GSSL HELP THREAD!!!

[garbage.island]

A post to share my experience and add to the knowledge base.

Built GSSL kit from PCB Grinder fairly easily - Gustav was great help with a couple clarifying questions. Powered it up and only Left channel was outputting. Confirmed it was output of compressor - both L and R inputs affected the meter which seemed to respond correctly to the other controls (ratio, threshold etc).

Tried reflowing some solder connections and swapping ICs and a bunch of other troubleshooting which didn't help but made compressor non-responsive (no meter, no compression of audio).

Reflowed again and tried some other troubleshooting, then powered up and blew one of the 10R resistors!

Replaced resistor, and went through again, this time extremely methodically, to remove any solder bridge or short and reflowed most components. This time, after being VERY attentive and methodical, IT WORKS! Sounds great and acts exactly how it's supposed to.

So to sum up - Make sure all of your solder joints are clean as I believe this caused ALL of my issues. I'm an intermediate at this - not a beginner (have built close to 20 kits from DIs to 500 series EQ/Pres to guitar pedals) and have built a few simple designs of my own.

 

[justinheronmusic]

I just finished my GSSL, but I have an issue. I have found this problem across the thread a few times, but either it didn't work for me or it was never resolved.

The issue:
The meter reads either -20 in bypass or -16 if I move the threshold CCW in the slightest bit. It will vary depending on how far CCW I move it, however, the meter is maxed out at 4:00. The meter is very erratic as I move to from fully CW to CCW.

79L12 and 78L12 are getting extremely hot, and I disconnected the bypass switch out to make sure it wasn't that. I also disconnected the meter lamp from the optional 12v. I am testing just under 15v (14.88) at the +/- 15v rails. All VCA and IC are oriented properly.

I am going to order another VCA tomorrow to try to swap it out but I don't think that this is the issue, however, as addressed on page 8 I should give it a try at least. Feeling discouraged 

 

[Potato Cakes]

If 78L12 and 79L12 are getting extremely hot then then they are probably getting voltage from somewhere they shouldn't, probably related to the same voltage going to the meter and apparently to your Threshold control. What is the voltage across your + and - terminals of the meter when the unit is on and idle and the needle is pegged? Start by figuring out where voltage is coming from to the meter. I would go over the solder joints on the cables between the main and ratio boards as those solder pads are very close together. And make sure you didn't cross any of the connections as I had on my first GSSL build. Some of those connections carry 12VDC which comes from the regulators in question, so that might be a clue. If you get really stuck, disconnect the ratio board and remove all of the op amps and start all +/-12VDC connection points around the 78/79L12 regulators. Also for fun make sure all of your op amps are getting +/-15V at the appropriate pins.

Everyone says they quadruple double bubble zigzag checked their solder work/component reading so surely it can't be that, but it rarely is not. Myself included.


Thanks!

Paul

 

[Harpo]

79L12 and 78L12 are getting extremely hot

something drawing more current than these +/-12V regulators can deliver.
As both are getting extremely hot, most likely not a single short, but a blown or wrong type opamp fitted (FI a TL071 instead of TL072) or opamp TL072, TL074 or sidechain-VCA (different orientation than audio-VCA's) fitted with wrong orientation, case else,
power down, remove all +/-12VDC supplied chips and power on again. 78L12/79L12 should stay cool, else locate and remove the short or 22uF caps with wrong orientation.
Power down, refit the TL072, power on, Vregs should stay cool, else replace broken TL072.
Power down, refit the TL074, power on, Vregs should stay cool, else replace broken TL074.
Power down, refit the sidechain-VCA, power on, Vregs should stay cool, else replace broken That218x.
Good luck

 

[Muraro]

Hey Guys,

I'm about to build the SSL clone from Gyraf and have some doubts related to the VCA THAT 2181 calibration... I know this topic has already been intensively discussed here, but since the latest information available is a bit old and since most of the links provided is not working anymore, I decided to create a new post to clarify and update the topic...
This is a collection of almost all threads related to this issue that I read in previous posts:

Since 2150 is gone, we need to replace it either by the 2180 or the 2181. The 2180 is the NON-trimmable version which does not requires any resistor in the "%" slots from the last PCB version in the gyraf site (rev.#12). However, the 2181 will requires some resistors in these slots to perform the calibration.

THAT provides a release note (number #137) that clarifies how to perform the replacement from the old version (2150) to the new one (218x).

http://www.thatcorp.com/datashts/dn137.pdf

Now, focusing on 2181 which requires calibration...

On the previous posts I found the fllowing links which are still working and might have the correct information:

https://web.archive.org/web/20100523224841/homepage.mac.com/marten.thielges/gssl/calibration.html

http://diy.fischerworks.com/gssl_vca.shtml

Thus, based on these information, I believe the "%" resistors should be as following for 2181AL calibration:
Number are related to the attached picture:

- 1: It should be 120k for the compressor ratio adjustment
- 2: Resistor should be 5k1 as per THAT documentation and information in the websites above.
- 3: For the SC signal, resistor connected to pin 4 can be omitted, since the 218X versions has a 27R resistor which will do the job. So that slot can be left empty.
- 4: Resistor should be 5k1 as per THAT documentation and information in the websites above.
- 5: Jumper at this position to make the connection between pin 4 from VCA and the next resistor which will connect to pin 2 from the trimpot.
- 6: This is the resistor R3 in the THAT documentation and will vary depending on the VCA version used. For 2181A use 680K, for 2180B use 220K, for 2180C use 130K.
- 7: Based on THAT information, there is no resistor between pin 3 and pin 5 for 218X series, so, this slot should remain empty.
- 8: It shall be a 50k trimpot which will adjust the distortion level.
- 9: Since the 218X version has the internal 27R resistor, this resistor is not required and this slot can be left empty.
- 10: Summing resistor shall be 47k (both).

Please, let me know if my understanding is correct and correct me if I'm wrong.
The idea is to put all the information I found for the calibration in this post to renew the topic for those will start the construction now.
Hope it can help the others!!!

Regards,
Marco

 

[Gustav]

The idea is to put all the information I found for the calibration in this post to renew the topic for those will start the construction now.

Please keep this post free for trouble shooting, there is a separate thread for modifications.

Google, GssL VCA resistors...and voila!

http://fucanay.fischerworks.com/diy/images/gssl/gssl_vca.html?fbclid=IwAR1lc0l_RVsshuCGqzjyy_I1-8LxFbif0rCYbClFnDJrAS3HsGsqty0DXMc

Gustav

 

[Monomood]

Hey guys,

I just finished the kit by DIY racked and after some trouble shooting we managed to get it up and running at my friends' studio.

BUT now something unpleasant is showing up. After the transport from my friend to my studio the left channel always has ~3db less volume than the right channel. The problem only occurs when the compressor section is switched on. turning any of the knobs has no effect on the problem at all.
The overall compressor section is working perfectly, make up gain is working and dry/wet also, but I always get this 3db less volume on the left channel. When the compressor is bypassed the volumes are equal.

I already swapped the quad VCAs and all other ICs left and right but this had no effect. Is there anything I can do to narrow down the problem.
Thanks for your help

 

[Harpo]

...the kit by DIY racked ...

This is the GSSL help thread, not a DR-MB2 help thread.

...the left channel always has ~3db less volume than the right channel. The problem only occurs when the compressor section is switched on...

so the problem seems sidechain related. Lots of spots to cause and fix this error, ie sidechain filter, turbo circuit, wet/dry circuit...
Just compare control signal voltages from left side and right side sidechain with audio signal present in order to detect differences causing your 3dB difference. 

 

[aaronmcphee]

Hi Folks.

Does anybody have a drawing or schem which shows how to hook up the eternal side chain to an input jack?, and secondly can I actualy do this now I have the Turbo mod fitted?

Sorry if its a noob question but I couldn't find much on the help thread that clearly explains this.

Thanks.

 

[aaronmcphee]

Sorry guys, more noobe questions (probably)

My understanding of the Turbo mod is that is will give give stereo compression as apposed to mono on the standard GSSL build, so I ran a test signal through the GSSL with Turbo engaged (But Sidechain filter switch off), I then panned the signal hard L&R and the compression level didn't change, I thought compression should increase on the sides??

Then I switched ON the External sidechain and did the same thing, with the tone panned hard RIGHT but this time I got a a few db's reduction in compression, but with it panned hard LEFT or centered, the compression returned back to where it was.

So I took the turbo out of the equation and this time did the same procedure, first with the external sidechain on, then off. In both cases I got a drop in compression when panning hard left or right.....as I would expect.

I'm not very technical when it comes to understanding signal paths etc,  but it seems to me this might be a turbo board issue? I say this purely because the turbo is not giving me increased side compression, and I get the RIGHT side compression drop only when the turbo is connected.

Any thoughts would be most welcome.

 

[aaronmcphee]

Sorry guys, more noobe questions (probably)

My understanding of the Turbo mod is that is will give give stereo compression as apposed to mono on the standard GSSL build, so I ran a test signal through the GSSL with Turbo engaged (But Sidechain filter switch off), I then panned the signal hard L&R and the compression level didn't change, I thought compression should increase on the sides??

Then I switched ON the External sidechain and did the same thing, with the tone panned hard RIGHT but this time I got a a few db's reduction in compression, but with it panned hard LEFT or centered, the compression returned back to where it was.

So I took the turbo out of the equation and this time did the same procedure, first with the external sidechain on, then off. In both cases I got a drop in compression when panning hard left or right.....as I would expect.

I'm not very technical when it comes to understanding signal paths etc,  but it seems to me this might be a turbo board issue? I say this purely because the turbo is not giving me increased side compression, and I get the RIGHT side compression drop only when the turbo is connected.

Any thoughts would be most welcome.

I think I may have found the possible problem that is causing the slight mis-balance I am getting between the left and right compression when the turbo is engaged, I opted to use a THAT2180 on  my turbo as to match what I have on my GSSL, I was studying the sidechain schems for the turbo and the GSSL and realized I have a 3k9 on the turbo where I have a 5K1 on the GSSL...should I change the Turbo 3k9 to a 5k1?

Would just like confirmation.

Thanks.

 

[Harpo]

I think I may have found the possible problem that is causing the slight mis-balance I am getting between the left and right compression when the turbo is engaged, I opted to use a THAT2180 on  my turbo as to match what I have on my GSSL, I was studying the sidechain schems for the turbo and the GSSL and realized I have a 3k9 on the turbo where I have a 5K1 on the GSSL...should I change the Turbo 3k9 to a 5k1?

3K9 is setting the correct bias current (the THAT218x sidechain VCA is powered by +/-12VDC).
5K1 is for the +/-15VDC powered THAT218x audio VCAs.

 

[aaronmcphee]

3K9 is setting the correct bias current (the THAT218x sidechain VCA is powered by +/-12VDC).
5K1 is for the +/-15VDC powered THAT218x audio VCAs.

Gotcha. It was just a hunch because I realised that I was using the 2180 and the turbo manual lists the 2150, I had changed it anyway but it made no difference to the slight drop on the right side.

I'm inclined to just live with it as its quite minor and the results sound pretty good. I saw an older thread in which somebody else mentioned watching the same youtube video by "Really nice audio"  when he panned the audio hard L&R with turbo engaged  the compression increased, and vise versa in GSSL mode, so I was expecting to see the same behavior on mine however LCR compression stays the same in Turbo. although I recall you replied saying this is quite normal?

Thanks again for all your help Harpo, Ive learned a lot from your with your guidance.

 

[Jan Zornig]

Hi.. I hope someone can help me, it is my second GSSL build, the first one worked like a charm but this one is killing me. I've checked all soldering point and can't find any faults and opamps are all okay, but it keep burning the 10 ohm resistor going to 79L12 input leg at the minus 12v rail, board is rev. 7 and power supply is 2 x 15v 230v 15va. The other 10 ohm resistor going to 78L12 is fine..

 

[Gustav]

Hi.. I hope someone can help me, it is my second GSSL build, the first one worked like a charm but this one is killing me. I've checked all soldering point and can't find any faults and opamps are all okay, but it keep burning the 10 ohm resistor going to 79L12 input leg at the minus 12v rail, board is rev. 7 and power supply is 2 x 15v 230v 15va. The other 10 ohm resistor going to 78L12 is fine..

The function of those 10R resistors going to the xx15 regulators,  is to make a filter on the supply but when they burn, you can think of them as acting like fuses. They burn because the current draw exceeds the capability of the resistor.

So when they burn, the only thing we can say is, something is causing you to draw too much current.

Can be anything causing a short, or anything demanding a higher current draw than expected (using a real bulb, instead of an LED, for example).

If you mean the 22R resistors (which are actually the ones coming up to the xx12 regulators), problem is the same, but isolated to the 12v supply you mention.

Gustav

 

[Jan Zornig]

Hi Gustav. Thank you for replying!

First of all there is no 22 ohm resistor coming from the  79L12 at the rev. 7  and I'm aware that the cause is too much current being drawn across the resistor, I'm not a total beginner and I'm using LED for the meter. I've attached the schematic for the rev. 7
I was hoping someone had come across similar problem and new what to do, it would spare me time seaching for the problem 

 

[Harpo]

Either broken 79L12 or more likely a broken or reverse oriented  fitted 22 uF cap between the 10R/79L12 and 0V reference voltage.  Edit: While at it, double check the rectifier is still a rectifier, ie each diode inside this part only conducting in one direction. A short pre voltage regulator is an overcurrent condition for the rectifier as well that is not built for this lot of current demand.

 

[Jan Zornig]

Either broken 79L12 or more likely a broken or reverse oriented  fitted 22 uF cap between the 10R/79L12 and 0V reference voltage.  Edit: While at it, double check the rectifier is still a rectifier, ie each diode inside this part only conducting in one direction. A short pre voltage regulator is an overcurrent condition for the rectifier as well that is not built for this lot of current demand.

Hi Harbo.  Thanks for your reply!
I've already checked for reversed 22 uf caps, but they are all fitted the right way according to the silkscreen and tried to change the 79L12 too but still the same.
The bridge rectifier is 2w06, it should be rated at 50v 2 amp, even tried a new one, but nothing changed. I measured all the electrolytic caps at the -12v rail too. 

 

[Jan Zornig]

Thank you for helping Harpo and Gustav. I found the fault, it was a bad batch of chinese 79L12, went to the local store and bought a couple new 79L12 and no smoking 10 ohm resistor, so now for some testing to see everything is okay. 

 

[gyraf]

Damn, dét er en svær fejl at finde - forfalskede komponenter ligger simpelthen uden for hvad man forventer..

Godt arbejde!

Jakob E.