My studio Mic compressor build

[11-90-an]

>I will recheck when I get home.  But from what I can see in the schematic, IC1, which is a TL072, has Pin 4 as + and Pin 8 as - in my schematic.  If google is not failing me, it should be Pin 4 - and Pin 8 +, right?


yes

>Now, I should point out that IC1 did not burn up, smoke, or even get hot (I touched all the other ICs when IC2 burned up), and when I took IC2 out, I connected the circuit for much longer periods, and nothing burned up, so you could be right...  But I do think it is backwards.

happy accident? 

[jfrabat]

Lots of things to update...

First of all, this was my solution to the swapped polarity:



I did this on both ICs:





Not pretty, but it works...

So now to the first test:

I then proceeded to adjust the 0V trimpot (the closest to 0 I got was ~15mV, so I left it there), played a little with the click trimpot (did not notice much of a difference), as well as swap the pots (and flipped both upper switches around, but have not yet swapped switches).  I also used an audio source to test.

Notice the sound at the begining...  Sounds like a helicopter!  What the heck is that?

I also noticed an issue with the limiter making noise.  I made it drastic on purpose, so you can here it, but this is what I mean:

Sounds like wind blowing in a microphone (it is worse in person)

At least we are making progress, right?

OVERALL ISSUES:
1. Compressor is a bit harsh, but you can play with adjustments to get around that (somewhat at least).
2. Noise gate not working as intended.
3. Limiter noise

So, intention is to start trying to fix the Noise Gate.  Where do I begin?

[jfrabat]

I just noticed another unexpected behavior; if I turn on the Noise Gate, and then flick it back off, the compressor stays in the Noise Gate activated mode (I have to turn off the entire thing and flick it back on for it to behave normally).  If you have any ideas on how to fix this, please let me know!

[jfrabat]

Notice the sound at the begining...  Sounds like a helicopter!  What the heck is that?

Well, about this one...  ehmmm...  well...  ehmmm...  so, if the source material has that, the compressor should reproduce it, right?    

Had not noticed the source had that sound before.  Sorry for posting that one!

[11-90-an]

I just noticed another unexpected behavior; if I turn on the Noise Gate, and then flick it back off, the compressor stays in the Noise Gate activated mode (I have to turn off the entire thing and flick it back on for it to behave normally).  If you have any ideas on how to fix this, please let me know!

Could your switch be broken...? 

[jfrabat]

I just noticed another unexpected behavior; if I turn on the Noise Gate, and then flick it back off, the compressor stays in the Noise Gate activated mode (I have to turn off the entire thing and flick it back on for it to behave normally).  If you have any ideas on how to fix this, please let me know!

Could your switch be broken...? 

Good point...  Tried it and it was the switch.  Thanks for that one!  I also put the switches in the right position, so that's 3 things crossed out from the to fix list! 

But the gate is still not working properly.  Depth pot is working as intended, but the gate simple is not opening for some reason.  Regardless of where the attack, release and trigger pots are set at.

[11-90-an]

oh well...

have you checked all your component values and transistor pinouts? especially the ones in here...

[jfrabat]

Sorry it took so long to reply; got ahead of myself, and worked a bit on my tube preamp (needed to clear some space on the workbench!)...  It's working, but has some bugs to fix still (no Phantom Power and pots are acting up and not working properly), but I will get to that later.  Will probably also finish the other preamp before deep diving into this one.  I need to get projects off the bench to make room to troubleshoot!  Unfortunately, ALL my projects need troubleshooting!

But just a teaser, here is how my stuff is looking so far:



If only it worked as good as it looks, we would be in business!  LOL!  At least the EQ is working flawlesly!

Anyway, back to the matter at hand.  This are my voltages at the gate side of the circuit (with no signal going through):

IC5:

1: -34.8mV
2: 1.3mV
3: 0V
4: 15.34V
5: 143.7mV
6: -13.39V
7: 14.75V
8: -13.66V
9: 4.7mV
10: -12.15V
11: -14.97V
12: -3.8mV
13: 2.8mV
14: 3.7mV

IC6:
1: 10.42V
2: 1.1mV
3: 0.1mV
4: 0~132mV
5: -0.9mV
6: 3.9~9.1mV
7: 6mV
8: 84mV
9: 0.9mV
10: 0.1mV
11: -14.97V
12: 14.7~13.7V
13: 14~14.26V
14: 14.26V

Q7:
E: 15.26V
B: 14.74V
C: -14.98V

Q8:
E: 15.33V
B: 14.75V
C: -12.67

Q9:
E: .720V
B: 100.8mV
C: -5.76V

oh well...

have you checked all your component values and transistor pinouts? especially the ones in here...

Well, I have the following readings (and, yes, I know you should not read resistors wile connected to the circuit, but it is all I can do without taking apart the whole thing and starting to desolder everything!):

R87: 0.964K
R89: 0.987K
R90: 11.86K

R86 and R88 I get a bunch of nonsense, so I checked color codes (not a guarantee that the resistors will be right, but the best I can do!).  They seem OK.  Diode is also in the right direction.  Q9 voltages are above.

[jfrabat]

Sorry it took so long to reply; got ahead of myself, and worked a bit on my tube preamp (needed to clear some space on the workbench!)...  It's working, but has some bugs to fix still (no Phantom Power and pots are acting up and not working properly), but I will get to that later.  Will probably also finish the other preamp before deep diving into this one.  I need to get projects off the bench to make room to troubleshoot!  Unfortunately, ALL my projects need troubleshooting!

Why do I always say stuff like this only to end up contradicting myself?  Well, finished the tube preamp.  Issue was a MAJOR design flaw in the circuit, which caused it to not work when you plug one single IC backwards.   

So, any takers on the compressor (noise gate, really) issues?

[11-90-an]

uh... my brain cannot comprehend voltage readings yet...   

Remember those strange markings/waveshapes on the original schem? If you have an oscilloscope, maybe try seeing what shapes you get! 

[jfrabat]

uh... my brain cannot comprehend voltage readings yet...   

Remember those strange markings/waveshapes on the original schem? If you have an oscilloscope, maybe try seeing what shapes you get! 

I SORT OF do have an oscilloscope...  I have one of these:



They are an oscilloscope kit.  Works OK, but the sensitivity is not the best.  I can track the input signal and output signal, but once they are reduced, they are gone.  For example, I used my tone generator at ~400Hz.  I could easily read the signal input and output.  But once the signal gets past R7, I lose it.

But, studying the diagram, I think I figured out the issue.  In my build, I have yet to solder SK10 (the jack for the external trigger for the gate, which is tip switched).  Since that jack is switched, it is sending the signal from the input to the gate to control the gate.  But since the switch does not exist, and it is not receiving any external signal either, the gate is simply saying that there is no signal, hence the gate should remain closed.

Is my logic correct?

[11-90-an]

I guess so...

Try doing it and see if it works...

[jfrabat]

Tried it out; soldered up the tip and the tip switch wires.  This is what we have:

Before (wires disconnected):

Now (wires connected):

As far as signal, I can now use the scope to see signals in the gate (as there is now a signal!). Well, at least I can trace it up to D5; I lose it at the cathode of D5.  Could it be an issue of D5?  The signal should not be modified by a diode, right?

Anyway, this is the scope at the input signal:



This is the signal at IC5 pin 2:



And this is the signal at IC5 pin 12:



Signal at IC5 pin 13:



Signal at IC5 pin 14:



And the signal at IC5 pin 8:



This last one is highly dependent on where the trigger pot is set (as it should be).  And it certainly looks like the square version of the image of the waveform on the original schematic...

[jfrabat]

Jack is in, and the issues are the same as when the wires were connected...

I have been thinking about this whole gate issue all day, and the whole signal dissapearing at D4 is baffling me.  I mean, I know there is a grounded cap there, but should not even part of the signal be reaching IC5b and d?.

I also checked Pin 10 of IC5, and I get this:



Not unlike the image of how the wave should look like at IC6D (I cannot see the signal past R71; it may be the scope is not sensitive enough).

Now, if I understand the article text correctly, the fully rectified signal (as in all the waveforms are to the negative side) at IC5 Pin 8 is compared to a reference voltage set by the trigger pot.  I know the rectified signal is at IC5 Pin 8, but how do I figure the trigger voltage?  Because of how it is connected, it should be be anywhere between V- and GND (which it is, in VDC), depending on position of the pot.  But isn't the audio signal AC?  If I measure DC voltage at pin 8 of IC5, it's at 14.4V, which is obviously going to be more than anything between V- and GND...  So the gate will always open, right?  Which is kind of what is happening, but not sure my reasoning is correct...

After all, if that signal is supposed to be compared to the trigger voltage, should we not be talking AC?  And I get no voltage at Pin 8 IC5 in AC...  Any ideas?  Also, if you can instruct me (so as to learn and not ask the same thing in the future), would be greatly appreciated!

[jfrabat]

Bump; anyone want to take a stab at this one?

[Rob Strand]

Bump; anyone want to take a stab at this one?

It's going to take a while for me to read over the thread.

Just a general comment:  When you follow the signal path through the circuit and you suddenly lose the signal, that doesn't always mean there is a problem.    For example when you lose the signal on one side of R7 that's OK.   The reason is pin 2 of IC2A is a "virtual ground" point.   It looks like ground because the feedback around the opamp IC2A makes it looks like ground.   In general if an opamp's +in input is grounded then the opamp's -in input will be a virtual ground.    In cases like this, skip to the opamp output.  If there's nothing at the opamp output then there is probably something wrong.    If you do see signal at the opamp output then it's OK to move on further.

IC5A pin 2 is another virtual ground point.

As far as tracing along the signal chain.  I'd start probing the signal at C8.   If you don't get a sine-wave there then the detection circuit for the noise gate cannot work.

The signal at IC5D pin14 should be a rectified signal with "humps".  See Fig 7.2.4, except you humps will go downwards and negative,
https://wiki.analog.com/university/courses/electronics/text/chapter-7

The signal at IC5C pin 8 should go on and off.   As you adjust the pot RV6, the width of the on and off signal should change.  If you can't see anything perhaps try reducing or increasing the audio level from the signal generator.

I need to catch up reading the thread.

[jfrabat]

Bump; anyone want to take a stab at this one?

It's going to take a while for me to read over the thread.

Just a general comment:  When you follow the signal path through the circuit and you suddenly lose the signal, that doesn't always mean there is a problem.    For example when you lose the signal on one side of R7 that's OK.   The reason is pin 2 of IC2A is a "virtual ground" point.   It looks like ground because the feedback around the opamp IC2A makes it looks like ground.   In general if an opamp's +in input is grounded then the opamp's -in input will be a virtual ground.    In cases like this, skip to the opamp output.  If there's nothing at the opamp output then there is probably something wrong.    If you do see signal at the opamp output then it's OK to move on further.

IC5A pin 2 is another virtual ground point.

As far as tracing along the signal chain.  I'd start probing the signal at C8.   If you don't get a sine-wave there then the detection circuit for the noise gate cannot work.

The signal at IC5D pin14 should be a rectified signal with "humps".  See Fig 7.2.4, except you humps will go downwards and negative,
https://wiki.analog.com/university/courses/electronics/text/chapter-7

The signal at IC5C pin 8 should go on and off.   As you adjust the pot RV6, the width of the on and off signal should change.  If you can't see anything perhaps try reducing or increasing the audio level from the signal generator.

I need to catch up reading the thread.

Thanks Rob.  Summary is that the signal is making it down C8 all the way through IC5c.  After it comes out of there, it goes into D7, but does not make it across.  Not surprised, because a cap there is taking it to ground.  But IC5b and IC6d are getting no signal at all.  On IC6a I get nothing on pins 1 and 3, but I get some noise (-30mV to 60mV).  The last two sentences are regardless of Depth or Trigger setting.

[Rob Strand]

Thanks Rob.  Summary is that the signal is making it down C8 all the way through IC5c.

OK, that's a start.

After it comes out of there, it goes into D7, but does not make it across.  Not surprised, because a cap there is taking it to ground.  But IC5b and IC6d are getting no signal at all.  On IC6a I get nothing on pins 1 and 3, but I get some noise (-30mV to 60mV).  The last two sentences are regardless of Depth or Trigger setting.

So D5 here, yes?

Once you get to the cathode (C8 side) of D5 the signals are DC.  So you no longer expect to see the signal getting through.

What you expect is the DC value to change depending on if the input signal is present or not.   So to test this area of the circuit you need  to measure the voltages with no signal from the signal generator, then apply the signal from the signal generator and re-measure.   With the signal generator present you might need to play with the controls to get the voltage to change.

So you would expect these.

Input           D5-cathode/C18      IC5B pin 7        IC6d pin 14

No Signal   -VE                            +VE                  +VE
Signal         +VE                           -VE                    -VE

Maybe a multimeter set to DC is better than an oscilloscope for this.

FYI,  when I look at the circuit diagram, the signal on pin5B pin 7 I've show here is opposite to the waveforms in the original article.  I hope I got it right.    Also the original article shows "ON" and "OFF",  I think "ON" means signal present  and gate "OFF" (not noise gate "ON").

Whatever the case, don't worry about the ON/OFF thing - the most important feature is the DC voltage changes with and without the signal present.

Don't forget you might need to play with the controls to get it to work with the signal present.

[jfrabat]

After it comes out of there, it goes into D7, but does not make it across.  Not surprised, because a cap there is taking it to ground.  But IC5b and IC6d are getting no signal at all.  On IC6a I get nothing on pins 1 and 3, but I get some noise (-30mV to 60mV).  The last two sentences are regardless of Depth or Trigger setting.

So D5 here, yes?

Yes, sorry, D5, not D7... 

Once you get to the cathode (C8 side) of D5 the signals are DC.  So you no longer expect to see the signal getting through.

What you expect is the DC value to change depending on if the input signal is present or not.   So to test this area of the circuit you need  to measure the voltages with no signal from the signal generator, then apply the signal from the signal generator and re-measure.   With the signal generator present you might need to play with the controls to get the voltage to change.

OK, got it.  Makes sense; D5 will limit the ability to be AC because only positives can flow through...

Maybe a multimeter set to DC is better than an oscilloscope for this.

OK, going back to the MM then!

So you would expect these.

Input           D5-cathode/C18      IC5B pin 7        IC6d pin 14

No Signal   -VE                            +VE                  +VE
Signal         +VE                           -VE                    -VE

Whatever the case, don't worry about the ON/OFF thing - the most important feature is the DC voltage changes with and without the signal present.

Don't forget you might need to play with the controls to get it to work with the signal present.

OK, here are my readings (obviously there are issues here!):

V+: 15.35V
V-: -14.94V

With no signal:

D5 (cathode): 14.11V
IC5 Pin 7: -0.508 (something is obviously wrong here!)
IC6 Pin 14: 3.7mV (also something wrong here!)

With signal:

D5 (cathode): 14.10V
IC5 Pin 7: -0.506 (regardless of DEPTH and TRIGGER pot positions; obviously wrong!)
IC6 Pin 14: -52.4mV (regardless of DEPTH and TRIGGER pot positions; obviously wrong!)

I also tried the readings with the signal with the switch for the bypass both on and off.  No change.  So obviously something is not quite working as it should be here.

[Rob Strand]

With no signal:


With signal:

I also tried the readings with the signal with the switch for the bypass both on and off.  No change.  So obviously something is not quite working as it should be here.

Man, that's weird.

Well we need to fix the voltages around D5 first.  If the voltages are bad there the whole circuit following won't have any hope of working.

One oversight on my part is your multimeter might be 1M ohm input impedance and not 10M ohm.   That's going to mess with the voltage on cathode side of D5.  Even then I can't make sense of the measurement because you get a positive value.

When you measured the D5 cathode voltage did you have the multimeter -Ve lead connected to ground and the multimeter +Ve lead connect to D5.?      Most of the time you want the multimeter -ve leads connected to ground so we can make sense of the sign of the measurement.

The next tests would be,
- measure the DC voltage on the anode side of D5 with an without signal.
- maybe recheck the sign of the voltage on the cathode side of D5 with and without signal
- measure the voltage on pin 6 of IC5b with and without signal.