All things G1176 - the new "repost" thread.

[schrobbelbop]

can anybody help me out wiring up my Rev J board with hariball trafo's/clearostat?

As far as I have worked out:

connect input trafo/input pot to "input pot"
output trafo to "S" and "F" connections op pcb trafo in. and X / Y Audio out (also going to VU)

wich resistors do i have to place near the output trafo connections?

Can i Skip the bypass switch wiring, because i dont have a switchable attack pot?

Can i connect my 12V lamp to the -10 / ground rails?

thanks!

 

[gswan]

Can i Skip the bypass switch wiring, because i dont have a switchable attack pot?

Probably not a good idea. You will need to disable gain reduction at some stage. If you don't have a switch pot then there are many other ways to do it using other switches.

Can i connect my 12V lamp to the -10 / ground rails?

Don't do this, the -10V rail is the bias rail for the FETs. If you inject noise into it by connecting it to other things then it will disrupt the FETs. Best idea is to connect two lamps in series and run them off one of the AC windings. Since the lamps are independent of the rest of the circuit they do not need to be referenced to GND.

 

[schrobbelbop]

such a smart idea with those lamps

then i'm gonna wire the bypass to the +8 switch.

almost done...

 

[schrobbelbop]

Ok i've tried my g1176 REV J.

Distorted output, Vu meter works.
No compression

Where to begin ?

 

[gswan]

Ok i've tried my g1176 REV J.

Distorted output, Vu meter works.
No compression

Where to begin ?

I presume you have performed all the setup tests and calibration procedures to verify the operation of each stage? If not, that's where you should begin.

 

[schrobbelbop]

stupid me i forgot R36... now it works! up to calibrating

 

[schrobbelbop]

Just a few questions left:

R78 (1.1K) is getting very hot, i've put 2* 2.2K 1/2W parralel. should I get 1 1.1K 1W?

Is it that a 20 ratio needs more input than a 4 ratio to start compressing?

 

[rooster21]

So, I think I found out why I wasn't getting any compression, the audio was never getting to the detection circuit. I probed around to see where the signal stopped, and found that they entered cr3 and cr4 but never came out the other end, tried replacing them and the same thing happened. Later on noticed that upon turning on the unit cr3 and 4 would get ~1.054V and gradually would drain to 0, 1mv at a time.

Now I'm not sure on how to attack this situation, how you you find what is causing this drain?

I rebuilt the PSU, and nothing changed so I'm pretty much stuck here, can anyone give any suggestions for this situation?

Thanks.
GDS

 

[gswan]

So, I think I found out why I wasn't getting any compression, the audio was never getting to the detection circuit. I probed around to see where the signal stopped, and found that they entered cr3 and cr4 but never came out the other end, tried replacing them and the same thing happened. Later on noticed that upon turning on the unit cr3 and 4 would get ~1.054V and gradually would drain to 0, 1mv at a time.

Audio is not supposed to come out of CR3 and CR4, this should be a control voltage offset by a proportion of Qbias and dependent on the audio fed to the GR amp. You need to check further upstream.

First, without any audio input, check the bias voltages around Q12-Q15. Alse re-check your Qbias calibration, since this depends on the operation of the GR circuit.

 

[rooster21]

So, I think I found out why I wasn't getting any compression, the audio was never getting to the detection circuit. I probed around to see where the signal stopped, and found that they entered cr3 and cr4 but never came out the other end, tried replacing them and the same thing happened. Later on noticed that upon turning on the unit cr3 and 4 would get ~1.054V and gradually would drain to 0, 1mv at a time.

Audio is not supposed to come out of CR3 and CR4, this should be a control voltage offset by a proportion of Qbias and dependent on the audio fed to the GR amp. You need to check further upstream.

First, without any audio input, check the bias voltages around Q12-Q15. Alse re-check your Qbias calibration, since this depends on the operation of the GR circuit.

          B     C          E

Q12 4.39   14.95 3.84
Q13 14.92   29.18 14.32
Q14 3.39   16.29 2.83
Q15 16.29  29.17 15.74

I measured all the test points on gyraf's schem, and they're all pretty close in that area, on average they're ~20mv less than gyraf's results.

 

[gswan]

Your DC bias voltages look good.
Now check that the GR disable switch is connected correctly (ie that audio is in fact entering at point 22 when GR is enabled). Also re-check all the resistors on the ratio switches, since these affect both the amount of audio entering the GR circuit and the offset applied prior to rectification (a kind of threshold).

 

[rooster21]

Your DC bias voltages look good.
Now check that the GR disable switch is connected correctly (ie that audio is in fact entering at point 22 when GR is enabled). Also re-check all the resistors on the ratio switches, since these affect both the amount of audio entering the GR circuit and the offset applied prior to rectification (a kind of threshold).

yep, the resistors are all correct and audio passes when GR is enabled.

 

[schrobbelbop]

Is it normal the 1.1K 1W resistor gets hot?

My meter reads -0.5dB when its just turned on
when on for 'bout 15 min it goes to +0.5dB.
I've single turn cheapass trimmpots, this might be the issue?

Love it's sound Thanks Gyraf/Hairball/Mnats, great work!

 

[gswan]

yep, the resistors are all correct and audio passes when GR is enabled.

Have you performed the gain checks of each stage and calibration of the gain reduction (FET Qbias)?

 

[rooster21]

yes I have, but I would like to redo it just to be sure. Some clarification on the procedure would be nice.

Measurements:
1. GR bias wound out
2. Release fully CW
3. Attack off (GR disabled)
4. Apply a 0.775Vrms 1kHz sine wave to the input.
5. Adjust the input and output controls so that the output of each stage is not clipping
6. Measure the input and output AC voltage values

these where the instructions i followed but I would like to verify the folowing:

1. GR wound out meaning ccw (least resistance) or cw (most resistance), I'm assuming least.
4. .775Vrms, this means roughly .39Vrms on pin 2 and 3 respectively, right?
5. should the input and output be adjusted to the maximum before audible distortion?

I'm sure most of this should be obvious, but i would rather ask the post false results.

Thanks.

 

[skal1]

OK  i have  had this 1176 j rev done for about a year now and i still can not understand how to calibrate this thing properly javascript:void(0);, if any one could walk me through the setup i would appreciate it a lot  thanks , as i am writing this all i can do is wonder where i am going wrong with the calibration steps.

well i am sitting in front of it now.

thanks

skal1

 

[gswan]

yes I have, but I would like to redo it just to be sure. Some clarification on the procedure would be nice.

Measurements:
1. GR bias wound out
2. Release fully CW
3. Attack off (GR disabled)
4. Apply a 0.775Vrms 1kHz sine wave to the input.
5. Adjust the input and output controls so that the output of each stage is not clipping
6. Measure the input and output AC voltage values

these where the instructions i followed but I would like to verify the folowing:

1. GR wound out meaning ccw (least resistance) or cw (most resistance), I'm assuming least.
4. .775Vrms, this means roughly .39Vrms on pin 2 and 3 respectively, right?
5. should the input and output be adjusted to the maximum before audible distortion?

For the gain tests you need to have the GR FET disabled, ie the Qbias trimmer set to where the FET is not conducting (most -Ve value on the FET Vgs).

Apply a signal to the input (the actual value is not important, somewhere around 100mV to 1V RMS should do). It's a balanced input so the signal goes across pins 3 and 3 of the XLR connector.

Now measure the signal value at the top of the output pot and wind the input pot up until you get a few Vrms there. Measure the input to this stage (input pot wiper) and calculate the gain. Do the same thing for the output section.

 

[gswan]

well i am sitting in front of it now.

And now what do you plan to do?

 

[skal1]

hi gswan


yeah i had another go at it  and i had to change the meter i was using because it did not have a  1db ref , went through the calibration again and now the attack is working as it should be , those bf245 have a very tight working window might have to change the 1 turn pot for a multi.

well it seems to work ok ,but then again what do i know , i think i will try  the transformer 1:2 mod for the oep , just a quick ?, is the oep tx 1:1 at the moment in the rev J  do i have to reverse it to get the 1:2 step .


thanks

skal1

 

[gswan]

yeah i had another go at it  and i had to change the meter i was using because it did not have a  1db ref , went through the calibration again and now the attack is working as it should be , those bf245 have a very tight working window might have to change the 1 turn pot for a multi.

Yes, the Qbias is very sensitive. If you plot the FET characteristics first it will give you an idea of where the optimum point would be and how sensistive it is. A 25 turn pot is a good idea for this adjustment.

Bear in mind that there will be some thermal dependancy in the FET adjustment, so don't calibrate it in freezing conditions and expect it to be the same on a hot day!