Tonebender mk1- some questions [SOLVED]

[r2c2]

Hi guys,

I'm new to the forum and looking forward to participate!
I just started with diy guitar pedal building (and electronics as well ) and decided to build a tonebender MK1 as a first one.
So far it's been a great learning experience.

I've got the gary hurst mk1 schematics with some modifications on my breadboard for the moment.
At this stage I can make it sound pretty good and by playing with the components I can alter the behaviour in various ways.
I haven't decided yet upon the final configuration for boxing it up. Mainly because I still have a few questions. Pretty much every forum thread on tonebenders I have read but can't seem to find answers to those questions so maybe someone can shed some light?

Here we go:
- With my guitar volume fully up it sounds great and responsive but as soon as I lower the guitar volume the sounds start to break up (like a flat battery). It depends of course on the attack setting where the problem starts, but on average I need to be minimum above 7-8 (from max 10) on the guitar output volume. I tried with a friend's strat and there it was worse (different pickups, maybe not set up correctly and too far from the strings causing a lower output than my strat?) , even at full guitar volume I needed to hit the strings hard, when playing soft the sound breaks up as well.
I didn't expect it to clean up like a fuzzface but also did not expect that the circuit would not function anymore with too low guitar output volume.
Is this normal for a tonebender mk1 or is there something I can/should do to solve this?

- Some people seem to suggest that you need at least 2V on the Q1 emitter? I didn't find any transistor where I can reach that voltage but can get it by increasing the 8k2 resistor (around 33K depending on the transistor). Is that a proper way to do it? It doesn't seem to change much to the tone of the pedal so don't know if I need to worry about it and just leave it at 8K2.
Anyway, does anyone know what the reason could be for having at least 2V on the Q1 emitter?

- Maybe a more general question now on the 500K level pot. Ground is connected to lug 1 and according to most resources lug2 has to go to the output jack tip and lug 3 to output of the board. By mistake I switched lug2 and 3.
And to my surprise it sounds very different, when turning down the level it just seems to reduce the output volume of the pedal without changing the tone. With the 'correct' pot connections the sound seems to become more brittle and lose bass response when reducing the level.
Can someone explain this? As it's just a variable resistor and no caps are involved in that part of the circuit I would except that it behaves the same?

The following information might be usefull:
Currently I have 3 AC128K transistors on the board (but I have other combinations that work too).
Q1 Hfe:75 lk:0,131mA
Q2 Hfe:70 lk:0,078mA
Q3 Hfe:90 lk:0,110mA

Transistor voltage measurements(resp low-high attack):
Q1C: -9.45 -9.41
Q1B: -0.88 -0.88
Q1E: 0.84 -0.84
Q2C: -8.65 -3.22 (I changed the 470K Q2B resistor to 180K to be able to get past -4.5V, max fuzz seems to be as expected at around -4.5V)
Q2B: -0,103 -0.157
Q2E: 0 0
Q3C: -9.2 -9.16 (I can reduce it to 8-8.5V as suggested by some - by increasing the Q3C 15K resistor - but not sure if this is really needed?)
Q3B: -0.016 -0.015
Q3E: 0 0

I currently switched the C3 capacitor from 25uF to 0.1uF (like the solasound schematics).
There doesn't seem to be much documentation about the differnces between the solasound version vs the gary hurst one but with the 25uF, it sounds fatter and more stacatto (gated?) while with the 0.1uF it has smoother/longer sustain.
Probably I'll make this cap switchable because both flavours sound great to me.

I have added a 100uF power supply bypass cap. It solves screeching & rumbling noises that I had with some of the transistors although the circuit with the current AC128 transistors seems quite stable (only on lowest attack settings I get some rumble noises without the cap).
I didn't add a 100R resistor in series because when I added it the noise are still there - albeit less than without the bypass (need some more experimenting to be sure).

schematics referenced:

[poiureza]

Hello,

I'm not much of a help but since nobody replies I'll give it a shot :

0.015V @Q3b seems VERY low and shows that those PNP trannies are leaking a lot (it shows for Q1 too). Q2 leakage seems fine.
Drop a different transistor in that position and look if things improve wrt guitar volume.
Some fuzz circuits simply need a strong enough signal to work. You can improve this by changing bias conditions but then it'll sound different too.
Maybe your trannies are simply just too leaky ...

Q1 is just a buffer I think, so moving Q1b closer to 4.5V improves headroom. Not really needed to have 2V IMO, 0.8V should be ok.

Your pot behaviour is a mystery to me. No idea. The standard wiring is a voltage divider though, not a (double) variable resistor as in your "wrong" wiring. I have also noticed tone changes without capacitive components in some circuits. Maybe it's due to parasitic capacitance of wirings, transistors etc ... Maybe the amp's input caps are involved, I have no idea TBH

Anyways, cap values is probably the single most important thing you want to play with while you test a circuit. Do not care in the slightest for what others have used in their builds. Tune them based on your guitar, your amp and your fingers.


100R ? I suppose that goes in series with the power. I never noticed a change either with or without, but I'd implement it.
Remember, most things you do in the power supply track is to help with non-normal situations (wrong polarity, too much voltage, non stabilized AC adapter, DC ripple etc etc ...). With perfect DC you wouldn't need any of this.

[r2c2]

thanks for the suggestions, Poiureza !

I think I found the solution for the volume input issue.
- Switched q1 for a higher leakage transistor
- better height setup of strat pickups
- and then probably the biggest reason, which is a bit more embarrassing When playing around with all the components on my breadboard, I must have put the 1M R1 resistor by accident on the q1 collector instead of the base...
I was so focussed on that q1C voltage reading that I missed that q1 base was off...
Correcting these things solved the issue.


It sounded allready great before but now my guitar volume control has become more useful in the interaction with the fuzz...
I think my voltages readings now also match better what is reported here: viewtopic.php?f=10&t=19594
Q1 AC125U Hfe:116 lk:0,211mA
Q2 AC128K Hfe:75 lk:0,106mA
Q3 AC128K Hfe:84 lk:0,085mA

battery-9.2V
resp min to max attack
Q1C: -9.18 -9.18
Q1B: -1.16 -1.16
Q1E: -1.15 -1.13
Q2C: -8.65 -3.80 (now a 220K resistor on q2b..just want to make sure I can reach 4.5V when switching the c3 cap and/or q1 transistor )
Q2B: -0,088 -0.151
Q2E: 0 0
Q3C: -8.89 -8.85
Q3B: -0.015 -0.015
Q3E: 0 0

note: low q3b seems to be in line with other mk1 voltage readings reported. Or do you consider -0.03V vs -0.015V a significant difference for the operation of q3?

c3 cap is now on a switch *(25uF to 0.1uF) and it's really a nice option to be able to switch between solasound/gary hurst flavour of the mk1 (that cap change seems to have the biggest influence on the tone).

I'm still wondering about that level pot 'issue' and why the order of lug2/3 for input/output matters. I prefer it the way around than suggested how it should be connected as it doesn't make the fuzz thinner sounding when reducing level.
For the moment it's on a breadboard and the pots are connected via alligator clip wires, so I'll check if the lug order still matters when mounted in the pedal with shorter and properly soldered wires...
I'll let you know when tested.
Or anyone else has any ideas to explain this behaviour?

Anyway...ready for the next step now, soldering it on a veroboard and mounting it in the enclosure. Wish me luck

[r2c2]

disaster...it doesn't work anymore on the veroboard (layout without trimmers: http://tagboardeffects.blogspot.be/2011 ... r-mki.html)
Everything fine at low attack but as soon as I increase the attack I get alot of screeching noises...
I used the exact same components as I used on my breadboard and triple checked my circuit that all components are in the correct location on the board..pretty frustrating it doesn't give me the same result.
The transistors should be fine as I socketed them. Can a socket cause problems with the behaviour of the circuit?
I measured the voltages of the transistors as well. Q2 and Q3 measure the same but Q1 measure around -1.6V on base and emitter instead of around -1.3V. I expected it to be exact the same as measured on my breadboard.
I tried another q1 transistor that gives a lower voltage (0.6) but same screeching noises..
Is it possible that going from breadboard to vero can alter the voltages? (I'm new to this so have no reference/experience yet with previous builds).
I allready checked for bad soldering and made sure there are no unintended connections between the lanes on the veroboard, but it seems fine. To be sure I reflowed the soldering but it made no difference.
Does it matter where you connect components to the ground of the board? The layout of the vero is of course different than the breadboard. For example, on the breadboard I had the Q1B connected via the 1M resistor to the ground bus and the Q2E with a jumper to the ground bus (Q1B -1M closest to the ground input connection), while on the vero layout the Q1E is connected via the 1M resistor to the hole right next to Q2E (which itself is connected straight on the ground lane , Q2E is now closest to ground input connection). But it's from a schematic point of view still equivalent, right? Or could such a physical layout difference cause different behaviour in reality?
next things I will try
- check the caps, maybe something went wrong when soldering them.
- set up again my breadboard so I can measure and compare things with my veroboard.

Any other ideas, things I can try to find the rootcause of my issue?
Any help would be appreciated so thanks in advance..

Cheers

[poiureza]

Well your Vero obviously has some issues.
The good news is that everything can be solved and it definitely can sound the same as on your breadboard.

Not sure I fully understood how 1M and Q1 are connected on the Vero.
Anyway, you can connect any component anywhere along a path for as long as nothing is put in series in between the locations.
Sometimes it's not trivial to see (especially once you start moving things around in a busy PCB layout).
So, if you're unsure, always ask yourself the question : "will my component see the same voltage if I connect it here or there ?", because it has to.

As for your problem, you really have to insert the actual breadboard trannies in your Vero build. Hfe and leakage are crucial values and no 2 Ge transistors have the same.
Sockets never do any harm. Make sure you pushed the transistor legs fully in.

But it seems you have already put the finger on it : Q1 bias is off. If you used the actual breadboard transistors you HAVE to read the exact same voltages. There is no magic involved.
Thus something close to Q1 is wrong. Check again, maybe you misread the resistor color bands, bad solder joint, touching wires etc ....
It has nothing to do with the actual transistor though : 1.3V vs 1.6V will not make or break the circuit. But it points to the problem being close to Q1

Edit : come to think, you do use accurate passive parts right ? Not some exotic, fancy and expensive stuff featuring +-50% accuracy ...

[r2c2]

Hi,
Super thanks for your input.
For sure i'm overlooking something but couldn't fix it yet.
I used the same components that were on my breadboard (not just the same values but exact same components)
In the mean time I triple checked that everything is on the correct location on the veroboard and measured every component to make sure nothing broke during soldering.
Layout and components are ok (i'll try to post some pictures).
I also wired up my breadboard again (different components but same type and values..except transistors i used the same 3 selected ones). On the new breadboard setup it's working fine..No squeeling and whizzing.
I measured voltages on every location on breadboard and Vero to compare and the readings are the same.
The 1.3 V vs 1.6 is indeed not a problem...Probably just a small difference because of temperature. When I measured the Vero again it was now at around 1.3 too.
I removed the 3pdt switch too just to make sure this is not the culprit but made no difference
The components on the breadboard are much further from each other compared to the Vero. Could that have some effect on a temperamental circuit like the tonebender?
I will replace the offboard wiring ..maybe an issue with one of them. I'll try solid core as well like i did with the breadboard although i don't expect much difference.

What you mean specifically with Q1 is not biassed properly? I used the same transistor (socketted) and get the same voltage readings (expect for some temperature drift) as my breadboard

Resistors and caps are new modern types and should be ok Quality (Panasonic caps and resistors have 5% tolerance).

[r2c2]

I tested now after replacing the offboard wiring but it didn't help so there was nothing wrong there.
But i think I found a solution.
As suggested in another thread a small cap (3 to 10nF) from input to ground helps to tame high frequency screeching.
viewtopic.php?f=19&t=406&p=238243&hilit ... k1#p238243
10nF is definitely too high (the sound becomes dull) but a 1nF cap seems to be sufficient in my case without making it sound dull and still sound plenty trebley enough. Just wondering what frequencies are filtered out with 1nF.
Also playing with other Q1 transistors seems to tame it.
So another Q1 and the 1nF cap solves the issue entirely.
So basically, my breadboard didn't seem to be representatieve enough to be used for making the final transistor selection.
Still don't understand why this is... Resistance of the breadboard breadboard itself has an influence on this sensitive circuit?
Anyway..i can move on ..finetune the transistor selection on the vero and then put back the offboard wiring.
Persistence is key.

[poiureza]

Good to see you sorted it out.

One thing that crosses my mind : I always cut unused parts on Vero i.e. I make holes to only keep the required part of each strip.
I don't know if it really does make a difference wrt antenna and stuff, but it can't harm either.

[deltafred]

Still don't understand why this is... Resistance of the breadboard breadboard itself has an influence on this sensitive circuit?

It is more likely to be capacitive and/or inductive coupling between vero rows than a difference in resistance giving you the high frequency instability (screeching).

High impedance and/or high gain circuits can be susceptible to instability caused by circuit layout. Keeping inputs and outputs as far away from each other as possible, adequate power supply decoupling and low value capacitors on inputs (to limit high frequency response) are all ways of taming it.

[r2c2]

Good to hear there's a logical explanation to this. Thanks, guys.
My build is now 100% finished and it sounds awesome.
I ended up using a 2nf on the input and a 100uF bypass power supply cap. Although I'm using a battery both were necessary.
I'll post some final build conclusions on the tonebender MK1 thread as it might help others.
Cheers!