Great Cheddar Debug

[euclid]

Hello all,

Just working through my first build from scratch -- R.G.Keen's 'Great Cheddar'. I've made a good number of silly mistakes, but am slowly honing in on a finished project. Unfortunately i've come across a problem I can't seem to sort out myself. Thanks in advance for reading through this and any suggestions you can offer.

I had it working, although the tone was quite 'thin' and output was quiet -- realized that i had used 4.7nF caps throughout in place of 47nF, and 4p7 caps in place of 4.7uF (!!). Changed these to the correct values and now am left with something that sounds thoroughly clipped/degraded on the rotary knob settings 2-4, poor but more consistent on setting 1. Output volume is about 1/4 of normal when the volume pot is turned to max.

I have traced the circuit with an audio probe, and the signal seems to be in good shape, full volume and sounding like a big cheese until it passes through U2. Probing pin 2/6, things resemble what I am getting at the output.

Checked and double checked the orientation of the chip, swapped in a new op-amp and got the same result.  Op-Amps are TL071ACP and Q1-Q3 are all BC 549. No intentional substitutions otherwise. Been through the circuit a few times to verify values since my early screw-ups.

I can upload photos of the build and sound clips if it is at all helpful. Otherwise, the significant digits (Keen's in brackets):

Q1:
C - 1.34 (1.246)
B - 0.54 (0.606)
E - 0.015 (0.034)

Q2:
C - 1.837 (2.81)
B - 1.34 (1.246)
E - 0.747 (0.620)

Q3:
C - 0
B - 0
E - 0


U1: pins 2,3,6 read 4.2V as expected

U2: pins 2,3,6 read 1.294 (!)


Thanks again for reading through this - any suggestions as to where to start looking for more problems are much appreciated. I would just remove U2 and jump straight to the output, but I know i'd be back trying to solve this thing in two weeks anyway....

[petemoore]

Q1:
C - 1.34 (1.246)
B - 0.54 (0.606)
E - 0.015 (0.034)
  I have trouble guessing what each active stage 'is'. In this case unless it's not boosting anything the collector has very small way to go before hitting Gnd. This doesn't look like a biased active buffer or gain stage.
Q2:
C - 1.837 (2.81)
B - 1.34 (1.246)
E - 0.747 (0.620)
  What is common to both transistors ? Q2 seems low collector/limited room to swing being biased near gnd.
Q3:
C - 0
B - 0
E - 0  This one appears to not have a resistance between it's collector and V+ [if it's a gain or buffer stage.


U1: pins 2,3,6 read 4.2V as expected

U2: pins 2,3,6 read 1.294 (!)
  This is dual opamp ? U2...
  Go to the voltage divider and measure that two equal resistors make the Vbias and that it is around the center of the supply rails.

[PRR]

> U2: pins 2,3,6 read 1.294 (!)
> suggestions as to where to start looking

Why is U2 pin 3 not at half of 9V as it should be? R16 R17 force half-voltage. C11 short could spoil that. Look for 68K where should be 680K, or solder-bridge in that area.

[euclid]

Thanks so much for the help..

Seems as though the supply voltage at U2 is the problem -- although after checking R16, R17 and looking for solder bridges around C11 (all verified as correct and without shorts) -- I wonder if something might have gone awry along the +9 line.  The voltage at the junction of C5 and R8 is about 1V lower than it should be as well.  I verified R3, R5 and R8 but all check out as correct.

Would there be any value in swapping out Q2, C13 or C14 or am I reaching...

[R.G.]

Thanks so much for the help..

Seems as though the supply voltage at U2 is the problem -- although after checking R16, R17 and looking for solder bridges around C11 (all verified as correct and without shorts) -- I wonder if something might have gone awry along the +9 line.  The voltage at the junction of C5 and R8 is about 1V lower than it should be as well.  I verified R3, R5 and R8 but all check out as correct.

Would there be any value in swapping out Q2, C13 or C14 or am I reaching...

At this point it's useful to conduct some thought experiments, then go start working on the board, and not just go swapping parts. Here are some concepts.

If there's no signal going in, just DC on the power pins, then **all capacitors can be replaced with open circuits and nothing should change**.
For DC conditions as read by a meter, all the places on the PCB that are shown as connected by lines in the schematic should be within millivolts of each other.

So we can do the following.
- hook your meter minus/black lead to the ground connection on the input jack; all places on the PCB which have the ground symbol on them simply must have a voltage that's nearly 0Vdc, within a few millivolts, as measured by the red/positive lead, or there is an open/high resistance connection between them. That has to be found and fixed or it's not going to work right.
- with the meter black/minus still hooked to the ground connection on the input jack, all places on the schematic that are connected to the power supply simply must be the same voltage plus or minus a cat's whisker, or it's not going to work right.
- you can extend this: for the whole board, places connected by a wire on the schematic have to be at the same DC voltage, or they are not actually connected on the board, no matter what the solder joints and components look like.
- it's best to do the measurements on the actual pins of the components, not the solder pad on the bottom of the PCB, because the solder joint to the pin itself may be the problem.
- if you have a place where the voltages are suspicious on either side of a capacitor, then *pull the capacitor out*. This will not affect the DC voltages on either side of the cap at all if the cap was good and was installed the right way around (if polarized). But it will change things if the cap was conducting DC when it wasn't supposed to.
- tone/volume controls are often isolated from DC voltages in well designed circuits; DC on the controls where there is a capacitor which ought to be blocking DC is a read flag; poor designs sometimes led DC onto the controls either by intent or omission, and this says something about the designer.

So to answer that pesky question about whether the power supply is not right from one chip to the other, get out the multimeter and start tracing where the power supply trace goes from part to part, and if it's ever different as you go along it, you've just located at least one source of trouble.

[euclid]

Brilliant, I've got it solved. Tuned out to be a bad joint at R13. Not so bad that it looked terrible, but i noticed on probing the lead a bit that the voltage was fluctuating wildly between .85 and 4.2. This gave me some hope, and re-soldering it yielded success.   

Thanks again to everyone for the help. I really appreciate the thoughtful and patient replies and ideas as to where to look.
RG, this is the second time you've bailed me out -- The debugging tips and concepts go a long way to making the process seem less daunting.

And now.. I'm going to go take a close look at this thing testing the voltages as suggested. If there was one of these, there are bound to be others....

[R.G.]

Brilliant, I've got it solved. Tuned out to be a bad joint at R13. Not so bad that it looked terrible, but i noticed on probing the lead a bit that the voltage was fluctuating wildly between .85 and 4.2. This gave me some hope, and re-soldering it yielded success.   

Good work!

And now.. I'm going to go take a close look at this thing testing the voltages as suggested. If there was one of these, there are bound to be others....

Good idea. If there is one place where using a shotgun approach makes sense, it is on solder joints. If you find one bad one, it makes sense to remelt *all* of them with a little flux and fresh solder just to be sure.

[stringsthings]

Brilliant, I've got it solved. Tuned out to be a bad joint at R13. Not so bad that it looked terrible, but i noticed on probing the lead a bit that the voltage was fluctuating wildly between .85 and 4.2. This gave me some hope, and re-soldering it yielded success...

And now.. I'm going to go take a close look at this thing testing the voltages as suggested ...

trouble-shooting a circuit is indeed one of the realities of DIY electronics ... anyone that's ever put together enough fuzz-boxes has run into the usual problems and come out of the process with some knowledge ... often, when you are able to trace a problem back to the source, and correct it, you can "gain"  a new understanding ...

tip #56.8:   keep a notebook of your trouble-shooting adventures ... chick's dig dude's that keep notes!