal on one channel)? Or does the needle still go hard right?

If it still goes hard right that looks like mis-wiring or a short somewhere.

Try connecting up the output of your Turbo board to a scope or a high impedance volt meter, still without connecting it back to point C. You should be able to see the control voltage jump negative as audio spikes come in, and then return to 0 as the audio drops. You can compare that to the voltage at point C i.e. the output of the other channels' detector on pin 14 of the TL074 to try to isolate the fault further.

Check that the side chain input is connected up OK (audio in on one channel with 0 DC offset) arriving at pin 1 of the 2181.
Your main board should only have the other channel feeding in for permanent turbo. Check the input levels are similar on pin 1 of the equivalent 2181 on the main board.

Check pin 7 of the TL074 has little or no DC offset and that it passes audio OK (could be attenuated depending on the control voltage)
Compare it to pin 7 of the TL074 on the main board....

Check the control voltage on pin 3 of the 2181. It should be very near 0V and only go slightly negative (very few mV) when compressing. Compare it to pin 3 of the 2181 on the main board.

Check the wiring of your ratio switch..... remember this has to be electrically isolated from the original ratio switching (using a different pole on the same 4x3 switch is OK)

 

[frazzman]

Thanks MeToo2, I'm still working through your suggestions but so far what I can say is:

- Meter stops pegging immediately as soon as point 'c' is disconnected, compression behavior seems fine whenever 'c' is not connected.

- Put together another turbo board with new components from scratch, rewired all connections to mainboard and ratio switch: same problem still !

- Can get some audio with release set to full and make up gain at full, but meter is still almost fully pegged.

- Tested voltages on TL074 at Pin 14:
Look good on mainboard:
at 2:1 = 0.17V
at 4:1 = 0.33V
at 10:1 = 0.37V

But not so good on turbo board:
at 2:1 = 0.17V
at 4:1 = 0.19V
at 10:1 = 0.25V

So I guess I can concede that there are some issues with the -/+ 12V going into turbo board?

None of my gssls have been turbo up til now, it seemed so damn easy but for some reason i've really hit a brick wall....... Thanks again

Very confused, whats the purpose of the 'c' connection anyhow? Havent got my head around the turbo board entirely...

The Turbo board is essentially a second side chain amp/ detector circuit copy of the one on the main board.

Point 'C' is where the output of the Turbo side chain amp/detector circuit and the original main board side chain amp/detector circuit connect back into the original side chain timing filter on the front panel daughter board.

There's a 1n4148 diode on the main board at the output of the original side chain detector, and also one 1n4148 on the output of the Turbo side chain detector board, meaning that point 'C' should be the most negative of the two control voltages = compress both channels by the max of (-CV_L , -CV_R).

Perhaps check for the correct orientation of the diode on the Turbo board connected to pin 14 of the TL074, and also that the output of the Turbo board isn't permanently stuck at Vcc or Vee, before connecting the Turbo board back to point C.

Thanks for your reply...! I'm still wrestling with this one...

Checked diode orientations - all good, checked for any shorts - all good, replaced VCA and TL074 - made no difference.
I noted that changing attack and release settings causes a popping noise and makes meter spring to 0 then straight back to pegging far right.

In my eagerness I noticed that I hadn't soldered the top of the board (newb error??? I haven't done any double sided pcbs) .... I guess this could be a contributing factor? I just went over all the diodes on the top of the pcb, but haven't done the resistors....

I can't think of anything that would make the GR meter go hard right in normal operations ....

So maybe check that your ±12V power supplies are holding up
Check pins 4 & 11 of the TL074 on the Turbo board for Vcc & Vss.

Double sided board shouldn't be the problem: the holes should be plated through and solder will flow in automatically anyway .....

Comparing the Turbo with the original detector is your friend. They are in essence electrically identical (expect of course for the physical lay out)

Try temporarily unhooking the connection to point C from the output of your Turbo board.

Does the unit compress OK (triggered by a signal on one channel)? Or does the needle still go hard right?

If it still goes hard right that looks like mis-wiring or a short somewhere.

Try connecting up the output of your Turbo board to a scope or a high impedance volt meter, still without connecting it back to point C. You should be able to see the control voltage jump negative as audio spikes come in, and then return to 0 as the audio drops. You can compare that to the voltage at point C i.e. the output of the other channels' detector on pin 14 of the TL074 to try to isolate the fault further.

Check that the side chain input is connected up OK (audio in on one channel with 0 DC offset) arriving at pin 1 of the 2181.
Your main board should only have the other channel feeding in for permanent turbo. Check the input levels are similar on pin 1 of the equivalent 2181 on the main board.

Check pin 7 of the TL074 has little or no DC offset and that it passes audio OK (could be attenuated depending on the control voltage)
Compare it to pin 7 of the TL074 on the main board....

Check the control voltage on pin 3 of the 2181. It should be very near 0V and only go slightly negative (very few mV) when compressing. Compare it to pin 3 of the 2181 on the main board.

Check the wiring of your ratio switch..... remember this has to be electrically isolated from the original ratio switching (using a different pole on the same 4x3 switch is OK)

 

[MeToo2]

Thanks MeToo2, I'm still working through your suggestions but so far what I can say is:

- Meter stops pegging immediately as soon as point 'c' is disconnected, compression behavior seems fine whenever 'c' is not connected.

- Put together another turbo board with new components from scratch, rewired all connections to mainboard and ratio switch: same problem still !

- Can get some audio with release set to full and make up gain at full, but meter is still almost fully pegged.

- Tested voltages on TL074 at Pin 14:
Look good on mainboard:
at 2:1 = 0.17V
at 4:1 = 0.33V
at 10:1 = 0.37V

But not so good on turbo board:
at 2:1 = 0.17V
at 4:1 = 0.19V
at 10:1 = 0.25V

So I guess I can concede that there are some issues with the -/+ 12V going into turbo board?

Have you actually measured your power, or is that a guess? A Turbo board can take quite a lot of current: 20-30mA or so if my memory serves me well.

It's looking more to me like a significant DC offset is creeping in somewhere. Rather than just blindly rebuilding and changing components, I humbly suggest that you continue testing with point C disconnected, feed in a mono test signal to both left and right channel inputs, and walk through the schematic one component at time starting from the input source at pin 1 of the That 2181 side chain VCA working towards point C (pin 14 of the TL074) to identify the point where the main board behaves significantly differently from your Turbo board for both AC and DC measurements on your multimeter relative to signal ground. It's only 1 off 9 pin VCA chip and 1 off 14 pin op amp to check. Be very careful to not short out adjacent pins. Shouldn't take you long. And that'll be where your problem is.

 

[frazzman]

Ahh thank you MeToo2....

I heeded your advice and guess what, found the problem and fixed it....

Turned out to be a power supply issue to the turbo board... tested voltages on TL074 & VCA on turbo board and compared them to mainboard...no suprise they were way off.

It was some dodgy molex connectors that were the deal breaker, next time screw terminals....

Anyway it all works perfectly now and comparing voltages between opamp on turbo & m/b net nearly identical results....

Thanks again for pointing me in the right direction

Thanks MeToo2, I'm still working through your suggestions but so far what I can say is:

- Meter stops pegging immediately as soon as point 'c' is disconnected, compression behavior seems fine whenever 'c' is not connected.

- Put together another turbo board with new components from scratch, rewired all connections to mainboard and ratio switch: same problem still !

- Can get some audio with release set to full and make up gain at full, but meter is still almost fully pegged.

- Tested voltages on TL074 at Pin 14:
Look good on mainboard:
at 2:1 = 0.17V
at 4:1 = 0.33V
at 10:1 = 0.37V

But not so good on turbo board:
at 2:1 = 0.17V
at 4:1 = 0.19V
at 10:1 = 0.25V

So I guess I can concede that there are some issues with the -/+ 12V going into turbo board?

Have you actually measured your power, or is that a guess? A Turbo board can take quite a lot of current: 20-30mA or so if my memory serves me well.

It's looking more to me like a significant DC offset is creeping in somewhere. Rather than just blindly rebuilding and changing components, I humbly suggest that you continue testing with point C disconnected, feed in a mono test signal to both left and right channel inputs, and walk through the schematic one component at time starting from the input source at pin 1 of the That 2181 side chain VCA working towards point C (pin 14 of the TL074) to identify the point where the main board behaves significantly differently from your Turbo board for both AC and DC measurements on your multimeter relative to signal ground. It's only 1 off 9 pin VCA chip and 1 off 14 pin op amp to check. Be very careful to not short out adjacent pins. Shouldn't take you long. And that'll be where your problem is.

 

[MeToo2]

Ahh thank you MeToo2....

I heeded your advice and guess what, found the problem and fixed it....

Turned out to be a power supply issue to the turbo board... tested voltages on TL074 & VCA on turbo board and compared them to mainboard...no suprise they were way off.

It was some dodgy molex connectors that were the deal breaker, next time screw terminals....

Anyway it all works perfectly now and comparing voltages between opamp on turbo & m/b net nearly identical results....

Thanks again for pointing me in the right direction

No worries. Glad you've got your Turbo working now. Enjoy  8)

 

[nadje]

Hey guys,

So i'm building a gssl and i have a question about the vca NE 5534 voltages. Before i've inserted my chips i measured the +15/ -15/ +12/ -12 voltages on the board and they are correct. I've inserted the ic's and when i turned on my unit the ne 5534 chips started to frie after a few seconds. They are wasted. I took out all the ic's checked the voltages and there is something really wrong (according to what i've read in the meta thread). On pin 6 of the 5534 i have +15 vdc and on pin 7 +4,7 vdc. Is there someone who has encountered this before?

 

[gyraf]

Check for shorts on the pcb side?

 

[masterboyle]

There is a space in the audio VCA section for a capacitor the value of which is not marked. I can't find it in the VCA emulation circuit schematic, maybe it is a supply bypass cap and is not included in the schematic. Based on my parts count I think they are the 100n polypropylene, because using them there would bring me up to the total of 25 which greg's BOM says are required. Any help appreciated.

 

[frazzman]

Hi guys,
I'm having some intermittent issues with one of my GSSLs retrofitted with a turbo board....
I want to redo the connections from the mainboard to the turbo board however I don't want to use the connections of the 10x molex connector pins for supplying power to the turbo board . I think its too congested and might be the source of my issues...
Can anyone recommend some alternate places on the mainboard to tap into +12V, 0, -12V to supply power to my turbo board that won't impede on the mainboard??

 

[3nity]

Guys.
Is it normal for the make up potentiometer to light up every time you turn it??
By light up i mean sparks....??
Its scary i even saw some smoke coming out!!

 

[3nity]

Alright sparks are fixed.

Im having problems calibrating the meter im using...i got distortion on the audio.

On another topic i used trimmers instead of the 27K resistors before each VCA..

help me !!!! ;D

 

[Harpo]

Im having problems calibrating the meter im using...i got distortion on the audio.

On another topic i used trimmers instead of the 27K resistors before each VCA..

Hi Jorge,
watch out trimmers/rheostats initial settings.
Depending on type of (<=1mA DC fsd) meter used, initial setting for the 4K7 meter trim will be about 2K for a 1mA meter showing 20dB fsd gain reduction. The feeding TL074 will not happily drive less than a 1K2 load.
For a GSSL in balanced out config these initial settings resistor values before the audio-VCAs will be 30K (double value of the feedback resistor in the VCA following current-to-voltage converting opamp stage), for unbalanced out this will be 15K (same value as the feedback resistor).
For a THAT218x (whatever type of VCA you might use in your build, they are all different) the recommended Input+Output signal current is 0.35mA rms or 0.5mA pk. In a standard GSSL the audio signal is always running thru the VCA, so in bypass condition both i+o signal currents are evenly shared. (with hardwire bypass you might weight currents a little to the VCAs input side, because output currents will be reduced by the min.2:1 compression ratio). The NE5534 line receiver will typically not swing higher than +/-13V from a +/-15V supply. 13Vpk/(0.5mA/2) is 13K. Higher value resistor will give less THD at the cost of noise figures and vice versa. A 10K in series to your maybe mistakenly shorted out rheostat/trimmer as initial setting might help to not fry the VCA from max.allowed signal currents.

 

[3nity]

Cool Harpo noted all your tips!
Thanks.

 

[gg85]

Hi, i have just finished my gssl build. Things weren't great on first power up. There were a few shorts causing different output levels as well as causing the unit to heavily compress when compression was engaged. That's all been sorted now, but I have an issue with my output levels. I have -14db going in, and I get -8.6db back out with no compression. I have been searching the pcb for any more possible shorts and cant seem to come up with anything. I've even swapped out IC's but still get the exact same problem.

The increase in level occurs before the VCA's (i'm guessing this is normal though). Using my scope I can see signal before the 27k resistors into the VCA's, but nothing after the resistor. No signal seems to be showing up after the two 47k resistors in the sidechain either. -15/+15 and -12/+12 rails seem to be fine.

Any suggestions would be appreciated greatly

George

 

[Harpo]

..The increase in level occurs before the VCA's (i'm guessing this is normal though)

The increase in level can occur in every active stage with surrounding level defining passive component of the circuit.
This can be at the debalancing line receiver (all 4 resistors at NE5534 same value 22K ?);
this can be at audio-VCA in bypass condition (pin3=0V) with following current-to-voltage converter (from latest schematic 27K in front of the VCA and 15K feedback resistor in the following I2V-stage) use 27K Rin and half this Rin value 13K5 (=2x 27K in parallel) for the feedback resistor (alternative use 30K Rin/15K Rfb) when connecting to a balanced out config; use same values 27K Rin/27K Rfb when connecting to an unbalanced out config;
this can be with negative control-(milli)voltage present at VCA-pin3; ...

Using my scope I can see signal before the 27k resistors into the VCA's, but nothing after the resistor. No signal seems to be showing up after the two 47k resistors in the sidechain either.

All fine. The -from schematic- right side of these resistors connect to a virtual ground node (opamps inverting input). This is a current node.

 

[gg85]

thanks for your help i found the problem....i had 15k resistors before the vca's instead of 27k. i noticed i have rev 7 pcb and the resistor value was changed in rev 9...

wouldn't have even thought of it if you didnt mention the issue being with a passive component so thank you very much.

george

 

[PeteC]

Hi all,

how critical are some of the resistor values in the GSSL ?

I have most parts from my stash but am short of a couple of resistors, but I can parallel up and get pretty close. 

how about the 3.3meg in sidechain circuit - I can parallel a couple of 5% and get 3.35Meg - OK to use ?

cheers
Pete

 

[PeteC]

And also any recommendations on where to best source 2u or 1u 19" ali cases in the UK ?

Rgds
Pete

 

[PeteC]

Hi Folks,
another GSSL on the way....

so today I got everything set up and have been stuffing the boards which I got from Chef in UK.

waiting on a some nice parcels from RS and Rapid which should arrive tomorrow with everything else I need apart from my meter ( still cant decde which one to go for ) and 19" rack case ( still cant decide ....etc ...)

Having built LOADS of tube audio gear this one is a real departure for me - so Ive been super careful with parts checking and soldering ( using my BIG "grannies" magnifying light !!!)

really pleased with how the pcbs are building up - my neatest work for a while!

been reading up this evening on the HPF add-on which I'd like to ionocporate - does anyone have pcbs or a graphic so I can etch my own.  I'd like to have HPF so that I can use on mixes which have heavy drum content etc.

Already thinking about next project and ordering - I would like a nice mic pre, have already decided I will build a second compressor though - perhaps a 1176. 

Whats a nice mic-pre without busting the bank.  PS - I have loads of tube parts around so a simple tube mic pre would be easy for me if anyone has recommendations .

I dont need lots of bells and whistles - just want a good circuit.

meanwhile - can anyone help with the sidechain HPF layout or PCB ?

cheers all, and thanks for a fantastic and helpful resource.

Rgds
Pete