GSSL HELP THREAD!!!

[Olegarich]

Checked for 4 times more, no shorts, bridges, everything in right place, "That 2180" is not inversed, orientation of components is good ,everything as is should be. (((

Do i have to see signal coming out from Sidechain VCA that 2180B??? cause i see signal in 1 pin. I've tried to change it with "that" from dbx emulation but stil nothing. Besides I've tried to push it hardly so it's really in place.

((((

 

[Harpo]

I have changed the 15k resistors in front of the audio vca's for 27k as noted in the schematics. So if I understand correctly you are saying change the 15k resistors connected to the output of the audio vca's (pin 8 ) to 27k ?

Same value. For unbalanced operation make both 27K or both 15K per L/R channel. 15k might be prefered for a -3dB cutoff at 106kHz instead of 59kHz with Rfb 27k if you don't change the 100pF compensation cap to 47pF accordingly for same LPF rolloff.

 

[Olegarich]

Cheking for right components for more 10  times. Gives me nothing. I wonder if the problem can be in modification I made for That 2180B so this mod is:

in the righ corner of PCB where it written with a big letter SIDECHAIN there are 3 resistors:

47R
47R
10K

In this thread there was post about to remove 10K and 47R near capacitor (22uF) is that might be the problem?

 

[Harpo]

keep the 47R in the middle (shunt arm resistor of the 470R/47R voltage divider that connects to THAT218x-pin3).
Remove the 47R that connects to THAT2180-pin4 or bend VCA-pin4 sideways so it doesn't connect to the VCAs socket.
Remove the 10K between THAT218x-pins3/5 (was ment for a different type of VCA to decrease noise modulation).
This will not be the reason for your problem, measuring a signal voltage in this current node.
Maybe read my last reply to your post again. Close up pics of the sidechain section from component side and trace side might help us to locate a fault.

 

[Olegarich]

Harpo,

First of all thanks for patience and help, really thank a lot.

I'm on my way to the studio where I'll make some pictures. I read you first post to me and now reread it, but perhaps i have to learn about current in-out, because it's not clear for me. But as I understand now, if I put 1 khz signal and with scope i can see it after Ne5534 and till sidechain vca 1 pin? but in vca pin8 there is no signal anymore and it's normal because it's about current in-out device? sorry if it dumb question. Probably have to learn more about VCA)))

One more time, thanks  man

 

[Olegarich]

Whuuuuu, things going well.... looks like now tl074 and tl072 are good have  - treshold pot controlling signal, haven't pluged yet to monitors but scope shows me that output sound changes with treshold ratio atack and release. let's move further!!!

 

[Olegarich]

My meter uses 6 volts for lighting, as far as I understand making 6 volts from 12 taken from PCB is not good idea. Would be ok to take from +15 and put a resistor? it's 20ma so I wonder which way to go

1.take from secondaries with resistor
2.take from pcb +15 with resistor
3.take from transformer and build small circuit with Lm317?

What do guys think?

 

[Harpo]

My meter uses 6 volts for lighting...

Could be incand.light bulb(s) or 2 or 3 LEDs in series for the more common suspects...

as far as I understand making 6 volts from 12 taken from PCB is not good idea.

Yepp. Don't connect additional load to these onboard + or -12V rails.

Would be ok to take from +15 and put a resistor?

You don't take from but connect between two potentials. Could be between +15 and 0V, between 0V and -15V or between +15V and -15V for these available DC voltage rails. Value and power rating of a resistor required to drop voltage depends on voltage differential and current drawn.

it's 20ma so I wonder which way to go

20mA might be LEDs max.allowed current that you don't want to exceed (and might be too bright) if your lightning source are LEDs for real.
Assuming there are series connected LEDs inside your meter, the voltage dropping resistor value needed would be (Vsupply - LED Vforward) / LEDs allowed current, could be (15V - 6V) / 0.01A = 900 ohm if you connect between +15V and 0V with this voltage dropping resistor in series and allowing 10mA current for your LEDs to not burn your retina. Resistors power rating required would at least be (15V-6V)=9V differential * 0.01A = 0.09W. Doubling up required min.rating, an 1/4 W rated resistor will do for these example values.

1.take from secondaries with resistor
2.take from pcb +15 with resistor
3.take from transformer and build small circuit with Lm317?

1. transformer AC voltage might be prefered if your lightning inside your meter are not (polarized) LEDs.
2. DC voltage (pre or post voltage regulator) might be prefered for supplying polarized parts.
3. Would work but requires some additional parts (rectifier, caps, voltage and/or current setting network), else you let the smoke out.

What do guys think?

Knowing what part you want to supply seems essential.

 

[Olegarich]

Thanks Harpo,

meter is 19-B1 I bought it here:

http://diypartssupply.com/index.php?main_page=index&cPath=3&zenid=u4j6thocnlom7lbb9bu79d8gi0
Couldn't find datasheet, but I've got answer, that:

it's LEDs, 5.6-7.6V at 20mA and they are connected to the small tabs

Probably gonna try with resistor beetwen +15 and 0

 

[Olegarich]

Allrighty so my Gssl compressor fully working! : : : ;D Two channels works well with no hum or distortion, meter working well, two channels same volume, all the controls, everything as it should. Actually I'm glad I had issue (there was short after I removed 10K corner resistor) ) because that made me to understand principle how it works. Harpo thanks one more time!

So next is turbo mod)

 

[Olegarich]

Hey

Is there any official help thread of turbo boards? I found some info there, some in other place and actually a bit confused because it's looks like different people saying different things. So maybe I can ask here.

1. I'm using 2180B in main board should I go with 2180B in Turbo board?
2. I removed 2 resistors on the main board near sidechain VCA, do I have to do the same on Turbo board?
3. There was a post that somebody put back those resistors when pluged a turbo board, so in the end what about those resisotrs?

thanks

 

[Harpo]

1+2: probably yes, except you want the left side sidechain operate different from the right side...
3: no idea what you're talking about. Could you post a link?

 

[main repairman]

Hi guys.
I just built my first GSSL compressor and I have a small problem.
I removed all the VCA chips (Harpo instruction - page 210) and measure the voltage with a DMM.

Input voltage NE5534 opamp pin6 1VAC (1V*22K/22K=1VAC)
Output voltage NE5532 pin7,1 0.55V (1V*15K/27K=0.55VAC), (0.55V*10K/10K=0.55VAC)
Input, output section is OK.

SIDECHAIN
Input level 0.1V sine ... (Harpo instruction - page 210)
Opamp TL074 pin7 0,14VAC (0.1*33K/47K=0.7VAC*2 (2x47K)=0.14VAC) this is OK.

And now I have a problem in this section:
TL074pin1 is 0.42VAC - is OK

about 0.1VAC at the junction 10K/20K/kathode 1N4148 ((Harpo instruction - page 210)
0.14VAC*20/20=0.14VAC at the juction - I have 0.8VAC :-(  Any idea?

and TL074pin14 0.14*1000K/20k=7V is a good result??


And TL074pin8 i have bad result other than Hapro instruction.

TL072pin1,8 is OK.

I need an oscilloscope to measure the correct results???

Any idea? Please help guys

 

[Harpo]

0.14VAC*20/20=0.14VAC at the juction - I have 0.8VAC :-(  Any idea?

Double check parts values, parts orientation and solder connections at the ratio switch as well as conducting pcb traces to/from and at center of ratio switch (using a magnifying glass might help to spot a short or a broken or lifted pcb trace).

and TL074pin14 0.14*1000K/20k=7V is a good result??

No. There is a lot more going on in this TL074-pin13 (current) node and FI this 1M resistor has other resistors in parallel and/or a subtracting current is injected with a working ratio switch ...

And TL074pin8 i have bad result other than Hapro instruction.

Followup from previous readings.
(probably my poor english but I will never instruct. That would kill the fun of DIY ...)

 

[main repairman]

0.14VAC*20/20=0.14VAC at the juction - I have 0.8VAC :-(  Any idea?

Double check parts values, parts orientation and solder connections at the ratio switch as well as conducting pcb traces to/from and at center of ratio switch (using a magnifying glass might help to spot a short or a broken or lifted pcb trace).

and TL074pin14 0.14*1000K/20k=7V is a good result??

No. There is a lot more going on in this TL074-pin13 (current) node and FI this 1M resistor has other resistors in parallel and/or a subtracting current is injected with a working ratio switch ...

And TL074pin8 i have bad result other than Hapro instruction.

Followup from previous readings.
(probably my poor english but I will never instruct. That would kill the fun of DIY ...)

Thank Harpo
This is my new measured voltage on the sidechain section.
Measured values ​​are ok or bad?

 

[Harpo]

This is my new measured voltage on the sidechain section.
Measured values ​​are ok or bad?

Dubious idea to disconnect the ratio switch. 
Are you sure your diodes are diodes for real, the TL074 is alife, you didn't connect any additional load to the +/-12V rails and your meter (with full batterie) can handle this frequency ?
Your 0.8VAC reading at halfwave rectifier output for the negative halfcycle at input is wrong.
Just running a simulation, with 0.14VAC 400Hz sine signal in (0.2Vpk, 0.4Vppk) it will swing to +0.2Vpk for the negative halfcycle at input and leave out (0V) the positive halfcycle, so still at 400Hz.
Keeping your 0.8VAC reading at this junction (by ignoring your voltage reading at TL074-pin1 and substituting the 20K fb with 160K), this inverted halfwave is fed to the summing node thru a 10K and the fullwave input signal thru a 20K, so for usual half of the first halfcycle cancels out and the other halfcycle adds up. With your disconnected ratio switch ensemble there is only the insane value 1M feedback resistor left, that would give (0.2Vpk*1M/20k*-1)=-10Vpk for the positive input halfcycle and opamp thankfully fails in the attempt to swing to ((0.2V*160K/20K)-0.2V)*1M/10K*-1)=-140Vpk because of voltage rail limits for the negative halfcycle. Your 2.8VAC is only the oscillating rest in a state of opamps try to better comit suicide.

 

[main repairman]

This is my new measured voltage on the sidechain section.
Measured values ​​are ok or bad?

Dubious idea to disconnect the ratio switch. 
Are you sure your diodes are diodes for real, the TL074 is alife, you didn't connect any additional load to the +/-12V rails and your meter (with full batterie) can handle this frequency ?
Your 0.8VAC reading at halfwave rectifier output for the negative halfcycle at input is wrong.
Just running a simulation, with 0.14VAC 400Hz sine signal in (0.2Vpk, 0.4Vppk) it will swing to +0.2Vpk for the negative halfcycle at input and leave out (0V) the positive halfcycle, so still at 400Hz.
Keeping your 0.8VAC reading at this junction (by ignoring your voltage reading at TL074-pin1 and substituting the 20K fb with 160K), this inverted halfwave is fed to the summing node thru a 10K and the fullwave input signal thru a 20K, so for usual half of the first halfcycle cancels out and the other halfcycle adds up. With your disconnected ratio switch ensemble there is only the insane value 1M feedback resistor left, that would give (0.2Vpk*1M/20k*-1)=-10Vpk for the positive input halfcycle and opamp thankfully fails in the attempt to swing to ((0.2V*160K/20K)-0.2V)*1M/10K*-1)=-140Vpk because of voltage rail limits for the negative halfcycle. Your 2.8VAC is only the oscillating rest in a state of opamps try to better comit suicide.

Harpo any idea to repair?
My problem is around TL and 074/pin2.3.1 and 12.13.14
Diode, resistores, OPAMP is correct.

Input signal set by your instukcii.
You write:
Sidechain: Reduce input level to-20dBV (0.1V RMS, 0.14V pk).
Why now set 0.14VAC RMS, 0.2VPK?

I connect section ratio and measured values ​​are different.
Connection point A, B voltage at pin 14 TL 074 is wrong.
How to calculate the voltage on the TL 074/pin14 ratio  2:1 = 4.8VAC

 

[main repairman]

This is my new measured voltage on the sidechain section.
Measured values ​​are ok or bad?

Dubious idea to disconnect the ratio switch. 
Are you sure your diodes are diodes for real, the TL074 is alife, you didn't connect any additional load to the +/-12V rails and your meter (with full batterie) can handle this frequency ?
Your 0.8VAC reading at halfwave rectifier output for the negative halfcycle at input is wrong.
Just running a simulation, with 0.14VAC 400Hz sine signal in (0.2Vpk, 0.4Vppk) it will swing to +0.2Vpk for the negative halfcycle at input and leave out (0V) the positive halfcycle, so still at 400Hz.
Keeping your 0.8VAC reading at this junction (by ignoring your voltage reading at TL074-pin1 and substituting the 20K fb with 160K), this inverted halfwave is fed to the summing node thru a 10K and the fullwave input signal thru a 20K, so for usual half of the first halfcycle cancels out and the other halfcycle adds up. With your disconnected ratio switch ensemble there is only the insane value 1M feedback resistor left, that would give (0.2Vpk*1M/20k*-1)=-10Vpk for the positive input halfcycle and opamp thankfully fails in the attempt to swing to ((0.2V*160K/20K)-0.2V)*1M/10K*-1)=-140Vpk because of voltage rail limits for the negative halfcycle. Your 2.8VAC is only the oscillating rest in a state of opamps try to better comit suicide.

 

[main repairman]

This is my new measured voltage on the sidechain section.
Measured values ​​are ok or bad?

Dubious idea to disconnect the ratio switch. 
Are you sure your diodes are diodes for real, the TL074 is alife, you didn't connect any additional load to the +/-12V rails and your meter (with full batterie) can handle this frequency ?
Your 0.8VAC reading at halfwave rectifier output for the negative halfcycle at input is wrong.
Just running a simulation, with 0.14VAC 400Hz sine signal in (0.2Vpk, 0.4Vppk) it will swing to +0.2Vpk for the negative halfcycle at input and leave out (0V) the positive halfcycle, so still at 400Hz.
Keeping your 0.8VAC reading at this junction (by ignoring your voltage reading at TL074-pin1 and substituting the 20K fb with 160K), this inverted halfwave is fed to the summing node thru a 10K and the fullwave input signal thru a 20K, so for usual half of the first halfcycle cancels out and the other halfcycle adds up. With your disconnected ratio switch ensemble there is only the insane value 1M feedback resistor left, that would give (0.2Vpk*1M/20k*-1)=-10Vpk for the positive input halfcycle and opamp thankfully fails in the attempt to swing to ((0.2V*160K/20K)-0.2V)*1M/10K*-1)=-140Vpk because of voltage rail limits for the negative halfcycle. Your 2.8VAC is only the oscillating rest in a state of opamps try to better comit suicide.

Harpo any idea to repair?
My problem is around TL and 074/pin2.3.1 and 12.13.14
Diode, resistores, OPAMP is correct.

Input signal set by your instukcii.
You write:
Sidechain: Reduce input level to-20dBV (0.1V RMS, 0.14V pk).
Why now set 0.14VAC RMS, 0.2VPK?

I connect section ratio and measured values ​​are different.
Connection point A, B voltage at pin 14 TL 074 is wrong.
How to calculate the voltage on the TL 074/pin14 ratio  2:1 = 4.8VAC

 

[Harpo]

Harpo any idea to repair?
My problem is around TL and 074/pin2.3.1 and 12.13.14
Diode, resistores, OPAMP is correct.

obviously not. I already told you my ideas to fix it.

You write:
Sidechain: Reduce input level to-20dBV (0.1V RMS, 0.14V pk).
Why now set 0.14VAC RMS, 0.2VPK?

With 0.1VAC signal in, the summed level of this L/R corelated signal at TL074-pin7 is 2* 0.1VAC*33K/47K=0.14VAC at rectifier input. Your measurement came to the same result.

I connect section ratio and measured values are different.

hopefully.

Connection point A, B voltage at pin 14 TL 074 is wrong.
How to calculate the voltage on the TL 074/pin14 ratio  2:1 = 4.8VAC

Kirchhoffs law might help you.
hint:Calculating the other ratio settings instead of the non-linear 2:1 setting probably will take some hours less when you have to ask. I'd run a sim and use a scope instead to take measurements inside a rectifiers feedback loop. With sidechain-VCA back in circuit, varying pre-rectifiers input voltage depending on generated control voltage, these readings will be different anyway.