MKI Tonebender - Gary Hurst Version - excessive sizzle and cackle

[Benjamos86]

Hey everyone! This is my first post on this forum. Hoping you guys might be able to give me some pointers on why I might be experiencing the issues that I am having with my MK I build. This pedal has been kicking my butt and while I'm fairly confident that I just haven't been able to get the proper transistor combo sorted out yet, I wanted to seek some clarification as it relates to this circuit. I've searched and searched for information on things like transistor voltages, target gain and leakage, etc. While there are quite a few threads out there that already broach these topics, it seems there is no "one size fits all" answer. Quite different than something like the fuzz face circuit. I've built the Gary Hurst version of the pedal, and while the pedal does work, I cannot seem to dial out the overly sizzly decay. It's become quite the time consuming and frustrating process of measuring and selecting transistors, plugging them in, only to still have essentially the same result each time I play through the pedal. I understand the MKI is primitive and sizzle/gating are commonly associated with this circuit, but mine seems overly obnoxious and unpleasant. I've owned MKI builds that didn't do this. The fuzz tone is not bad on mine at all, but the way notes fade out really drives me a bit nuts. It has quite a gate to it as well. Things I'd like to seek information on are the following:

What are target voltages on each legs of the transistors that I should shoot for?

Should Q1 base and Q1 emitter voltage essentially be equal? Every time I put transistors in my circuit to test, I do a voltage check first and have taken notes on different combos and values. I am noticing the voltage on Q1 base and emitter are basically equal. Is this normal or indicative of a problem?

What should I be aiming for in terms of leakage and gain on my transistors? I see what small bear shoots for in their MKI transistor combo, but have read forums where people have shared positive results with some values that differ quite substantially from one another. Everyone has their own ear but what's more crucial here? Gain or leakage or both?

I've noticed if the transistors are too leaky, the pedal will squeal and make noise even if I'm not playing or touching the guitar. It seems like the window of satisfactory leakage and gain(more leakage though) is VERY small.

What's the potential main influencer of sizzle on the note decay? It's excessive in my build. I can see maybe a little but it's quite intense with mine. No combo I've put in the pedals has really made a huge impact up to this point. I've measured Q2 collector voltages anywhere from 2v up to 8v, with sizzle being present in all scenarios.

Are there some transistors that just don't work in this circuit? Some people definitely seem to believe so. I have picked through my whole stash which consists currently of:

2N404 and 2N414 (STI)
SFT353
GT402B
AC132
MP16B
3AX31C

So far the best combo I had noted down so far was:

Q1 = MP16B. Roughly 66 hfe, 104 mA leakage
Q2 = SFT353. Roughly 70 hfe, 590 mA leakage
Q3 = 2N404. Roughly 100 hfe, 62 mA leakage

Voltages with this combo:

Q1 collector: 9.18
Q1 base: 0.517
Q2 emitter: 0.47

Q2 collector: 5.81
Q2 base: 0.131
Q2 emitter: 0

Q3 collector: 9.09
Q3 base: roughly 8-10 mV
Q3 emitter: 0

I've seen people suggest q3 needs low to no leakage. I've seen other say this for q2 instead. I've seen others say that all positions needed higher leakage to work for them. The uncertainty with how to approach this circuit is something that I guess keeps a lot of builders away. I'd really like to get this thing dialed in and so I'm hopeful someone here can guide me!

I've tried higher leakage trannies in Q1 and Q2. I've tried high leakage in q1 and low leakage in q2. I've tried higher leakage in all positions. I've tried as many combos as my stash has allowed and am essentially not making much progress. I am considering investing in some more transistors, but wanted to put this thread out to the world first.

I appreciate any tips or suggestions anyone has! Thanks for reading!!

[duck_arse]

welcome to the forum.

what is a Gary Hurst? you would have a circuit diagram you could post here so we know what we are talking about, wouldn't you? also, those transistors are pnp, yes? and the circuit is positive ground, yes? [see why always the circuit diagram?] so we would expect all your voltages to show a minus sign, whish is a good indicator of correct method.

[Benjamos86]

Thank you for the reply!

To my knowledge, there was an original Sola Sound circuit, that was then followed shortly after by the "Gary Hurst" modified version of the circuit.

And you're absolutely right, it does use PNP Transistors and I should have specified the voltages as (-). My apologies!

I'm having quite a difficult time trying to  upload photos to this thread on my phone all the sudden. The "postimages" link keeps providing me with a bad gateway for some reason,

You can see the Gary Hurst schematic showed in this thread:

https://www.diystompboxes.com/smfforum/index.php?topic=102140.0

And the layout I followed is here:

http://tagboardeffects.blogspot.com/2011/01/gary-hurst-tonebender-mki.html

[idy]

Welcome!
People are going to need pics of your build, it's not just a fetish.
The first thing we see, that makes us suspect :pilot error" is the transistor voltages.
The most primitive among us read them and think:

"even I know a working transistor has a diode drop between B and E.'

That would be about .3v for Ge, .6 for Si.

[antonis]

Hi & Welcome..

I am noticing the voltage on Q1 base and emitter are basically equal. Is this normal or indicative of a problem?

For Ge BJTs, V_BE should lie between 250mV and 350mV where for Si ones between 500mV and 700mV..
(of course, the above values depend on various factors like temperature, Collector current, operation region, device power rating, etc..)

In your circuit, those V_BEs may be much lower 'cause Q1 & Q3 are mainly biased via leakage current..
(transistors barely active for quiescent operating point..)

P.S.
Q3 Collector voltage needs to be much lower (in absolute value terms) so you either need a BJT of higher leakage or a bias resistor of value higher than 8k2..

[Benjamos86]

Some more up close images of the pedal are below. Was just test again and still getting the same issues. I don't have any solder jumps between the board traces where there shouldn't be. I've compared my layout against the one I followed in the below picture and it looks correct to that. Still measuring q1 base and emitter voltages very close to one another. Current transistor combo:

Q1 = MP16B, 66 hfe, 104 microamps leakage
Q2 = SFT353, 71 hfe, 594 microamps leakage
Q3 = 2N404, 106 hfe, 62 microamps leakage

Having a higher leakage transistor in Q2 is definitely controlling the amount of noise at idle when not playing. Earlier I tried a combo of:

the  same transistor as above in Q1,
the Q3 from above moved to Q2,
and another MP16B at Q3 which measures roughly 100 hfe, 176 microamps leakage.

The pedal made an excessive amount of noise at idle (fizz, crackles, etc) all without touching the guitar. 

With the first combo described above in the pedal, voltages are:

Q1 collector: -9.14
Q1 base: -0.545
Q1 emitter: -0.492

Q2 collector: -5.71
Q2 base: -0.131
Q2 emitter: 0

Q3 collector: -9.04
Q3 base: -8mV
Q3 emitter: 0

Now, if I put a more leaky device in Q1, another SFT353 with roughly 83 hfe and 295 microamps of leakage, and change Q2 to a MP16B with approx 100 hfe and 175 microamps leakage, I observe the following:

Q1 C = -9.14v
Q1 B = -1.78v
Q1 E = -1.77v

Q2 C = -5.77v
Q2 B = -141 mV
Q2 E = 0

Q3 C = -8.9v
Q3 B = on my meter this value just fluctuates like crazy with this setup. Jumping around from 0 up to around 30 mV. It no longer shows me a fairly steady value as it did when measuring the combo described above.
Q3 E = 0

If I play through this combo, again I get the unwanted sizzle action, you can hear it present even when playing, as if it's below the fuzz tone and as soon as you stop playing, the sizzle becomes more prevalent as the note fades. It's also noisy at idle.

All these values are measured with the attack pot maxed out.

[duck_arse]

I squint closely at your provided, and I feel a little uneasy about your Q1 and 2 pinouts. I can convince myself they are in backwards, but I'm sure you will persuade me otherwise. and they wouldn't mess your Q3 anyway.

can we also see the solder-side, please? and as for postimg - they are very often bad gatewaying me, too.

[antonis]

As long as you don't read and follow suggestions concerning Q3 bias, whatever you do with Q1 & Q2 will walk in vain..

[duck_arse]

I read them, but I don't follow them.

[antonis]

That's why you'll always be down-under..
(or up-over, dependent on beholder viewpoint..)

[Benjamos86]

I will try to get some pictures of the bottom side of it if I have some time later today.

I really don't think the pinouts are backwards or anything like that. Like I said, the pedal works and makes a fuzzy sound as you'd expect, just with it's fair share of unwanted artifacts going on underneath and while notes decay. I had attached a picture of the circuit layout that I followed in my last post.

One thing I will also mention. If I turn the attack knob all the way down, the pedal still works as it should, albeit with a less fuzzy and saturated sound which is to be expected. And the unwanted  sizzle completely goes away.

Note decay with the attack knob down is fine. Kinda makes me go back to thinking that I just don't have enough transistors in my possession to come up with the magic combo yet.

[andy-h-h]

Now, if I put a more leaky device in Q1, another SFT353 with roughly 83 hfe and 295 microamps of leakage, and change Q2 to a MP16B with approx 100 hfe and 175 microamps leakage, I observe the following:

Q1 C = -9.14v
Q1 B = -1.78v
Q1 E = -1.77v

Q2 C = -5.77v
Q2 B = -141 mV
Q2 E = 0

Q3 C = -8.9v
Q3 B = on my meter this value just fluctuates like crazy with this setup. Jumping around from 0 up to around 30 mV. It no longer shows me a fairly steady value as it did when measuring the combo described above.
Q3 E = 0

If I play through this combo, again I get the unwanted sizzle action, you can hear it present even when playing, as if it's below the fuzz tone and as soon as you stop playing, the sizzle becomes more prevalent as the note fades. It's also noisy at idle.

All these values are measured with the attack pot maxed out.

These look about right for the attack on full, with no input signal.  Q2 collector should increase to over -8v, close to -9v (roughly) with Attack on the lowest setting.

As someone who has sat at a breadboard for hours testing transistors on a MKI - it sounds like you have transistors that are not suited for the job.   Apart from that, sockets can become loose - double check that they are firmly seated. 

Good luck

[Benjamos86]

Thanks for chiming in!

It seems from the time I've spent researching, that I am able to get myself into the ballpark on transistor voltages that I have seen others mention in terms of "target values". Q2 voltage does go up to around -9v with attack on minimum. It's seemingly functioning as it should! But it's almost as if I'm compensating between positions to hit target voltages while falling outside of the suggested gain/leakage values for one or more of my gain stages with every combo tried thus far.

Just for clarification, it seems for Q2, the target collector voltage with attack on full should be somewhere slightly above half of the supply voltage. Would you agree something like -4.5-5.5 volts is the target here?

I consistently see people suggesting Q3 collector voltage should be around -8 to -8.75 volts with attack at max. The more leakage in Q3, the lower I observe this voltage to be when measuring. I seem to observe more idle noise and "whistling" when Q3 is higher leakage. Tuning this circuit is like walking on a tight rope!

I am inclined to agree, I simply don't have the ability to come up with the proper combo with my available stash of components, hence why I have decided to buy some more transistors to add to my collection.

I have ordered some more transistors over the weekend. For kicks, I ordered one of the MKI sets from SmallBear. Really curious how that will work out. Also ordered some pretty sweet RCA 2N414 and some Russian MP42B trannies from eBay (we shall see how that works out).

I'll be getting my order from SmallBear this evening so I will be sure to report back with my findings!

[antonis]

Each BJT bias point can be individualy trimmed via its particular leakage and bias resistor value..
(no more no less)

http://www.geofex.com/Article_Folders/zonkmach/zonkbst.pdf

[Benjamos86]

So I received the MKi transistor set from Small Bear yesterday. I was watching my kids solo last night so I didn't have much time to sit down and really take voltage measurements or mess with it too long. But I did throw the transistors in the pedal and plugged it up to have a quick play.

It sounds really great when playing and has more sustain than it ever has up this point, it also feels more easy to play under my fingers (not as intense of a gating effect). But it still has the same sort of sizzle to the end of the note decay that is just not a pleasing sound to listen to. Even more annoying now though is this weird behavior and unwanted noise when not playing the guitar but adjusting the attack knob on the pedal. Like I can turn the attack knob down from max, and then if I turn it back up, the pedal with start to "squeal" or whistle or hum/buzz like crazy. As if the pedal is trying to produce a sound when there is no signal from the guitar. It's pretty intense. I was so frustrated I stepped away from it for the rest of the night.

At this point, I'm questioning my off board wiring. I'll sketch up a schematic at some point that represents how I wired it and share it here. I've built a handful of fuzz pedals. Didn't do anything radically different with this one in terms of wiring.

Will post voltage values and transistors specs later. from memory though they were:

Q1 (OC75) = 80 Hfe, .300 mA leakage
Q2 (can't remember transistor number but it was in a TO-5 package) = 110 Hfe, .03 mA leakage
Q3 (same as Q2) = 105 Hfe, .05 mA leakage

What about the resistor from Q1C to Q2B? Would this have any effect here? I used a 470 Kohm instead of a 180k because it seemed people expressed that the 470k made the pedal easier to bias. Seems some people suggest the 180k increases gating and reduces sustain, while the 470k increases sustain but can effect the sizzle on the note decay.

I have this circuit breadboarded right now to so I intend to go back to tweaking on there. Just trying to see what other suggestions anyone might have! Really appreciate all the input so far.

[duck_arse]

the hurst variant shows 470k, and you built the hurst, didn't you say? that resistor sets the bias for Q2, along with the 2k2 to ground - assume the 50k pot is to ground. [it doesn't connect Q1C to Q2B, it connects Q2B to supply.]

if you've changed the transistors, we want a new photo. we also want a photo of all your offboard wirings, please.

[Benjamos86]

Yes I did and you are correct, it does show the 470k in the Hurst version circuit. I'm just pulling for straws I guess, trying to hone it on all things that could be at play here.

I will sit down tonight and sketch up how all the wiring is done and post more pictures with the new transistors installed!

Thanks all

[andy-h-h]

Have a careful listen to this clip and note the tail end decay    https://www.youtube.com/watch?v=MbZGYMPqu-4&ab_channel=pinstripedclips      also have a close listen to this  https://www.youtube.com/watch?v=YJFnV3Eqy_s&ab_channel=GrahamGreen   

Does it sound like either of these, as this is not uncommon on a MKI.    While mine don't tend to squeal, there can be some noise when moving the attack while not playing.  It settles down within a second or so.

[Benjamos86]

Ok guys, sorry it took me a few days to reply back. I listened to both of those clips and it does exhibit the behavior from the second clip showing the original pedals. I had come across this video myself a week or so ago and thought it was interesting that my pedal was behaving in a similar manner.

With the transistors in place from the SmallBear set, the pedal has sounded it's best with the most sustain and best fuzz tone in general.

I did have some time to take some pictures of the pedal over the weekend and also sketched up how it was wired. Take a look and let me know if anything looks incorrect or not wired properly. I am pretty confident it is wired up fine.

I am coming to the realization that it might just be what it is and that's it.

At this point, what might make sense it to post a video to YouTube and let you guys hear it.

[Benjamos86]