About to build a Tonebender MKI. Any advice?

[rousejeremy]

I'm using Ivlarks vero layout here http://tagboardeffects.blogspot.ca/2011/01/tonebender-mki.html
From what I have read so far, biasing can be difficult with this circuit. Since his layout is using trimmers I assume it will be much easier.
I have searched the forum and read a little about using slightly leaky transistors in certain spots, and various opinions on gain ranges for Q1, Q2 and Q3.
Can anyone offer some advice on bias, leaky germaniums and gain for this circuit?

[Arcane Analog]

If you have a lot of germanium to choose from skip the trimmers and build it with stock values. It is all about the germanium/leakage and the interaction of the stages. The trimmers are not going to dictate that interaction properly and they are just going to complicate the build. If you have them OC75s and/or AC125s work well in this circuit. I have built a few of these and it became alot easier after I scrapped the trimmers.

Build it on breadboard and swap your transistors until you get close. Make lots of notes on the combinations you like and keep swapping transistors with the same or very close HFE and leakage. You can get a beautiful sounding MKI and then copy the transistor selection with new transistors having the exact same HFE and leakage and it will sound completely different. 

Dropping the trimmers and auditioning alot transistors is the best advice I can offer.

[rousejeremy]

Thanks. What about voltages? What should I be aiming for?

[Arcane Analog]

The FSB thread on the MKI is a good place to find information.

[mac]

Read about the zonk machine at GEOFEX.
YOu have to experiment with different transistors at Q2 and Q3 until you get a set that sounds good.

mac

[LucifersTrip]

It's funny. There are long threads on various forums with no conclusive ways to build this one well. You usually get the standard "try a bunch of combos until it sounds right" comments.

Though, after building a bunch that sounded excellent, I was going to post a "guide" with some explanations...but the response for vintage stuff here is so underwhelming, I didn't...

I can still remember much, so I'll try to help from memory.

I think the first misconception is that it should sound like an FZ-1 or 1A. It doesn't.  Secondly, many shoot for certain gains in the different slots (a common set would be 60 / 100 / 60), but I have found that leakage is far more important as long as the gains are 50+...and as usual, the model # transistor means almost nothing. I almost always use transistors with leakage of at least 200uA for the 3 positions.

The first thing I do is simply breadboard stage one and play thru it. You should get a decent clean boost that doesn't sound misbiased. You want the notes to fade out nicely.  I've seen a large range of voltages for Q1E. The last one I built I marked down as .6V.

Secondly, a load of discussion goes into biasing Q2 and whether to use a 180K or 470K on Q2 base. It's no big deal whichever you use. When you're at lowest fuzz setting, Q2 collector is close to 9V and when you increase that fuzz pot, the voltage drops. The max fuzz is usually in the 3-5V range. If you use 180K, you'll probably hit 3-5v at only 5 - 15K instead of at 50K, so all you have to do is replace the 50K pot with a 10K one...very simple. Look at the Zonk. They only have around 3K there!
http://fuzzcentral.ssguitar.com/schematics/zonkIschem.gif

With a 470K, you will be able to use most or all of the 50K pot range to get to 3-5v. So, just measure the K where it's max fuzz (around 3-5v) and put the exact value pot there (use a resistor across the lugs to get the exact value you want...remember, it's only a variable resistor).  If you drop the C voltage much lower (1-2V and lower), you'll get a thin, synthy, gated fuzz...and it will also get noisy sometimes <5V, so set the max fuzz where it's cool for you. If you pick a Q with the wrong leakage, C will be too high when fuzz is at minimum, so you must choose one that sounds good at min (C voltage not too high).  With 180K, the last one I set max at 4V with a 7K pot. At min fuzz setting, it was in the high 8V range (I think ~ 8.75 - 8.9V).

Q3 works similarly to Q2, but the C voltage is much higher. I have found it's best in the 7.5 - 8V range. So, I just choose a transistor that gets me in that range using the stock 15K. The leakage is usually 300uA+. I usually use a trimmer instead of the 15K to fine tune it, but you want to be close to 15K. If the transistor doesn't have enough leakage C will be closer to 8.5 or 9V and just like Q2, won't fuzz well and you'll get more gating. The last one I set to 7.75V with the trimmer at 13K. Set the voltage too low and the volume will drop and the tone thinner.

After I've chosen Q2 and Q3, I always throw a few different Q1's in with similar leakages, just to see if it sounds any better that my initial choice.

I still have these handy:

2N1144 (72 / 210uA)
2SB376 (69 / 250uA)
2N1038 (54 / 380uA)

[Arcane Analog]

YOu have to experiment with different transistors at Q2 and Q3 until you get a set that sounds good.

All three transistors - not just Q1 and Q2 - need to tailored to the circuit.

I agree on leakage being the paramount factor. I think the reason that there is no clear cut maner to do this is because each transistor works and sound differently in the circuit. Even transistors with almost idetical HFE/leakage cannot always be swapped out. Lucifer, I am surpised that the unit you built with those numbers wasn't a crackling mess on the fade out based on my eperiences. I like Q1 to have the most leakage and Q3 the least for the circuit to minimize or completely remove the crackle and sizzle. It just goes further to my point that the individual transistors are the key and set gains/leakages are not necessarily going to get you close.

[LucifersTrip]

YOu have to experiment with different transistors at Q2 and Q3 until you get a set that sounds good.

Lucifer, I am surpised that the unit you built with those numbers wasn't a crackling mess on the fade out based on my eperiences.

nope...long smooth sustain and nice decay. you will get noise if you set max fuzz at too low voltage for q2, set q3 voltage too low or if you don't have a nice decay for the q1 stage by itself (which is why I test that first before adding q2 & q3).

remember, closer to 9v for q2/3 is less fuzz and gatey, so if I really wanted less noise, I'd just set max fuzz to 6v and I'd set q3 C voltage 8.5V+

[rousejeremy]

This is good. Although there are tons of thread talking about the various important points behind the circuit, it's nice to get them all in one.

[Arcane Analog]

nope...long smooth sustain and nice decay. you will get noise if you set max fuzz at too low voltage for q2, set q3 voltage too low or if you don't have a nice decay for the q1 stage by itself (which is why I test that first before adding q2 & q3).

remember, closer to 9v for q2/3 is less fuzz and gatey, so if I really wanted less noise, I'd just set max fuzz to 6v and I'd set q3 C voltage 8.5V+

Long and smooth sustain is not really a characteristic of a vintage MKI. What type of gating do you have?

[LucifersTrip]

Read the history at DAM
http://www.stompboxes.co.uk/History.html

"The MKI was a three transistor circuit that was based upon the Gibson built Maestro Fuzz-tone that was modified by Hurst to produce more sustain than its American counter part."

As I wrote above, I can easily dial in gating by combinations of increasing the voltage on Q3 and turning down the fuzz

[Arcane Analog]

Read the history at DAM
http://www.stompboxes.co.uk/History.html

"The MKI was a three transistor circuit that was based upon the Gibson built Maestro Fuzz-tone that was modified by Hurst to produce more sustain than its American counter part."

As I wrote above, I can easily dial in gating by combinations of increasing the voltage on Q3 and turning down the fuzz

More sustain than the FZ-1 doesn't really have any bearing on what I said and that comparison is of zero relevance. 

You can tune your MKI to sound more like a MKII but you are entirely missing the point.

[Arcane Analog]

As I wrote above, I can easily dial in gating by combinations of increasing the voltage on Q3 and turning down the fuzz

And for the record, the voltage on Q3 does not fluctuate at all - just a tiny bit. The voltage of Q2 is the one that changes alot with the fuzz pot.

EDIT:

MKI: Primal, gritty, spitty, jagged and a little tasty gate. Not a long and smooth sustain - it last for a few seconds and then fades with just a little bit of crackle.

Long and smooth sustain is a MKII. You can certainly try to tune more sustain into a MKI but that defeats the entire point of making a MKI.

[LucifersTrip]

As I wrote above, I can easily dial in gating by combinations of increasing the voltage on Q3 and turning down the fuzz

And for the record, the voltage on Q3 does not fluctuate at all - just a tiny bit. The voltage of Q2 is the one that changes alot with the fuzz pot. Perhaps you need to revisit the circuit.

What are you talking about? I only said that I can adjust the voltage of Q3 to get any type of gating I want. I've seen Q3 C voltage reported anywhere from ~ 7.5V to 8.5V+ ...and guess where you can dial in more gating? 8.5V

The question you should be asking yourself is:
How much gating, if any did the original designer intend...and can you prove it? 

You can tune your MKI to sound more like a MKII but you are entirely missing the point.

Mine sounds nothing like a MKII. Missing the point? To make a MKI that sounds exactly like everyone else's or to copy some ideological tone you've heard somewhere...maybe at DAM.  I've heard demos of original MkI's and Zonk machines with tones and gating amounts all over the place.

[Arcane Analog]

Haha. Wow.

[LucifersTrip]

EDIT:

MKI: Primal, gritty, spitty, jagged and a little tasty gate. Not a long and smooth sustain - it last for a few seconds and then fades with just a little bit of crackle.

Right...I can easily tune mine that way....just did. set Q3 to 8.75V w/ 9.6 supply.  But again, what makes you think all vintage MKI's sound that way when the circuit is so finicky with transistors?

[rousejeremy]

Well the first batch I tried out were all around 100, as recommended in the Zonk Machine article on Geofx (thanks for that tip btw). I've got it tweaked with the trimpots (by ear) and it's sounding pretty good. It sustains and then the notes kinda trail off in what sounds like intermodulation or something.

I'll experiment with different gains a little later.

[Arcane Analog]

Thanks for the offer Lucifer but I think I have better things to do that roll around the mud with you.

[LucifersTrip]

These are not originals, of course...but Brad usually knows what he's doing.

Are these poor MKI's because they have longer, smoother sustain?   Before I set Q3C to 8.7V, these are very close to what I have.

http://www.stompboxes.co.uk/History.html

"The two pictured MKI's on this page when side by side do sound different."

"The output level is well over double that of the other MKI and when correctly dialled in has fierce amounts sustain with truly astounding clarity for such an old device."

[Arcane Analog]

Edit: Not worth it.