"K" style transparent overdrive build help/tone debug please

[Graham YC]

Hi! First time poster and haven't built anything for 28 years, so please go easy on me and if I've broken any protocol, please say! Full description with links to the schematic, voltages around the circuit etc are below, as per the debugging 'How To' sticky.

My build basically operates cleanly and noiselessly, with a familiar sound to the distortion and nice clean boost, but it suffers from one issue: as the gain pot turns above 20% or so, quite a lot of clarity/top end is swept out of the sound.  I did have some support from the supplier, but I know I can't hassle them forever over something that could just be down to my level of competence.  One suggestion is that there may be a pinch point in the gain circuit somewhere, draining off those upper frequencies.  I really can't find it, and if there's something else it could be – even tolerances or known issues – or anything obvious from the data below, I'd be grateful to know, before I start ripping up IC sockets on a tight board.

Name and origin of circuit: Boutique Transparent Overdrive kit from Fuzz Dog

No mods, and only the supplied subs/options on some caps: All values followed for the Silver version, but with the C14 suggestion of 6n8 for a more useful tone control range and 100n for C2 (more bass in the clean section).  I removed the 100n once, switching it for the original 68n value, which only proved that the 100n did indeed provide more bass, so swapped it back.  I originally fitted the suggested 150n alternative for C6 as well, but later swapped it for the original value of 82n.  It made little or no difference to the sound that I could detect, but the 82n remains in place.
 
All resistor values colour-code-checked since fitting and multimeter-checked beforehand.

I've swapped the TL072s' places once, just to check nothing was wrong with either; all fine.

All electrolytic capacitors test more positive at the anode than the cathode.

I burnt out the centre pad of the tone pot, removing and refitting to access C2, and have had to put a connecting wire in place.  Nothing has changed in use as a result.

It is not a positive ground circuit. (Edit- virtual ground 4.5V from charge pump)

Voltages at IC pins and diodes, all taken with no input connection and all pots roughly half way:
IC1
1) 4.56
2) 4.56
3) 1.04 [Edit - that reading may be low, due to the impedance of my meter, as per reply below]
4) 0.00
5) 4.44
6) 4.56
7) 4.53
8) 9.13
IC2
1) 4.56
2) 4.56
3) 4.50
4) -8.72
5) 4.50
6) 4.56
7) 4.56
8) 16.46
IC3
1) 9.13
2) 4.66
3) 0.00
4) -4.27
5) -8.71
6) 5.48
7) 4.97
8) 9.13
D1
A) 0.00
C) 0.00
D2
A) 0.00
C) 0.00
D3 (LED)
A) 1.90
C) 0.00
D4
A) 9.13
C) 12.93
D5
A) 12.93
C) 16.46
D6
A) 0.00
C) 9.13

[antonis]

Just to have a reference..



Check for appropriate VB (4.5V) on R2 upper leg, then on the node of R29/R30..
(if it's OK, then "something" is pulling down VB on the node of R2/C1/pin3..)

[Graham YC]

Thanks antonis

The upper leg of R2 measures VB, 4.5.  So does the node of R29/R30/C20a.  So I assume something is pulling on IC1 pin 3/R2/C1 as you suggest.  (Edit - Just to confirm, that node is definitely 1.04V; I checked.)

I'm examining the tracks and joints involved with a magnifying glass.  Some of the pin 3 node is obscured.  To see all of the pin 3/RC/C1 node, I will have to remove the tone pot (again), IC1's socket, C1, C2 and maybe C4.

Is there anything else I can test (maybe even with IC1 removed from its socket?) before doing that?

By the way, I have 0.00V at the junction of R1/C1/R31.  I assume that is correct, without an input signal.

Much appreciated.

[Slowpoke101]

Welcome to the forum.

Check the orientation of C7. If it is reversed it can cause problems similar to what you have described - amongst a multitude of seemingly unrelated problems (it is a Klon after all ).

Edit: Also, the low voltage on pin 3 of IC1 is most likely due to your meter loading that point. Pin 3 is supplied with 4.5V (Vref ) via a 2M resistor and if your meter has an input impedance of 1M a voltage divider is now formed which should read close to the voltage you reported.

[anotherjim]

Almost certainly meter impedance loading giving a low reading on IC1-3. Note that the output of that amp pin1 is reading correctly, which almost always means that is what the amp is "seeing" on its pin3.

[Graham YC]

Thanks all.  It's a dirt cheap XL830L meter so I assume you're right. 

I really wanted to find that C7 (1uF tantalum) was the wrong way around when you said that, but sadly not. I've got the + symbol going to the square pad, which I traced along the tracks to R7/R8 just now, to make sure.  To check the value, the marking is 105 35, which I guess is in pF (10 and 5 zeros) and 35V rated.

Where next please? ...if all the voltages look OK

[willienillie]

What clipping diodes did you use?  Did you try anything different?

[Graham YC]

What clipping diodes did you use?  Did you try anything different?

The D9E germanium diodes supplied.  I haven't tried anything else. 

My basic understanding of the circuit is that these shape the clipping, especially at higher gain, rather than affect the attenuation of a distinct band of upper or mid-upper tone - but I'm here to learn!  This attenuation can be heard well before the gain gets to 50%.  It's like you're dropping a slider on a graphic EQ while turning up the gain with the other hand.  The heavier clipping in the top 30% or so of the gain reintroduces a bit of brightness, but not in the same way, or equally across the whole band that has diminished up to that point.  Turn the gain back down below 25% or so, and all the clarity and 'air' returns.

[idy]

You maybe understand that when you turn the gain up or down you are also turning the volume down and up on the clean side, and that the clean side is full-ish frequency response and the dirty side is mid-rangey.

If there is a short on that side, or a wrong value in C7,8,11,12 for example, this side would have even less treble than intended. If turning the gain up far enough to get clipping and the clipping is not being low-passed, then C11/12 is not the problem, as that would be there to smooth out the clipping. But "K" lovers may know exactly who that filtering is tuned and where....

Are you sure the midrangey thing is a bug and not the way these "K" things work?

[Slowpoke101]

Are you sure the midrangey thing is a bug and not the way these "K" things work?

I connected my clone of this thing back into my pedal-board just to check. Its tone was fairly even across the gain range. No obvious peaks or troughs so I'd say that there is a problem.

Definitely check the values of C2, 3, 6, 7, 8, 11, 12 and R3, 4, 5, 7, 8, 9, 10, 11, 15, 16.

But first, you mentioned that you moved the tone pot and damaged the centre connection pad on the PCB. Set your meter to a low ohms setting (or continuity ) and with no power applied test for connection between pin 1 of the tone pot and R21. Then test for connection between pin 2 and C14 and finally check pin 3 to R18.

It is also possible that the tone pot may have been damaged by excessive heat but check the other stuff first before worrying about that.

[PRR]

> dirt cheap XL830L meter

Input resistance not specified by maker. One reviewer says 1Meg on DC. So if you measure a battery, then measure again but with 1Meg resistor in series, you will see about half the un-loaded voltage. As 470k and 1Meg are popular bias resistors, low reading is probably correct.

[Graham YC]

Firstly, thanks for all the really helpful responses. 

I knew I should have started with a simple Fuzz Face or something! The way the filters work against each other in this circuit clearly leads to difficulty in tracing issues around.  I'm taking my understanding from analysis here.  It would make sense to me that any problem in values or unwanted grounding of the signal from the network of C6, C8, R9, R10 and R11 would cause what I'm hearing, while the 'feed forward network 2' referred to there seems to pass on the full range just fine until it is faded out on the second gang of the gain pot (as the gain goes up).  I guess the latter possibility (pinch/grounding), suggested by the supplier, becomes more likely the longer it stands un-disproved.

I've checked for continuity as suggested by slowpoke; all fine, though I can't get to C14 with the tone pot in place on the board.  I guess that continuity is proven on the middle pin by a change in tone when the tone pot is turned?

I'm going to take some time to check the values mentioned for a third time, if not more, but I'm pretty sure everything is in order there.  In the meantime, can I post pics to this forum without external hosting please? (couldn't find a how to on that).

[Slowpoke101]

Posting pictures is not difficult. When the "Post reply" screen is open, have a look towards the lower left corner of the text window. Just below that point is a link titled "Add image to post". Make sure that you have your picture files where you can find them, click the link and follow the prompts.

Willienillie asked about the clipping diodes in a previous post. There was a case some months ago where the germanium diodes where causing some odd tonal and level problems. Changing the diodes fixed the problem. You may wish to consider removing those diodes and trying some other diodes.

Looking forward to see some pictures. They can really help.

Edit: I think that if when you vary the tone control and it makes a noticeable difference to the sound you can be assured that the centre terminal (pin 2 ) is connected to C14. I do suggest that you also check that the gain control pot is connected to the correct circuit elements too. Double-sided plated-through circuit boards are good but it is easy to damage them without realising it.

[antonis]

>3) 1.04 [Edit - that reading may be low, due to the impedance of my meter, as per reply below]<

Don't get me wrong Graham, but I'd further suggest to take another measurement with DMM probes shunted by a 100k (or even 10k) resistor..
(easily obtained via aligator clips or so..)

[willienillie]

Willienillie asked about the clipping diodes in a previous post. There was a case some months ago where the germanium diodes where causing some odd tonal and level problems. Changing the diodes fixed the problem. You may wish to consider removing those diodes and trying some other diodes.

Yes, I do believe those cases resulted from germanium diodes being damaged by soldering heat.  D9Es should sound pretty good in this circuit, but I've learned the hard way that any time there's a pair of Ge diodes (clipping, octavia), best to socket them and try a bunch.  Even without damage they can sound radically different from one set to the next.  I recently re-built my Dist+ with sockets, and some diodes sounded just awful, it took about a dozen sets to find a pair I liked.  (D9Es were pretty good but I went with some matched 1N59s.  I only have three 1N270s, the original spec part, but no combo of those worked out).

[Graham YC]

OK, great.  I will work through everything mentioned in turn.  I've ordered some 100k resistors which will be a couple of days.

The nature of the clipping seems good and I got the germanium diodes in very quickly (<2secs) heatsunk, but I won't rule them out as an issue just yet.

I hesitate to display my modest efforts, but here we go. Be cruel to be kind.  The back of the board was taken before fitting the last two pots during original construction.  The burnt out pad 2 of the tone pot is (just) visible in the other shot and a connecting wire now bridges the gap.





[/url

[willienillie]

I'm not convinced those TL072s are genuine TI parts.  I don't claim any expertise, but all the known-real ones I've encountered have had printed markings, not etched in like those appear to have.  That said, they may be perfectly functional opamps, even if they're not actually TI.  But since they're in sockets, it would be easy to swap with something else, and see if the tone improves.