My studio Mic compressor build

[jfrabat]

So that got me thinking  maybe check these points
- IC5 pins 1, 2, 3
- IC5 pins 12, 13, 14
- D3 cathode
- IC5 pin 9

are not shorting to other pins on IC5.   I'd be really looking at pins 5, 6,7 if IC5 first.

OK, no 2 pins in IC5 have continuity between them (I checked all of them).  Same goes for D3 Cathode

Using the DMM continuity tester, check there is no shorts to ground on any of these pins,
- IC5 pins 1, 2, 3
- IC5 pins 12, 13, 14
- D3 cathode
- IC5 pin 9

An also check this pin *is* connecting to ground,
- IC5 pins 3

Only pin 3 has continuity to GND.

1) I have a feeling you need to re-do the test with the 100pF cap soldered between IC5 pin 1 and pin 2.

OK, so connect increase to 100pF this time around between pins 1 and 2?  I had a 10pF last time around, so just checking.

2) Another thing you can try is soldering a 100nF cap across the power pins of IC5, pins 11 and 4.    If that's no good maybe one 100nF from pin 11 to ground then another from pin 4 to ground.

How do I know if it is no good or not?  If it smokes it is no good? (Not being sarcastic; I really do not know!)

It's starting to be a tough one to work out.

I will let you know in on a little secret...  They are ALL tough to work out for me!  LOL!

By the way, the second board is all populated (except for the IC's and the off board stuff).  Off board stuff tends to take some time to get done, but I hope to have it done by this weekend.  May even correct the LED order so that the LEDs light to the left and not the right as I have it now (although I kind of like having the limiter in the middle between the gate and compressor, convention is that compressors move backwards).

[Rob Strand]

OK, no 2 pins in IC5 have continuity between them (I checked all of them).  Same goes for D3 Cathode
Only pin 3 has continuity to GND.

OK that's good.   But it means there's no obvious faults to cause problem.

OK, so connect increase to 100pF this time around between pins 1 and 2?  I had a 10pF last time around, so just checking.

Yes.   Try 100pF.

How do I know if it is no good or not?  If it smokes it is no good? (Not being sarcastic; I really do not know!

If you check voltages:

Do the changes and check the DC voltage on pin 14 IC 5, maybe pins  1, 12, 13  also.   No signal is fine.

And they match-up with my table in Reply #43.     Most should be zero.  The main one to check is IC 5 pin 14.   If that's -12V and not close to zero then the problem is still there.

I will let you know in on a little secret...  They are ALL tough to work out for me!  LOL!

Debugging is always hard.   If you are fixing the same brand and model of TV everyday you can come up some quick fix recipes but when you build something for the first time there's no "fix it" recipe at all.

By the way, the second board is all populated (except for the IC's and the off board stuff).  Off board stuff tends to take some time to get done, but I hope to have it done by this weekend.   May even correct the LED order so that the LEDs light to the left and not the right as I have it now (although  kind of like having the limiter in the middle between the gate and compressor, convention is that compressors move backwards).

Good luck with it all. 

[jfrabat]

yea that back-trace would be useful...

forgive me for my ignorance, but is the depth pot supposed to control the LED indicator?

Sorry, I missed your answer.  In short, yes.  If you reduce the DEPTH, the gate starts opening and letting some sound through, and the LED is indicating this (by changing color).  At least that would be the logical conclusion for me!

[Rob Strand]

I've been trying to make sense of your measurements.
Take the measurement from Reply #66 with no input signal and no100K to +15V or -15V.

If we have no signal the DC levels should just be constant.  Nothing is changing.

IC5 Pin 1:
No Signal: +2.19V

D3 Anode:
No Signal: +2.65V

D4 Cathode:
No Signal: +2.65V

All these are the same voltage, but as you said they fluctuate a bit.
If the DC voltages are constant then pin 1 is somewhere in the middle of those, say +2.4V.

D3 Cathode:
No Signal: +5.59V

D4 Anode:
No Signal: -5.21V

However if the DC voltage is constant, how can  the DC voltage on D3 cathode be greater than pin 1?

If doesn't make sense that the DC voltage voltage on D4 Anode be negative.   With no input signal it should be zero.
The voltage on pin 1 of the opamp is positive so it can't produce a negative voltage on D4's Anode.

If I5a was oscillating then the diodes  would rectify the signal from pin 1.   The output would swing +/- 13V.
In that case we would see a negative voltage on D4's anode and a positive voltage on D3's cathode.
Each diode would half-wave rectify the output from pin of the opamp, If we lose 0.6V due to the diodes
the output would swing +/- (13-0.6) = +/-12.4V.   

If the opamp is oscillating as a sine-wave:
When we measure the half-wave rectified signal with a multimeter the measurement will be 0.32 times the peak, so 0.32 * 12.4 = 4V.

If the opamp is oscillating as a square-wave:  When we measure the half-wave rectified signal will be 0.5 times the peak, so 0.5 * 12.4V = 6.3V

So we might expect +4V to +6.3V on D3's cathode and -4V to -6.3V on D4's anode.
We are seeing +5.6V and -5.2V, very suspicious as the measured values are in the middle of what we expect to see if IC5a is oscillating.

IC5 Pin 2:
No Signal: +1.4mV
Signal: +1.4mV

Pin 2 near 0V means IC5a is working.


There's another thing we can check.

IC5 Pin 14:
No Signal: -12.94V

IC5 Pin 13:
No Signal: -4.25V
Signal: -4.82V

D3 Cathode:
No Signal: +5.59V
Signal: +3.69V

We expect the current through R67 and R68 be be equal (*assuming no shorts* ---> but we haven't measured any shorts).  From ohms law, since R67 and R68 are the same value that means we should be seeing the same voltage across each.

R67:   Vdrop = (5.59 - (-4.25))     =  9.84V
R68:   Vdrop = (-4.25 - (-12.94))  =  8.69V

That's close but not quite as close as I would expect.

[jfrabat]

OK, put the 100pF cap in:



Put the IC back in.  This is what I get (no signal):

IC5 Pin 9: -12.5V
IC5 Pin 10: 140mV to -11.83V (depending on RV6)

IC5 Pin 12: -5.67V
IC5 Pin 13: -4.11V
IC5 Pin 14: -12.65V

I also think the bypass switch may be backwards, so I tried with the switch in the opposite direction just to be sure, but it makes no difference.  Tried it also with another new IC, but no change.
With no IC, pins 12, 13 and 14 have about 1.1mV each.  Where the heck is that V- coming from???  AAAAHHHH!!!!  I even thought maybe there was something making a connection only when the legs of the IC were in the socket, so I measured with a lead inserted in the socket, but no change...

OK, looking at the schematic, pin 9 and pin 14 have R68 and 69 touching them.  But without the IC in, both are in the single digit mV range, so the -12V is not coming from there.  I don't get it!

[Rob Strand]

Put the IC back in.  This is what I get (no signal):

IC5 Pin 9: -12.5V
IC5 Pin 10: 140mV to -11.83V (depending on RV6)

IC5 Pin 12: -5.67V
IC5 Pin 13: -4.11V
IC5 Pin 14: -12.65V

I also think the bypass switch may be backwards, so I tried with the switch in the opposite direction just to be sure, but it makes no difference.  Tried it also with another new IC, but no change.
With no IC, pins 12, 13 and 14 have about 1.1mV each.  Where the heck is that V- coming from???  AAAAHHHH!!!!  I even thought maybe there was something making a connection only when the legs of the IC were in the socket, so I measured with a lead inserted in the socket, but no change...

OK, looking at the schematic, pin 9 and pin 14 have R68 and 69 touching them.  But without the IC in, both are in the single digit mV range, so the -12V is not coming from there.  I don't get it!

I can understand *if* the circuit is going nuts then you could get weird negative voltages.
However, I can't make any sense beyond that unless IC5 is actually oscillating!

As far as a fix goes I think leave the 100pF in and try the 100nF caps on the power rails.   The ground points on the 100nF caps are best placed near IC5.   The 100pF and 100nF cap thing is a common fix for oscillations.

It's very tempting to short out both diodes and see if at least the DC voltages come back to normal.   You might be best to do that *after* you put in the 100nF caps.

[jfrabat]

I've been trying to make sense of you measurements.
Take the measurement from Reply #66 with no input signal and not 100K to +15V or -15V.

If we have no signal the DC levels should just be constant.  Nothing is changing.

OK, just to be sure, I am retaking all the readings here:

RV6 Pin 2 @ -12.04

IC5 Pin 1:
No Signal: Oscillating, but hovers at around 0.4V

D3 Anode:
No Signal: Oscillating between 0.69V and 0.82V

D4 Cathode:
No Signal: Oscillating between 0.59V and 0.72V

D3 Cathode:
No Signal: +3.94V

D4 Anode:
No Signal: -6V

However if the DC voltage is constant, how can  the DC voltage on D3 cathode be greater than pin 1?

If doesn't make sense that the DC voltage voltage on D4 Anode be negative.   With no input signal it should be zero.
The voltage on pin 1 of the opamp is positive so it can't produce a negative voltage on D4's Anode.

If I5a was oscillating then the diodes  would rectify the signal from pin 1.   The output would swing +/- 13V.
In that case we would see a negative voltage on D4's anode and a positive voltage on D3's cathode.
Each diode would half-wave rectify the output from pin of the opamp, If we lose 0.6V due to the diodes
the output would swing +/- (13-0.6) = +/-12.4V.   

If the opamp is oscillating as a sine-wave:
When we measure the half-wave rectified signal with a multimeter the measurement will be 0.32 times the peak, so 0.32 * 12.4 = 4V.

If the opamp is oscillating as a square-wave:  When we measure the half-wave rectified signal will be 0.5 times the peak, so 0.5 * 12.4V = 6.3V

So we might expect +4V to +6.3V on D3's cathode and -4V to -6.3V on D4's anode.
We are seeing +5.6V and -5.2V, very suspicious as the measured values as in the middle of what we expect to see if IC5a is oscillating.

IC5 Pin 2:
No Signal: +1.4mV
Signal: +1.4mV

Pin 2 near 0V means IC5a is working.


There's another things we can check.

IC5 Pin 14:
No Signal: -12.94V

IC5 Pin 13:
No Signal: -4.25V
Signal: -4.82V

D3 Cathode:
No Signal: +5.59V
Signal: +3.69V

We expect the current through R67 and R68 be be equal (*assuming no shorts* ---> but we haven't measured any shorts).  From ohms law, since R67 and R68 are the same value that means we should be seeing the same voltage across each.

R67:   Vdrop = (5.59 - (-4.25))     =  9.84V
R68:   Vdrop = (-4.25 - (-12.94))  =  8.69V

That's close but now quite as close as I would expect.

Retaking all the readings:

R67:
Side 1: 4.35V
Side 2: -4.34V
4.35V-(-4.34V)= 8.69V

R68:
Side 1: 4.35V
Side 2: -13.12V
4.35V-(-13.12V)= 17.47V

WAY OFF with the last measure!

With IC5 out (just for kicks and giggles):

R67:
Side 1: 11.7mV
Side 2: 11.8mV
Basically 0...

R68:
Side 1: 12.6mV
Side 2: 12.7mV
Basically 0 again

I can understand *if* the circuit is going nuts then you could get weird negative voltages.
However, I can't make any sense beyond that unless IC5 is actually oscillating!

As far as a fix goes I think leave the 100pF in and try the 100nF caps on the power rails.   The ground points on the 100nF caps are best placed near IC5.   The 100pF and 100nF cap thing is a common fix for oscillations.

It's very tempting to short out both diodes and see if at least the DC voltages come back to normal.   You might be best to do that *after* you put in the 100nF caps.

OK, putting the 100nF in.  Let me take readings and report back

[jfrabat]

I can understand *if* the circuit is going nuts then you could get weird negative voltages.
However, I can't make any sense beyond that unless IC5 is actually oscillating!

As far as a fix goes I think leave the 100pF in and try the 100nF caps on the power rails.   The ground points on the 100nF caps are best placed near IC5.   The 100pF and 100nF cap thing is a common fix for oscillations.

It's very tempting to short out both diodes and see if at least the DC voltages come back to normal.   You might be best to do that *after* you put in the 100nF caps.

OK, I did this:



That is what you meant, right?  With that, I get these voltages:

RV6 Pin 2 @ -12.04
No Signal

IC5 Pin 1: Oscillating between +0.33V and +0.4V
D3 Anode: Oscillating between +0.62V and +0.71V
D4 Cathode: Oscillating between +0.47V and +0.22V
D3 Cathode: +3.80V
D4 Anode: -5.65V
Pin 14: -12.87V

[Rob Strand]

That is what you meant, right?

Yes.  Maybe better if the ground point was near pin 3, close to the chip.

It doesn't look like it has done anything to fix it.

Wait a bit until I can catch-up reading your last two posts more thoroughly.

[jfrabat]

That is what you meant, right?

Yes.  Maybe better if the ground point was near pin 3, close to the chip.

It doesn't look like it has done anything to fix it.

Wait a bit until I can catch-up reading your last two posts more thoroughly.

I accidentally deleted the old board file when I went to fix the polarity, so I have to manually check for GND...

[Rob Strand]

I accidentally deleted the old board file when I went to fix the polarity, so I have to manually check for GND...

Well, pin 3 itself should be ground anyway.

You accidentally deleted your PCB file?   Oh man, that sucks.

[Rob Strand]

Retaking all the readings:

R67:
Side 1: 4.35V
Side 2: -4.34V
4.35V-(-4.34V)= 8.69V

R68:
Side 1: 4.35V
Side 2: -13.12V
4.35V-(-13.12V)= 17.47V

WAY OFF with the last measure!

Maybe the positive side of  R68 side was negative?  That makes more sense.

Doesn't matter though.   At this point we are pretty much stuck.   Everything we have tried more or less has the same problem.

When I see problems like this it's often something we are assuming is true which isn't really true.

It might be time to change something.  Put a shorts across D3 and D4 then measure the DC voltages on,
IC5 pins 1, 12, 13, 14
and D3 cathode.

They should all be zero.  The idea is to find something that makes sense by simplifying the circuit.

[11-90-an]

Doesn't matter though.   At this point we are pretty much stuck.   Everything we have tried more or less has the same problem.

Hows the other PCB going, Felipe? 
If the other one doesn't work, then we can assume it's the PCB's problem... or the same mistake...

[jfrabat]

Maybe the sides of  R68 side was negative?  That makes more sense.

Doesn't matter though.   At this point we are pretty much stuck.   Everything we have tried more or less has the same problem.

When I see problems like this it's often something we are assuming is true which isn't really true.

It might be time to change something.  Put a shorts across D3 and D4 then measure the DC voltages on,
IC5 pins 1, 12, 13, 14
and D3 cathode.

They should all be zero.  The idea is to find something that make sense by simplifying the circuit.

No, Indouble checked polarity on both resistors.  One side + and the other was - in both cases.  I double checked because I found that odd.

Do I do the short on D3 anothento D4 cathode (IIRC the schematic; not at home right now!)?

Doesn't matter though.   At this point we are pretty much stuck.   Everything we have tried more or less has the same problem.

Hows the other PCB going, Felipe? 
If the other one doesn't work, then we can assume it's the PCB's problem... or the same mistake...

Soldered all the wires for the jacks last night.  The jacks I will install when it is ready for the bix, as these wires are a bit stiff, and they can break from handling (single core; only kind I have that fits the PCB hole).  I have to check if I have enough switches, though.  Still need to wire all the pots.  I will do that today.

[Rob Strand]

No, Indouble checked polarity on both resistors.  One side + and the other was - in both cases.  I double checked because I found that odd.

That is weird.   I'll have to try to make sense of that. 

How confident are you that the PCB matches the schematic?

Do I do the short on D3 anothento D4 cathode (IIRC the schematic; not at home right now!)?

Just put a just across D4 and a short across D4.  It takes the diodes out of the circuit.   The circuit will then work
like an inverting gain stage followed by a differential gain stage.   The main point being the DC voltages
should make sense ... if everything is working  .

[Rob Strand]

I can't make sense of the point where R67 and R68 join being at a positive voltage.

That join point should be the same voltage as pin 13 IC5.

In previous measurements pin 13 IC5 was a negative voltage.

Measure the voltages on R67 and R68 and find which side is positive.
Check pin 13 IC 5 is also positive.

At the point where you got the positive voltages, use the continuity test to check R67, R68
and pin 13 IC 5 are all joined together.

If we can't make sense this the PCB might not match the schematic.

[Rob Strand]

If you have any doubts the PCB doesn't match the schematic, you could take one of the blank boards and check all the connections around:  IC5a, IC5d, R64, R65, R66, R67, R68, R69, D3, D4.

[jfrabat]

EDIT: Values corrected (no big changes)

How confident are you that the PCB matches the schematic?

Well, I did it in Eagle, so THEORETICALLY, it should match, as the connections are carried over from the schematic.  I do know the TL072 and NE5532 have the voltages the wrong way about because I had them like that in the schematic...  The original one, at least.  I now fixed the schematic, but lost the board design in the process...

I can't make sense of the point where R67 and R68 join being at a positive voltage.

That join point should be the same voltage as pin 13 IC5.

In previous measurements pin 13 IC5 was a negative voltage.

Measure the voltages on R67 and R68 and find which side is positive.
Check pin 13 IC 5 is also positive.

At the point where you got the positive voltages, use the continuity test to check R67, R68
and pin 13 IC 5 are all joined together.

If we can't make sense this the PCB might not match the schematic.

You know, I have been using the old MM because the new one had trouble locking on to some voltages.  But that one had stopped working some time back, and only recently started working again.  I tried to measure again with the new one, but, again, it cannot take a reading on Pin 1 or D3a and D4c.  So I looked for a third voltmeter (cheap Radio Shack one I have in my Jeep), but that one has the same issue (no readings).  It tries to lock the voltage range (both only have auto range), but cannot do it.  Only the old one is giving me a reading...  HOWEVER, the old one ALSO gives me a reading with no leads connected!  Look at the video:

So it is QUITE possible that the readings in Pin 1, D3a and D4c are just the MM going crazy and not actual readings, which could explain why they change every time!  Maybe it is just 0V and what we are seeing is the MM going crazy!

The other thing I was thinking is that I used the sleeve of the output as a GND.  But what if the issue is that there is some voltage there?  I measured it with the negative lead grounded to the jack, and got -0.5mV, so not sure that is the issue, but just wondering.

Anyway, I retook the readings (with the new MM), and this is what I got:

RV6 Pin 2 @ -12.13V
No Signal

IC5 Pin 1, D3 Anode and D4 Cathode: No reading

D3 Cathode: +4.35V
D4 Anode: -5.95V
Pin 14: -12.86V

With the diodes bypassed:

RV6 Pin 2 @ -12.08
No Signal

IC5 Pin 1: No reading
Pin 14: mostly no readings, but I did get some quick shots of 0.-0.428 and -0.486 (and some others in between)
D3 & D4: Same as Pin 1

I also took readings with the old MM and the diodes bypassed:

RV6 Pin 2 @ -11.94V
No Signal

IC5 Pin 1: Oscillates around 0.5V (less than 50 mV each way)
Pin 14: Oscillates around 0.5V (less than 10mV each way)
D3: Oscillates around 0.8V
D4: Oscillates around 0.7V

If you have any doubts the PCB doesn't match the schematic, you could take one of the blank boards and check all the connections around:  IC5a, IC5d, R64, R65, R66, R67, R68, R69, D3, D4.

I checked a couple of days ago.  Connections that should be there, are.  Not sure there is one there should not, but considering how Eagle works, the chances of that are low.

[Rob Strand]

RV6 Pin 2 @ -12.08
No Signal

Just quickly, you need IC 5 pin 13 etc.

RV6 is past all the problem area.  Also,  the voltage supposed to be negative and doesn't depend on IC5 working.

[jfrabat]

Just quickly, you need IC 5 pin 13 etc.

RV6 is past all the problem area.  Also,  the voltage supposed to be negative and doesn't depend on IC5 working.

Just included it in case it affected the voltages.

Here are all the pins:
Pin 1: Oscillates around 0.5V (less than 50 mV each way)
Pin 2: 1.4mV
Pin 3: 0V
Pin 4: 15.23V
Pin 5: 50mV
Pin 6: -0.637V
Pin 7: 4.21V
Pin 8: Oscillates around -0.9V
Pin 9: Oscillates around -0.35V
Pin 10: -11.85V
Pin 11: -14.89V
Pin 12: Oscillates around -0.35V
Pin 13: Oscillates around -0.43V
Pin 14: Oscillates around -0.37V

EDIT: This is all with the diodes bypassed

I also am working on the other board to compare.  I am missing a couple of switches, but I can tie the cables together to simulate the switch.  I only need to solder the switched jack to start testing on that board (but, then again, I have not tested the other functions of that board, and the LEDs and input and output jacks are not soldered, so I cannot test it right now either!).