Debugging The Depths by EQD on Vero

[snk]

"the orange multiplier"... I get it! thank you!

[snk]

I am wanting you to be testing the resistance value of each resistor. that circled resistor is one of a pair of 10k's, except that the multiplier on one looks red, and the other, the one that counts to 100k, looks orange, which would be ...... 100k.

Duck Arse, thank you so much for your hawkeye and expertise : that silly little resistor is changed, and now everything seems to work fine ! 
Kudos!

[duck_arse]

mmmm, hang on a minute, you ain't finished yet. what did you find when you measured all those other resistors, any others out-by-10? and yes, that band on the [5 band] resistors is called the multiplier, sorry for the lack of explaining.

I think that 10k-100k is one half of the Vbias divider string for the opamps, which would have pulled the voltages way low, 10 to 100 instead of 10 to 10. but you've measured them all again since, haven't you? and you're going to post them here for us to check, aren't you? it seems only fair.

[snk]

Hi,
When I measured the resistors, they all matched the expected values (except, of course, the wrong one. Also, I did not test the 2.2M and 3M, as my DMM doesn't go above 2M).
I spent some time yesterday playing with the effect, and so far, it seems to work fine.
Today, I built an enclosure for the circuit, but I will take some time tomorrow to measure the values once again (and yes, that's fair, given the help provided). By the way, did you mean the resistor values, or the voltage at the opamp pins (with the opamps removed)?

[duck_arse]

if you have a bad/wrong/backwards opamp, it will throw the voltages off. and because the opamp outputs follow the voltage on their inputs, they can confuse and confound when other errors are present. so, at times, we pull the opamps and measure the voltages on the empty pins. that can/will/should show things like correct supply connections and correct bias voltage and things shorting when they shouldn't. if you have, for example, a 10k//100k divider instead of 10k//10k, it becomes very apparent.

I keep looking at your board photo, and comparing the red bands I can see [that are actually red] to the bands on the 47k resistors, and they [47k's] seem to be very orange to me. and it looks possible, from how the circuit shows those resistors connected to the opamps, that if 470k's were instead fitted, in pairs, their would be no discernable audio difference [to if they was 47k]. but I only think this, I don't know it for a fact.

and now that you have experienced a resistor colour band reading problem build, I can tell you the professional pcb assemblers would/will always align the resistors such that the tolerance band was always facing the same direction. this shows you've taken the attention in the first place, and makes it less fumble for when you [or the testing tech that follows production] come to read the values when debugging, or when others have to read to debugg. and it counts double for the blue-bodied resistors, just because they are harder.

[snk]

if you have a bad/wrong/backwards opamp, it will throw the voltages off. and because the opamp outputs follow the voltage on their inputs, they can confuse and confound when other errors are present. so, at times, we pull the opamps and measure the voltages on the empty pins. that can/will/should show things like correct supply connections and correct bias voltage and things shorting when they shouldn't. if you have, for example, a 10k//100k divider instead of 10k//10k, it becomes very apparent.

Thank you for the technical insight.
I'm going to check the voltages and report back.

I keep looking at your board photo, and comparing the red bands I can see [that are actually red] to the bands on the 47k resistors, and they [47k's] seem to be very orange to me. and it looks possible, from how the circuit shows those resistors connected to the opamps, that if 470k's were instead fitted, in pairs, their would be no discernable audio difference [to if they was 47k]. but I only think this, I don't know it for a fact.

I checked yesterday all the resistors, which measured as expected. I tested again all the 47k a couple minutes ago, and it was fine... So i guess it was maybe a lighting issue while taking the picture?

and now that you have experienced a resistor colour band reading problem build, I can tell you the professional pcb assemblers would/will always align the resistors such that the tolerance band was always facing the same direction. this shows you've taken the attention in the first place, and makes it less fumble for when you [or the testing tech that follows production] come to read the values when debugging, or when others have to read to debugg. and it counts double for the blue-bodied resistors, just because they are harder.

Thank you for the tip. I had never paid attention to that, but now it becomes obvious, and makes such kind of debugging way faster. After all, this is what these color strips are for

Now, I'll go measuring the pins voltage before boxing the circuit into an enclosure

[snk]

New chips measurement (errh, without the chips).
It's totally different than the previous measurements :

CHIP 1
0
0
2.20
0
0
4.30
4.11
882

CHIP2
0
0
moving (between 1.8 and 4.80?)
0
4.30 (changing slightly, around 4 and 4.50 it seems)
0
0
8.83

CHIP3
0
0
around 4.30 (varying slightly)
0
around 4.30 (varying slightly)
0
0
8.83

[duck_arse]

[just so we won't argue with you, it's always good to include the IC pin number itself.]

the four opamps non-inverting inputs connecting to the ldr's [pins 3 and 5] we can allow to wobble, because the resistance of the ldr will be changing and varying the way the meter measuring loads that reading. but they should all be the same, as the connections are identical - so "moving (between 1.8 and 4.80?)" looks out of place.

the input opamp non-invert pin (+) connects to Vbias via a 1M resistor, so measuring at pin 3 will again load down the reading - yours is ok. the other half of that oppie is connected differently, with Vbias to the (-) inverting pin, pin 6, and then thru another resistor to pin 7, so those readings, pin 7 a little lower than pin 6, seems right.

all ground and supply pins measure same, so that's good, and all the other pins are not connected to DC except when the IC's are fitted, so they are good. so, only pin 3 of chip 2 needs attention - perhaps remeasure to make certain.

but if it works and you're happy with the sound and it's in a box - forge on! and ignore me.


oh, and to really, REALLY belabour the point - this stuff is all apparent when you have a circuit diagram, complete and matching your build, at hand to refer to, to compare with, and to discuss at. none of this can be divined from a layout diagram.