Paia Gator - nothing comes out

[salicet]

Hello,

after leaving that not so well working Boss Slow Gear behind, I build the Paia Gator from this layout http://www.diystompboxes.com/smfforum/index.php?topic=89647.0;prev_next=prev and unfortunately ran into some unpleasant building results. It's seems dead.
I used all the "original" ICs (IC1-OP07, IC2-RC4136, IC3-CA3080) except for a 100k pot for level because I ran out of 250k. After checking all the bridges and testing the breaks negative for continuity, I decided to look for help.

Since there seem to be some successful builds around this forum, I hope to get some hint on that.

I use 3PDT Bypass-switching and the bypass works but when engaged all stays silent.
Audio probing says there's signal on IC1 and hitting pin 1 of IC2. After that everything is dead.
I checked the voltages on each IC and comparing them with a previous post, the voltages seem strange:

IC1

1: 8,9 V
2: 4,5 V
3: 2,9 V
4: 0 V
5:  0,14 V
6: 4,45 V
7:  8,9 V
8: 8,85 V


IC2

1: 1,1 V
2: 1,1 V
3: 1,3 V
4: 1,3 V
5: 1,1 V
6: 1,1 V
7: 0 V
8: 1,3 V
9: 1,1 V
10: 1,3 V
11: 8,9 V
12: 1,3 V
13: 1,1 V
14: 1,3 V


IC3

1: 0,14 V
2: 1,2 V
3: 1,1 V
4: 1,3 V
5: 1,1 V
6: 1,3 V
7: 8,9 V
8: 0,15 V

Does somebody has an idea what could be wrong? I powered with a 9 V power supply, haven't tried battery (don't intend to use any on that).
Could it possibly be a dead IC?

I would deeply appreciate some help!
Thanks!

Alex

[GibsonGM]

Have you tried using an audio probe (AKA, your best friend) to listen in the signal path, and see where you lose signal?

Issues like this are often solder bridges causing shorts.  Did you socket the IC's?

I am not familiar with the IC pinouts, but be sure your ground and V+ pins read 0 and about 9V...

[armdnrdy]

Have you tried using an audio probe (AKA, your best friend) to listen in the signal path, and see where you lose signal?

Copied from above:

"Audio probing says there's signal on IC1 and hitting pin 1 of IC2. After that everything is dead."

[PRR]

> Could it possibly be a dead IC?

Where did you get your LM3080? Non-functional fakes flood the market.

But pin 7 IC2 has to be +9V, not zero. That's your work, not a dud chip.

And IC3 has more than 8 pins. And pin 7 must NOT! be 9V, it must be zero (suggesting a major power reversal).

[salicet]

Thanks,

I'm afraid, I've mistaken IC2 and IC3 in the text:
IC2 is of course the CA3080, so corrected:

IC2 (CA3080):
1: 0,14 V
2: 1,2 V
3: 1,1 V
4: 1,3 V
5: 1,1 V
6: 1,3 V
7: 8,9 V
8: 0,15 V

IC3 (RC4136)

1: 1,1 V
2: 1,1 V
3: 1,3 V
4: 1,3 V
5: 1,1 V
6: 1,1 V
7: 0 V
8: 1,3 V
9: 1,1 V
10: 1,3 V
11: 8,9 V
12: 1,3 V
13: 1,1 V
14: 1,3 V

With the audio probe, the signal hits pin 1 on the RC4136 and vanishes after that.

But pin 7 IC2 has to be +9V, not zero.

And IC3 has more than 8 pins. And pin 7 must NOT! be 9V, it must be zero (suggesting a major power reversal).

So, according to the confused IC labeling pin 7 voltages on both seem correct at last? I bought the CA3080 over here in Germany at Banzai http://www.banzaimusic.com/CA3080E.html
Unfortunately a very pricey chip.

Regards, Alex

[Seljer]

Something is wrong with your vref supply, on all the pins of IC2 and IC3 something is dragging it down to 1.1V . For good measure check the values of R20 and R21. If you have the two ICs socketed remove them seperately and then check the voltages on the pins.

The first opamp works because R1 and R2 set IC1 to 4.5V seperate of the rest of the circuit.

[GibsonGM]

Have you tried using an audio probe (AKA, your best friend) to listen in the signal path, and see where you lose signal?

Copied from above:

"Audio probing says there's signal on IC1 and hitting pin 1 of IC2. After that everything is dead."

Tired, missed that  ;o)   

[salicet]

For good measure check the values of R20 and R21.

That was a good one! R20 had a bad contact. I fixed that one and measured again, with somewhat better results:

IC2 (CA3080)

1: 0,14
2: 4,5
3: 4,5
4: 4,1
5: 4,4
6: 4,5
7: 8.9
8: 0,14


IC3 (RC4136)

1: 4,5
2: 4,5
3: 3,5
4: 4,5
5: 4,5
6: 4,5
7: 0
8: 4,5
9: 4,5
10: 4,1
11: 8,9
12: 4,5
13: 4,5
14: 4,5


I get a signal through the probe on pin 3 (quite loud) and on pin 4 (slightly distorted) on the RC4136. The level of audio was adjustable by the treshold pot. Once again after that all is quiet as a coffin.
Still, I tested voltages with IC2 & 3 unsocketed, I don't know, if that would help identify a busted chip?

so this is without the ICs:

IC2

1: 0,01
2: 4,5
3: 4,5
4: 0,1
5: 4,4
6: 0
7: 8.9
8: 0,14


IC3

1: 0,8
2: 4,5
3: 0,6
4: 0,4
5: 4,5
6: 0,3
7: 0
8: 0,2
9: 4,5
10: 0,2
11: 9
12: 0
13: 4,5
14: 0

Thanks, Alex

[Seljer]

Those voltages look much nicer

Measure IC2 pin 4 (or IC3 pin 10 because its the same trace) while strumming, the voltage should jump around when the envelope section triggers, mess with the threshold anddelay settings to get it slow enough that your multimeter catches it.
The audio probe findings make sense, pin 3 of IC3 is already the envelope so nothing audible should be there.

Is there still nothing on the output side of IC2?

[salicet]

I tested those pins you suggested and there isn't much of a party there
When strummed and depending on the treshold there were jumps in the milliVolt dimension, like going DOWN from 4,3 Volts to 4,12 Volts, climbing up again when the note starts to decay.

I suppose the RC4136's hemicycle nick is the up-position marking pin 1+14, curious because there is a dot on the other side?! But I guess, I got it right.

Unfortunately still nothing on the other side of IC2. Could it possibly be busted? How could you test a dead IC?

Alex

[duck_arse]

pin 1 is on the bottom left. we had a thread on pip mouldings:

http://www.diystompboxes.com/smfforum/index.php?topic=104701.0

can we see photos of yr work?

[Seljer]

The effect seems to work by moving around the ground pin of the 3080. If you want to test your CA3080 chip remove IC3 and grouind pin 10 (just temporarily jumper it to pin 7). If the 3080 functins should be able to get audio running through it.


I believe the voltage on IC3 pin 10 should be dropping much lower when the envelope section triggers.

Those chipless voltages actually do help. Something funny is going on.
Pins 2, 5 and 9 of IC3 are at 4.5 volts whichs is correct. 12,13 and 14 also make sense.

But if you look at the chain of the inputs and outputs in the circuit's order from pin 1 to pin 10
1: 0,8
3: 0,6
6: 0,3
4: 0,4
8: 0,2
10: 0,2
So if you look at the schematic without the opamps, by the falling voltages you'd deduce that a phantom current is flowing from left to right through the line of resistors from somewhere around C4 (or maybe IC3 pin 1)! The circuit has got rather high value resistors in it so any residue of the flux from soldering may be enough to screw it all up. That or a bodged up trace cut or something.


edit: or these three are also strange from the voltage with IC3 in the socket
1: 4,5
2: 4,5
3: 3,5
The mechanism of the feedback scheme around the opamp is to adjust the output (pin 3) until the negative input pin 1 matches the reference voltage at pin 2. Which it does appear to be accomplishing in some way. But it doesn't make sense because pin 3 is at a lower voltage, meaning theres current flowing from pin 1 to pin 3.
There should be no current flowing through R12. So all the current from the output flows through the R14 and D1, then through R13 which is attempting to drag pin 1 to ground. In theory, the output voltage should be 4.5volts + whatever drop you encounter over D1 (you can ignore R14 if the diode is conducting), so higher than pin 2, somewhere between 4.5v and 5v, not lower! So again I'm guessing theres current from the 9v supply traces sneaking in somewhere around pin 1. And because the chipless voltages point the same conclusion I'd wager that the chip is fine.


edit2:
and to check if IC3a works correctly, you can remove IC2 and jumper pins 2 and 6 and should have (quiet) audio passing through

[salicet]

Wow! Thanks for the detailed diagnosis, Seljer!
I'll take a look at the schematic and check the circuit tomorrow and maybe post a picture. Good suggestion on testing the ICs...
One more thing: what I really don't understand in the scheme is the connection from the level pots' lug 1 to V ref. Normally it would just go to ground, wouldn't it?

What is V ref in this circuit anyway?

Alex

[Seljer]

Vref is the reference voltage, half the supply voltage, provided from the voltage divider made out of two 10kiloohm resistors at the bottom left in the schematic.

The CA3080 is an operational transconductance amplifier, which basically means you can adjust it's gain by adjusting the current going into pin 5 (current going into, which implys that the pin 4 is at a lower voltage). The level pot sets this current. This is what striked me as kind of strange in the design, because usually you'd usually use the chip by grounding pin 4 normally and then adjusting the current on pin 5 with your envelop or LFO, which is how the Dynacomp compressor and Small Stone are set up. No idea why this circuit is the way it is, but it it works then it works  

[salicet]

Ok, the tests of the ICs showed a slightly but recognizable distorted signal when IC3 (RC4136) was removed and pin 10 was set to ground. The audio was quiet though, definitely way quieter than bypass.
IC2 (CA3080) removal showed, as you suspected a, quiet signal that was clear (undistorted) but quiet (below the level of the signal when IC3 was removed).

Should I worry about IC2 beacuse the signal is quiet and dirty?
I will check all the connections at the inputs of IC3b and report.

Thanks, again,
Alex 

[Seljer]

It just occured to me that the CA3080 has got a current output, not a voltage output, hence it needs the output opamp to be installed for it to function correctly (so current can flow througho R7). Try it with IC3 removed and along with pin 10 grounded as before, jumper together pins 14 and 13. That should recreate normal conditions and should be able to verify the output of the CA3080 with the audio probe as well as checking if the level control works.

And I also just realised that the circuit is missing an output decoupling capacitor.

[PRR]

I'm hung on a roofing project and not watching this.

> CA3080 has got a current output, not a voltage output, hence it needs the output opamp to be installed for it to function

Alternatively wire ~~10K from Out to a Vref. That *should* hold the Out at a mid-DC voltage and allow good audio to be developed out of it. (I think.... I have not re-reviewed the schematic.)

Yes, the 3080's Iabc and V- are connected unusually. Not wrong, just mind-bending.

[salicet]

Try it with IC3 removed and along with pin 10 grounded as before, jumper together pins 14 and 13.

Tested with those results: I get VERY low audio from pin 6 of the CA3080 which you can adjust with the level pot clockwise.
When pins 13+14 of IC3 socket not jumped, the signal appears louder (but quieter than bypass) and can be raised (and amount of distortion) when turned anticlockwise.

Alternatively wire ~~10K from Out to a Vref. That *should* hold the Out at a mid-DC voltage and allow good audio to be developed out of it. (I think.... I have not re-reviewed the schematic.)

That made the audio signal clear and as loud as the bypass but without any effect of the level control.

I checked the connections for continuity to Vref or V+ on all ICs, they appear to be correct. My knowledge of ICs is poor but why doesn't the CA3080 have a ground pin?
Sorry for filling this thread up so much with no big progression from my side. I'll try to take pictures tomorrow and post them. Maybe this could help.
Unfortunately, I have no spares for IC2 and 3.

Alex

[PRR]

Schematic:
http://i703.photobucket.com/albums/ww40/luckyfeline7/schematics/GatoGator11.gif

> why doesn't the CA3080 have a ground pin?

'3080 data:
http://www.ti.com/lit/ds/snosbq5a/snosbq5a.pdf

The NEgative power pin is pin 4.

With a single-battery (or equivalent) supply, we would conventionally connect pin 4 to Ground.

However. In the '3080 the supply voltage isn't too critical. What does matter is the _current_ into pin 5, which controls the chip's gain. And pin 5 connects to pin 4 through a diode (transistor).

In this Gator, instead of nailing pin 4 to the ground and jiggling pin 5, they choose to wire pin 5 positive through a large resistor, and pull-down pin 4 to increase the gain. This may simplify the "side-chain", all that stuff across the bottom of the schematic which computes "some function" of the input signal and applies it to the '3080 control port.

I'm driving my truck. To get "more", I normally push the gas pedal down to the floor. But, in concept, I could instead hold the gas pedal steady and push the floor up to the pedal. Well, that doesn't make much sense with a 6,000 pound truck and a 1 pound foot. But the supply pins of the '3080 are much easier to push-around than a truck floor. And apparently this "inverted" action made sense to the Gator's designer.

[salicet]

Here are some pics, maybe not the best quality, a bit blurred on the solder side. So far, I checked the solder connections on the stripboard for unwanted contact to the upper/ lower rows.
I also checked the switch.

I suppose, the circuit would work without a decoupling cap, but I can put one in, if I reach a successful working state. Would 100 nf be enough?

Alex