Experienced MKI builders? How to best approach transistor selection?

Started by Gristlepig, April 13, 2014, 11:16:26 AM

Previous topic - Next topic

Gristlepig

Hey folks,

A lot has been posted about the MKI circuit - with lots of seemingly contradictory/inconclusive info as to best Q1,2,3 gains, leakages, etc....

From what I've gathered, trimpots are to be avoided and a process of swapping transistor after transistor in and out of the circuit is the best way to go.

Piles of transistors later, I've still got varying degrees of spittiness, and general doo doo happening on the decay of this thing. I know this is super common and is the bane of most many a MKI Tone Bender build...

My question is, what is the best was to approach transistor selection for each gain stage in this pedal?

Those who have build several badass MKI's... is there a specific approach you take to this circuit?

A method, perhaps... to maybe minimize the madness?

For example, is is worthwhile to audition transistors in each spot, and then listen to the output of that stage with an audio probe?

I've spent what seems like weeks swapping hundreds of transistors in and out of this piece of vero (it's way too noisy on breadboard) literally wearing out the sockets! I'm trying to take notes as I go along, but they're not providing particularly useful data...  :icon_frown:

I'm not about to give up om this circuit, but I'm really starting to get frustrated!

I'm going by D. Main's Sola Sound Schematic using the 470k collector resistor and a 33uf electrolytic in place of the 25uf electrolytic and .15uf coupling capacitors in place of the .1's. I had the same headaches with the stock values...



The law, in its majestic equality, forbids the rich as well as the poor to sleep under bridges, to beg in the streets, and to steal bread.
- Anatole France

Electric Warrior

What transistors are you using and how do they measure?
I had some problems with my own build having too much of a sizzly decay at some settings, but I guess that might have been normal. Most vintage units only have a narrow sweet spot on the Attack knob without gating or sizzle.
Increasing the resistor between Q2's base and 9V to 470k made it useable over the whole range for me.
Between 180k and 470k seems to be the range that was used in vintage units from what I've seen so far. Apparently they tweaked that resistor for every pedal. I've seen leads of cut out resistors and several resistors in parallel there.

Gristlepig

I've used A LOT of different transistors....

From trios of real-deal Newmarket NKT275's, to Tungsram AC187's...

I've tried combinations of OC's, 2N's, Japanese... Mixing and matching gains and leakages...

I've been focusing on tying to get a set of OC75's to sound good, with hfe/leak 65/220, 114/187, and 87/320... Seems to be right within what many claim to be usable numbers.

But something you just said really got my attention...
"Most vintage units only have a narrow sweet spot on the Attack knob without gating or sizzle"

That little nugget of info has eluded me throughout my reading on this circuit...

I've def gotten some great tones from a very small portion of the attack knob's rotation. Is this the most I should expect from this circuit?





The law, in its majestic equality, forbids the rich as well as the poor to sleep under bridges, to beg in the streets, and to steal bread.
- Anatole France

Electric Warrior

No, you should be able to make it behave better than that with OC75s and 470k at Q2's base. Maybe go a little lower with Q2's hfe. Maybe Q1's too. You could experiment with different leakages as well; luckily OC75s have quite a range.