GSSL add-on help thread

[gyraf]

This thread dedicated to all sorts of add-ons and modifications to the GSSL - in order to clean up the original "vanilla" help thread.

Discussion and tracing errors in turbo, sidechains, relay bypass, CRC power cleaning, 202VCA's, etc etc..

Jakob E.

 

[DrywOwens]

Thanks, Jakob!

I Searched thru the vanilla GSSL help thread and found one hit from 11 years ago that was somewhat similar but thought I'd ask-

if you want the short version of my question, skip to the bottom

I'm modding an existing GSSL with it's own PCB design that includes the turbo mod on it's main board. I wanted sidechain filtering  ability added on. I used more or less the same schematic as this one- http://sounddiy.free.fr/images/SSL/Doc/Sidechain/Sidechain_Sch.pdf

with the exception of leaving out a few of the filters. and I duplicated the circuit since this is a turbo GSSL. The compressor also already has external sidechain ability, so I wired it so that when you engage the external, the existing DPDT relay switches over to the new sidechain filter board.

when I connect the output of the filter board to the main board (at the 47k resistors prior to the VCAs), the unit goes into a strange behavior where the GR shown on the meter reads inversely proportional to the threshold pot setting.

TLDR since this is kind of an oddball in that I didn't build the compressor that I'm troubleshooting, I'd like to know if anyone can shed light on what conditions would have to be in order to make the GR inversely proportional to the threshold setting? Just looking for where to start looking, and I've never built a GSSL before. Sorry if this is a dumb question.

Thanks!

 

[gyraf]

..maybe a DC offset at filter output? Should be AC-only...

Jakob E.

 

[innercityman]

Thanks for creating a thread for GSSL mods, as it began to be hard to search on the almost 400 pages of the GSSL thread.
I posted on the regular thread about my problem with turbo mod, when I realized that one side was not going in the side chain (the turbo one). I made few tests beginning by disconnecting the turbo and coming back to the original design and everything was fine... both sides compressing when muting one side after another when feeding the unit with a stereo signal. I checked my voltages on the pin 4 and 11 of the TLO74 of the turbo board and I had the +12 and -12 as expected. I also checked continuity between the two TLO74 (on turbo board and GSSL board) and it was ok two. So Gustav advised me to try to switch the TLO74 and the VCA of the turbo  with the ones on GSSL to see if the problem was coming from one of them. Unfortunately I soldered them without sockets so when trying to desolder, I damage the board so I ordered a new one and populated it again with new components and today I've been able to make few tests.

What I'm experiencing now with the GSSL inserted on the main mix buss of my console is that both sides are compressing but one is compressing a lot more than the other when panning all the way left or right. When letting the pan knob of the console master to the middle, the sound is ok, balanced and compressed the same way on both sides.

If someone has an idea, I'd be thankful.

Also I wanted to know what means the letter on the THAT 2180 A, B or C. I'm asking this because the letter on the chip I put on the turbo is different from the 2180 on GSSL. (Maybe stupid question...)

Pierrick

 

[innercityman]

BUMP

 

[Harpo]

..of the TLO74..

There is no letter 'O' in this part. It's a TL074. Makes a difference when searching for the parts datasheet.

What I'm experiencing now with the GSSL inserted on the main mix buss of my console is that both sides are compressing but one is compressing a lot more than the other when panning all the way left or right. When letting the pan knob of the console master to the middle, the sound is ok, balanced and compressed the same way on both sides.

Whatever magnitude 'a lot more' might be, with THAT2180 sidechain-VCAs, double check, you only fitted one 47R resistor per sidechain-VCA (keep the one connecting to pin3. Leave out the 47R that connects to pin 4, ment for a 2150 type VCA.)

Also I wanted to know what means the letter on the THAT 2180 A, B or C.

Maybe read the parts datasheet ? (A-grade comes with up to 0.01%THD, B-grade comes with 0.02%THD, C-grade comes with 0.05%THD, measured @1V, 1kHz, 0dB)

 

[innercityman]

Hey Harpo, thanks for the reply.

Sorry for my poor knowledges but I'm a DIY newbie and GSSL was my first build. I just built an EZ 1073 and it works perfectly.

(keep the one connecting to pin3. Leave out the 47R that connects to pin 4, ment for a 2150 type VCA.)

You talk about the 47K resistors right ? I did like in the turbo installation guide... I lifted one leg of one of the 47K and wired it to the SC input of the turbo, and left the one next to it soldered on the GSSL board. Sorry again but which 3 and 4 pins are you talking about...? pins on turbo board ? And what do you mean by "leave out"...? has it to be desoldered ?

 

[Harpo]

You talk about the 47K resistors right ?

No. A 47R. (that is 46953 ohms difference to 47K)

I did like in the turbo installation guide...

...describing parts values for a 2150 type VCA.
You fitted 2180 type VCAs on Turbo board and main board instead.
On main board you probably already left out the 47R that connects to this type VCA pin 4. (This type VCA has 8 pins in a SIL package. Pin 4 is the 4th pin of this VCA, counting from left. Have a look at the datasheet.) Essentially, you want both L/R sides behave the same, so use same parts values. If you have this VCA mounted in a socket, it might be easiest to bend the VCAs pin 4 sideways, so it doesn't connect to its socket.

 

[innercityman]

Ok Harpo, I understood... Yes I looked at the GSSL layout and there is only one 47R resistor beside the side chain 2180, and there is 2 on turbo board. What I don't understand is why the turbo design is made for a 2150 (and it's printed on the board) but in the BOM it's a 2180...? I guess because the kit from Gustav like the one I built goes with 2180s, so why nothing is mentioned about it on the turbo build guide?

Essentially, you want both L/R sides behave the same

Of course yes, I use it as a buss compressor like everyone do I guess, and the unit behaving the same from left to right is what should be at least expected from a stereo comp.

Anyway, thanks a lot for your precious help. I'll do that this evening and let you know.

Pierrick

 

[innercityman]

Hey Harpo,

I've been able to do a  test after I've desoldered the 47R which was connected to pin 4 of the turbo 2180, following your advice. So what I can say is that there is still a difference between left and right side when fully panning from left to right. The difference is not as big as when there were the two 47R but approximately, left side is compressing about 6db more then right side when fully panning my master channel pan knob.
So I was guessing if maybe another resistor needs to be unconnected to the 2180 as I saw on the GSSL schematic that the side chain VCA is a 2181. So is the difference between 2181 and 2180 only that one is pre trimmed and not the other? Anyway, this won't help me figure out what is going wrong... What I can say is that the turbo board's side chain still hasn't the same values than the one on the main board as the signal is not treated the same from one to the other. I saw on the GSSL schemo that the 10k connected to pin 5 of the VCA is also connected to a 3k9 resistor but on the GSSL layout it's a 5k1, so Harpo, do you think this could make the difference? I checked on my turbo board and there's a 3k9, so do you think I would have to change it for a 5k1?

I think it would be interesting that people who installed the turbo mod in their GSSL with 2180 or DBX 2150 could make the same test and give some feedbacks about what happens when inserting the GSSL in the master output of their mixer or sound card, and mute one side then the other and see if the sound is the same both sides... Or is it just me who experience this strange behavior???

 

[Harpo]

...I saw on the GSSL schemo that the 10k connected to pin 5 of the VCA

the 10K once was an ineffective try to decrease noise modulation of the DBX2150 VCA. I'd leave this resistor out. YMMV.

...is also connected to a 3k9 resistor but on the GSSL layout it's a 5k1

just one of the little errors. For the +/-12VDC supplied sidechain part of the circuit you want a current setting 3K9 resistor connecting to this THAT2180 VCAs pin 5, so the turbo silkscreen shows the correct value for Iset  (12V-4*0.7V diode drop)/3900 ohms=2.4mA and the 5K1 at the sidechain VCA on main board (would be only allowing 1.8mA) needs the change to 3K9. (Keep the 5K1 resistors at the +/-15VDC supplied THAT2180 audio VCAs).
I have no idea what parts and parts values you populated at what generation of board.

 

[innercityman]

Thanks for the answer, I was about to replace the 3K9 on turbo board by a 5k1. The version of my GSSL is the last one I guess, revision number 11, bought from Gustav about 8 or 10 months ago.  For the part values of the turbo, I just followed Rochey's bom that I downloaded from his site, but it certainly needs to be updated as It was done in 2008. For the GSSL, I bought Gustav's full kit (PCB + components) also the bom is downloadable from his site.

Therefore in summary... The 47R connected to pin 4 of the THAT 2180 on turbo board needs to be left out (I've done it already), the 5K1 on GSSL needs to be replaced by a 3K9, and the 10K (both on GSSL and Turbo Board) need to be left out too right ?

 

[main repairman]

Hi. I build GSSL with HPF (by Steffen) and i need your help.
Say me please, which function have this trimpot 10k on picture in this SC HP.
Value R7 is correct in the feedback loop?  Because input voltage opamp pin 6 is 1VAC and output voltage pin 7 is min. 0.08VAC and max. 0.916VAC. (1*1/12=0.08), (1*11/12=0.916).  Output voltage pin 7 has be 1VAC or i'm wrong??? Which output voltage i have to set? Thanks

 

[innercityman]

I made some progresses... I changed the 5k1 resistor near the side chain VCA on GSSL for a 3k9 like on the turbo board, and the improvement is clear, but there is still a tiny difference between the GSSL's side chain and the turbo one. maybe 3 dbs of compression.

 

[andow]

Hello everybody!

I'm in the process of populating my GSSL board with turbo, cavendish and supersidechain mod. I have a few questions regarding the cavendish mod:

In the PDF from Expat it says that the capacitors should be "greater than" 22uF/47uF, so what are reasonable values to use? I guess the 47uF coupling cap can't be big enough, but what about the 22uF? In the API312 e.g. I saw 12.5uF...

Another thing I noticed is that the output of the EA2503 transformer is taken from two windings in series from both the primary and secondary and the input is fed to one secondary winding. I never saw that before, what's the reason for this configuration?

The PDF also states that by using the trim pots on the cavendish board it is recommended to remove the 7k5 resistors on the main board. Since there is no such value, do they mean the 15k feedback resistor of the 5532?

Thanks!

 

[Gustav]

Thanks for the answer, I was about to replace the 3K9 on turbo board by a 5k1. The version of my GSSL is the last one I guess, revision number 11, bought from Gustav about 8 or 10 months ago.  For the part values of the turbo, I just followed Rochey's bom that I downloaded from his site, but it certainly needs to be updated as It was done in 2008. For the GSSL, I bought Gustav's full kit (PCB + components) also the bom is downloadable from his site.

Therefore in summary... The 47R connected to pin 4 of the THAT 2180 on turbo board needs to be left out (I've done it already), the 5K1 on GSSL needs to be replaced by a 3K9, and the 10K (both on GSSL and Turbo Board) need to be left out too right ?

Saw your comment in the main thread..

I'd basically try  to copy the  set-up around the SC VCA on the main board.

The supply voltage going to the SC VCA on the GssL boards is 12V, and I dont know if you are tapping the turbo board from  12V or 15V supply rails, but it may help if also check and take that into consideration.

Gustav

 

[DrywOwens]

Hi everyone!
I have a question regarding a switchable turbo mod on my SSL clone. My "turbo section" is already integrated into my PCB. The trouble is that every time I switch it off (thereby summing L+R into one detector circuit), my gain reduction increases dramatically, making it difficult to make a fair A/B assessment. Of course this is to be expected, as the "turbo off" switch is summing the signals into one detector VCA, so that VCA is getting a hotter signal. I figure I can add some resistors to the "turbo off" switch position to trigger a 3db each, or 6 DB total reduction of the signals being summed to the detector VCA in order to avoid the excessive gain reduction. My question pertains to which value resistor(s) I should be using to attenuate the signals being summed to the detector VCA. Or, what impedance shall I use to calculate the resistor values? Sorry for the basic question and thanks in advance for answering!

 

[innercityman]

The supply voltage going to the SC VCA on the GssL boards is 12V, and I dont know if you are tapping the turbo board from  12V or 15V supply rails, but it may help if also check and take that into consideration.

Nice to see you back. In fact I already checked my voltages  and I measured +12v and -12v on pin 4 and 11 of the TL074 on turbo board.

I found a blog on the net of a guy who built a gssl with turbo with the exact configuration as me. He explain how he wanted to retain the HPF switch functionality and how he wired it to keep this function, and he exactly did what you told me to do... Use a DPDT switch instead of the SPDT and lift a leg of the 22uf cap on turbo. So I asked him to make a test for me by sending a stereo signal on the GSSL and muting one side then another or panning the signal from left to right, and he told me that he didn't have the behavior I got. He had exactly the same signal and amount of compression on left and right side. He populated the turbo board following the BOM from expat, so I guess he didn't even do what I did, which were to leave out the 47r resistor connected to pin 4 of the VCA on turbo board and replace the 5k1 by a 3k9 on GSSL. SO I'm really lost

 

[andow]

Hello everybody!

I'm in the process of populating my GSSL board with turbo, cavendish and supersidechain mod. I have a few questions regarding the cavendish mod:

In the PDF from Expat it says that the capacitors should be "greater than" 22uF/47uF, so what are reasonable values to use? I guess the 47uF coupling cap can't be big enough, but what about the 22uF? In the API312 e.g. I saw 12.5uF...

Another thing I noticed is that the output of the EA2503 transformer is taken from two windings in series from both the primary and secondary and the input is fed to one secondary winding. I never saw that before, what's the reason for this configuration?

The PDF also states that by using the trim pots on the cavendish board it is recommended to remove the 7k5 resistors on the main board. Since there is no such value, do they mean the 15k feedback resistor of the 5532?

Thanks!

Anybody?

 

[spreadspell]

Hi,

in the SideChain Board BOM there is a Relay 24V DPDT 2800 Coil. So i found in local store and use instead of this an Omron G2RL-2 (24VDC) , and i was wondering if this is the cause that i'm not hear any proccessed sound when compressor button is ON. Also the 3 Leds (meter, power button, compressor button) working properly.

Thank you...