GSSL Turbo problem ! FIXED :-)

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

MartyMart

Well-known member
Joined
Feb 28, 2006
Messages
2,340
Location
Berlin for a while
Fitted and tested voltages @ the TL074 all OK, when bypassed passes signal
but engaged the VU meter pegs and all output is lost !

I'm sure that I understand the very clear instructions but perhaps I've done
something stupid.
The only broken and re-made connection is the lifted 47k, all others are additional
connections to what's there already.... right ?

My 12v/-12v/0v are not taken directly from the 10 pole connection, but at the first
available "node" and these voltages test fine in circuit.
Point "C" is taken at the top of the diode which connects directly to pin 10 of the edge
connection - so that's also the correct "node"

What have I done wrong ? ... Keith !
Ta,
Marty.
 
When you say 'engaged', I'm guessing that you mean you made it switchable?

The turbo output is always connected to the unit's GR control voltage, so nothing changes there... -Did you make it switchable? (I'd better ask, before I continue my assumption!) -If it only goes ballistic when the switch is engaged, it sounds like there's something wrong with the switch wiring perhaps...

SO you have ±12V at pins 40 and 11, right? -Howzabout DC voltages at pins 1, 7, 8 and 14, with a silent input.

Keith
 
I mean engaged with the original GSSL bypass switch, that's all.
Have made this "Turbo only" for the moment.

Pin 1 0.68v
Pin 7 0.13v
Pin 8 0.19v
Pin 14 -3.38

Thanks,
MM
 
Right. The first three op-amp stages look good. the last (pin 14) looks unneccesarily negative.

The schematic for this stage is EXACTLY the same as for the GSSL, the exact same pins of the TL074. -Pin 12 should measure 0Ω to 0 volts when powered down, pin 13 should measure 0 Volts at all times when powered up, pin 14 should ONLY go below negative 0.5V or so during compression action.

-This is one of the two boards which I built -right? -If so, then I know it's been tested and works correctly, so just check that the wiring to and from the ratio switch is correct.

AS a pointer to begin with, the inverting input (pin 13) should be connected to the POLE (wiper) of the switch. The resistance across R6 (1meg) should change with the ratio: about 1 MΩ for 10:1, about 400kΩ for 4:1 and about 200kΩ for 2:1... (resistances measured with the power OFF)

Can you just check those for me?

Cheers,

PS I typed this all up once before and then closed the window by accident... :Grrrrrr:

Keith
 
OK, across R6 I have 620k @ 10-1 400k @ 4-1 and 192K @ 2-1
Not far off but no where near the 1M @ 10-1 !

General +/- 12 voltages read +12 and -12.5

Was the original GSSL speced with a 4 pole switch ?
Just wondering if that pole is totally redundant ....

MM.
 
Well, if you're measuring across the 1MΩ resistor and the resistance is changing, then it must be switching resistances... You should be able to measure and compare switched resistances with the corresponding 1MΩ resistance in the original GSSL board.... That's the 1MΩ resistor with one end connected to point 'A' and the other end going to point 'B'.

I'd suspect the board, but in this case I checked all three boards that I built out, and the only problem that I encountered was when I put a diode in backwards on one of them... (duh!!!). -Once I'd put it in the right way, everything was just spiffing, just plugged in and worked.

Double-check the ratio switch connections for me, would you? Make sure that the pole is on the right spot etc. -That's all that I can think of...

Keith
 
Thanks Keith, switch wiring is all fine, I'll check the bottom of the main board
perhaps I popped a resistor whilst soldering on the "new" stuff - did it with the boards all still in place.

MM.
 
Troubleshooting notes:

TL074 voltages at pins 1, 7 & 14 on the turbo board should all match those on the same pins on the main board. Pin 8 is used for a different function on the turbo compared to the main board, so don't look for a match on that pin.

Then send the SAME audio signal (preferably constant, like a sine wave) to both inputs and check for the following on both the main board and the turbo, for comparison.

Pin 7: Is the audio signal recovery buffer after the sidechain VCA. Scope for a "rising and falling" audio signal as the threshold is turned down and up. There should be no DC offset when no signal is present.

Pin 1: should be a mostly rectified waveform... all bumps in the same direction. There should be no DC offset. Again, the bumps should rise and fall in amplitude as you wiggle the threshold control.

Pin 14: may have a DC component (during silent periods) on it which may well change absolute value when different ratios are selected. The DC voltage should be the same as the main board at all times, so if the main board 074 Pin 14 goes to different DC values, so should the turbo 074 pin 14.

...Pin 8 on the Turbo 074 is a DC buffer which should wiggle up and down as you wiggle the threshold control around. So long as pin 8 matches pin 10, it's doing its job correctly.

If you don't have a scope (I don't know if you have one or access to one, Marty) then take silent-input DC measurements at all the pins and then signal-present input at all the pins... Pin 7 should be unchanged, but pins 1 and 14 may give different readings on the main board (the exact amount of difference depending on how the front panel settings are, and how much the unit's trying to compress the signal at any given moment) but the turbo should at all times give the SAME readings on pins 1, 7 & 14 so long as both channels are seeing the same input signal.

Stick with it, bro!

One other thought: -What's the DC voltage reading at point 'C'? (i.e. where the two diodes sum). -Is is about 2.5volts negative?

Keith
 
[quote author="SSLtech"]
Stick with it, bro!

One other thought: -What's the DC voltage reading at point 'C'? (i.e. where the two diodes sum). -Is is about 2.5volts negative?

Keith[/quote]

Erm ... it's -7.95 volts !!! so that WAS a good question :wink:

Point C, which is pin 10 on the ribbon connector goes round a trace to a single
diode the top point of which is where I connected to the turbo point "C"
Have I miss read that trace ?

MM.

EDIT` - Gotta run, leaving for Berlin at 7am !!
Will continue on thursday afternoon, Ta mate.
 
Sounds like you're not at the right point 'C' then... maybe the trace HAS been misread, perhaps...

I think we're onto the problem. If the problem was coming from the turbo board, the voltage on the OTHER side of the summing diode (point C) would be about 0.6V dropped... but it's actually higher instead...

We'll get back to it whan you return.

Keef
 
For the record, I took some TL074 DC voltage measurements on my unit here... test condition: silence at both inputs.


Code:
Ratio   Main PCB pin 14 voltage    Turbo pin 14 voltage
 2:1              +0.206V                  +0.206V
 4:1              +0.323V                  +0.324V
10:1              +0.357V                  +0.358V


 Main PCB pin 1 voltage    Turbo pin 1 voltage
           +0.005V                  -0.002V

 Main PCB pin 7 voltage    Turbo pin 7 voltage
           -0.053V                  +0.088V

Turbo pin 8 varies between approx. 0VDC and +2.5VDC as the threshold control is turned through its complete range of travel; Approx. +1.2VDC at the mid point.

From this, we can see that pin 1 should have basically no DC present (which makes sense), pin 7 can have some DC present, maybe in the region of ±100mV (this is 'hidden' from the world by a diode, so this also seems to be quite reasonable) and pin 14 should have about +0.20VDC when set to 2:1, about +0.32VDC when at 4:1, and about +0.35VDC when set to 10:1.

Point 'C' should sit at about 0VDC when no compression is taking place. -Increasing gain reduction (compression) sends this voltage progressively further NEGATIVE... It's still very bumpy here (before the time constants & smoothing section) so there's no easy correlation between measured DC voltage and gain reduction, but a couple of volts of measured negative DC will equal a LOT of gain reduction.

Keith
 
Hi Keith, I'm back - thanks for taking the measurements.
Have just checked the GSSL layout and my point "C" connection is fine, it's
the diode that is the first component from the "main PCB" point C and I have
soldered to the top of this.
I'll check some voltages and perhaps swop out the TL074 ( I have dozens of them )
Otherwise I just can't see what's wrong with my build, will check my "0" volts point
as it appears to be sitting at about 0.03v at present !

MM.

EDIT : Just a thought, would it make ANY difference as my main board uses
2180LB's and therefore has all the trimmers/1M's / 10k's missing as per the
setup table posted by ..... erm ..... Matt ?? ( sorry if the name is wrong )
 
[quote author="MartyMart"]would it make ANY difference as my main board uses
2180LB's and therefore has all the trimmers/1M's / 10k's missing[/quote]
no difference.

Just for 'Gits and Shiggles', I unplugged the 'signal' connector on my turbo to see if you can take any measurements, and the voltages go completely haywire, so that's not a test which we can do.

Unwire your diode solder-point to point 'C' and temporarily tape the end of the wire, then see how the voltage measurements are on the turbo board.

Check voltages at pins 1 & 7 (silent input) then pin 8 as the threshold control travels through its range, then pin 14 at three different ratios.

[edit] It should however be possible to check the turbo in "stand-alone mode" if you unplug the signal connector, but clip the "Signal in" and "CV in" both to ground. -Is this an easy test for you to try? -Did you connectorise yours?[/edit]

Also, as for the 0.03Volts on your '0V', it depends on your meter... Does your meter have a separate "DC mV" range (as many flukes do)? -If so, be sure to use that one.

Keith
 
OK, results following your test suggestion to the letter :

Pin 1 0.68
Pin 7 0.14
Pin 8 0.18 to 0.22
Pin 14 @ three ratios -3.41 -5.18 -5.86
All with silent input.
I hardwired so not easy to do the other test and DMM is a cheap £8 one so I
don't have that particular Fluke setting.

Meter still "pegs" fully clockwise !
 
Pin 10 same as Pin 8 0.18 to 0.22
Pin 14 from -5.18 to -5.86

Would have to be a "new" solder bridge, perhaps on the bottom of the board
Will check, probably tomorrow morning now ... have to go see my "Lady" !

Seems "odd" having just done the few Turbo connections quite carefully ??

Mucho thanks Keef, I'll get there ....

MM.
 
I messed up with the pin 8 theory...as soon as you disconnected the 'Sidechain CV' connection, it should sit around 'nothing' volts... call me silly.

However, with point 'C' disconnected, the turbo board can no longer cause the main board to compress, so if your GR meter still pegs, maybe we should disconnect the turbo for a moment and go back to making sure that the GSSL hasn't sprouted an internal problem of its own as a result of the disturbance...

Keith
 
Just had the main board out - bitch to do 'cos the case is a little small and
it's all a tight fit around the I/O and VU meter ! - I didn't build the main board up,
I bought it and re-fitted correct caps and checked it out though.
It was operating fine until I started fitting the Turbo.
Couldn't see any problems that could have just been caused, however, I connected
point "C" under the board to pin 10 and now have almost -10v at the point "C" on
the turbo board !!
I simply made the connections for Turbo as cleanly as possible, without disturbing
the build very much .... :-(

I need to check it out tomorrow as I've run out of time for now.
Will remove Turbo and get it back to stock.

Cheers,
Marty.
 
Cool. -For now, maybe sniff around to see if you have any flying 3-conductor connections to use temporarily (even XLRs!) for the DC power and signal connections.

I would be thinking about the board, but it was tested and working. All three boards were swept and checked to match perfectly...

Bummer... I WON'T let it beat me though!

:wink:

Keith
 
Keith, OK so removed the turbo and it all went back to "normal" operation
right away !
Didn't find any nasty solder blobs or "dodgy" looking connections on the back
of the main board and "top of the Point C diode" returned to 0v too.
I used a new 47k and 100r ( where connections are made for turbo ) as the fitted ones
seemed to have quite flimsy leads on them, new ones are much thicker.

One observation though, I/O op-amps read -15v pins 4 and +12.8v pins 8 ?
Is my meter LED taking away from the +15 supply a little ?
I have the Meter connected direct from the +15v "jumper" between the two VCA's
via a 12k resistor. ( just for the LED of course and heatshrinked )

Any more thoughts ?

Regards,
Marty.
EDIT : One comment on the Turbo board, I found quite a lot of flux on the
top of it, so have scrapped it all off and given it a clean up.
 
Back
Top