About to build a Tonebender MKI. Any advice?

Started by rousejeremy, February 07, 2013, 04:40:18 PM

Previous topic - Next topic

Arcane Analog

It all boils down to whether you are happy with "in the ballpark" or if you want to achieve a certain sound. I go through dozens of transistors. Trim pots only prolong the inevitable fact that you need to audition lots of transistors. I find OC75s are the easiest to dial in for this circuit. No trimmers required.

LucifersTrip

I have to disagree about the trimmers. If you understand what's happening, they are your friends.

Here's a perfect example (start 1:22):
http://www.youtube.com/watch?v=jkn29p1aLdI&list=UUqQDtCG7rijf_O408zPaqpA&index=42

The first thing you need to know about Q2 & Q3 is that they are biased very standardly. Nothing unusual....and just like many other fuzzes, the hot spot will be at around 1/2 the supply voltage (~4.5 - 5v). But, with the MKI, you don't want that. You want it almost misbiased, much nearer 9V (8V+) so you get that dirty gating effect (the most common cause of gating in circuits like this is misbiasing).

The next thing you need to know is that the higher leakage the transistor in (Q2/3), the lower Q2/3C voltage....and the lower the collector voltage, the hotter, the sound...and the hotter the sound, the more noisy it will be. That is precisely why it is recommended not to use a high leakage Q3...unless you trim it, of course. For example, in my last one that I wrote about earlier in this thread, I used Q3 = 2N1038 (54 / 380uA). The voltage was ~ 7V on Q3C,  so I adjusted it up to ~8V (Note: just realized that was with a weak battery...with strong battery, closer to 8.5V).

In the video,  blondegraemey has a trimmer on Q3. You have 2 choices. If you sub it for the 15K, you can alter the bias much quicker, but I have found you can't get it to gate as well. I found a trim for the 8.2K is better to tune in the gating. What blondegraemey did was adjust one of those 2 resistors to raise Q3C voltage closer to 9V till the noise disappeared.

===============

A trimmer for Q2 is not commonly used since you're setting the bias with the attack pot, but you still can use a trimmer for the 1.8K limiter to set the lowest attack. Note that the max attack Q2C voltage is different when using a 470K or 180K. With the 470K, you wind up closer to 5V and with the 180K, you could get closer to 2V and even lower. When you go past the "hottest" spot (~ 1/2 supply), the sound will get thinner and more compressed. It's your choice. Again, you can set the max attack anywhere you want by changing the size of the attack pot. Just like with Q3C, the larger resistance from base to ground will lower the Q2C voltage...and a Q2 with too high leakage will also lower the Q2C voltage and will set the the low attack setting too hot. You probably want it to be 8V+ at low attack.

===============

A trimmer on Q1 can be used to set how hot (gain) the overall circuit is. I rarely use that since it's the easiest of the 3 transistors to choose, but hfe seems to matter as much or more than leakage.  The higher the hfe, the hotter and noisier the circuit can be.  You can choose one in the classic 50-70 range but you can always trim it to produce less gain and the circuit will be less noisy....especially if you chose a Q1 with too high hfe.

Here's a great post from 2009 that Arcane quoted some info from.
http://stompboxes.co.uk/forum/viewtopic.php?f=10&t=37

Dragonfly
"A trim pot on the Q1 emitter definitely influences gain, and would allow you to use a wider range of germaniums. You should be able to "get away with" using a little bit higher gain device. I agree that Q1 likes that 50-70 range (if you're not using a trimmer)."

===============

The bottom line here is not to depend on trimmers too much, but to use them to tweak to exactly where you want the sound. In other words, you shouldn't have to tweak a trimmer too far from the stock values.
always think outside the box

Arcane Analog

#42
If you look at many the professional builds there are no/few bias values tweaked. You do not need to do that. They simply dropped in transistors that worked. David Main, PigDog, Castledine, SonicVI, Germs - none of those gents use trimmers and they do not drastically tweak - if they even do any tweaking at all - any of the resistors you have trimmers on.

What Lucifer seems to skip over is that some transistors simply will not work well in the MKI no matter what you do or how you try to tweak it. Period. He can try to throw out as much half-understood theory as he can muster but the bottom line is that you need to audition transistors. You are experiencing it yourself in that you have tried tweaking that transistor selection with trimmers and you still are unhappy with it. You can hit the volatges exactly as Lucifer is preaching and still not have it sound properly biased. That the way the MKI is. Breadboard the circuit and swap transistors until you have it where you want it and go from there. You could be playing transistors that will never get the sound you want.

All of these quotes are found in the FSB thread and none are by me:

- Don't bother with the trimpots, you'll drive yourself nuts. Just plug and play until you get a good transistor trio. I've built mine all stock. (Phibes)

- In case of MKI you cannot depend on numbers. It's bitch to bias it properly... and usually with trannies in correct gain/leakage bucket it sounds like sh*t... only swaping transistors till it sounds good could help here IMO. (Beedotman - Turretboard.org)

- I'd just breadboard, no need for a test bed. Tone Benders are pretty low are the part count so it should only take 5 minutes to put the circuit together. With Germaniums, it's a MUST to test run them before soldering them to a board. Having say 15 good transistors to try is better than just going with the 3 and crossing your fingers. (Phibes)

- Stock Mk-I circuit. Tone controll is a 100K lin pot that blends in a 22uf in parallel with the 10nf output cap.
Biassing consists of about 10 hours of transistor swapping and a box of OC75's. (Radiotron - Mr Morphine)

rousejeremy

You mention OC75's a lot. Unfortunately the only Mullards I have are a couple OC140. I appreciate everyones input here. Once this pedal is finished and out the door I'm deleting every MKI schematic and layout I have.
Consistency is a worthy adversary

www.jeremyrouse.weebly.com

Arcane Analog

I never said you need Mullard OC75s. I also mentioned AC125s work just as well. They are cheap and plentiful. You can buy bags of 150 pieces for $60.

If you follow my advice the circuit is not nearly as difficult to build as you are finding.

I know you are getting voltages, trimmer, etc suggestions. Ignore them. This circuit is not that difficult to build.

You can build MKIs on breadboard with stock values and plug in transistors that you have sorted for gain/leakage until you get a trio you like. You are not going to have success sticking to a few transistors and trying to make them work. Plain and simple.

Arcane Analog

Also, plugging in transistors that you have measured will teach you alot more than simply twisting a trim pot to hit voltages that may or may not work.

The glaring fault in Lucifer's advice is that this is not your typical circuit where relying on voltages will ensure that you get a decent sounding pedal. That is just a crutch for the MKI that you need to toss. It is good as a reference but that is about it. I build them with zero attention to voltages and I know several other people who do the same. I do record them when I am finished.

You can get the exact voltages of the best sounding MKI on the planet dialed into your circuit and it might still sound bad. The interaction between leakage is paramount in each stage and each germanium device will react different - even when they have the same HFE/leakage properties and even if they are the exact same brand/type of transistor.

No amount of futzing with trimmers and voltages can make up for that. Some transistors will simply not work in this circuit.

LucifersTrip

I guess you didn't read what I wrote:

"The bottom line here is not to depend on trimmers too much, but to use them to tweak to exactly where you want the sound. In other words, you shouldn't have to tweak a trimmer too far from the stock values." You still need to find transistors that are close to what you're shooting for.

All the info I gave is not half understood theory. Those are simple facts about how transistor voltages & leakages relate to gain in this circuit that would definitely help anyone that hasn't built this circuit loads of times.

There are just as many good builders that use trimmers/non-stock values as ones that don't. This can be said about almost any circuit. If the builder is fortunate enough to have enough supply to choose from, yes, of course, you won't need trimmers. The builders that don't use trimmers are mostly sellers/dealers/boutique guys who need to sell their product as exact clones.

Remember, you showed numerous videos above with varied amounts of gating. That is the perfect reason to use a trimmer or an external bias pot.

always think outside the box

reverberation66

I think ultimately lucifer's trip and arcane analog are both right.  I've built several of these and trimmers can definitely be useful in dialing in a good sound, especially if you have a limited selection of transistors to choose from.  I also feel that you can go with the stock values and  if you have a good selection of transistors you can play around with it a bit and find a combination that will sound good as well.  This is a notoriously finicky circuit and will probably require some fine tuning no matter how you build it.

LucifersTrip

the funny thing is that many times I've trimmed the circuit within the error of the resistor values.

for instance, an 8.2K with 5% tolerance can be 7.79K to 8.61K and with 10%, it can be 7.38K to 9.02K

it is very common that my Q3 trimmer is set around 7.5V.  it just needed that little bit to tweak out the unwanted noise and bring the voltage closer to 9V
always think outside the box

Arcane Analog

Quote from: LucifersTrip on March 08, 2013, 05:05:53 PM
There are just as many good builders that use trimmers/non-stock values as ones that don't. This can be said about almost any circuit. If the builder is fortunate enough to have enough supply to choose from, yes, of course, you won't need trimmers. The builders that don't use trimmers are mostly sellers/dealers/boutique guys who need to sell their product as exact clones.

Remember, you showed numerous videos above with varied amounts of gating. That is the perfect reason to use a trimmer or an external bias pot.



Gating is simple transistor selection.

Who are all of these quality builders you speak of that make MKIs? I made a list of builders that have videos available as reference. None of those builders except DAM make production runs. They are all built to order and most do not make replicas.

Arcane Analog

#50
Damn phone cut off half of my response.

Thunderbird, Radiotron, etc. All excellent. All no trimmers. Jimmy Behan does one sometimes with an external pot. I am interested in your quality builder list.

LucifersTrip


Quote from: Arcane Analog on March 08, 2013, 07:15:13 PMThunderbird, Radiotron, etc. All excellent. All no trimmers.

so if instead of  using a trimmer set to 7.5K to put Q3C exactly where I want it, I went thru a bucket of 8.2K's until I found one that measured 7.5K, I can claim I didn't use a trimmer?  genius

Quote
Gating is simple transistor selection.

gating = biasing the transistor. do you think the MKI will gate more or less if you choose a transistor that puts C voltage closer to 9V?

QuoteI am interested in your quality builder list.

why would I spend time looking thru 5+ years of dam, fsb & youtube to find vids for a single person thinks he can judge which is better?
always think outside the box

Arcane Analog

My list was from the top of my head. I expected the answer you provided.

LucifersTrip

always think outside the box

Arcane Analog


Arcane Analog

Quote from: reverberation66 on March 08, 2013, 06:06:29 PM
I think ultimately lucifer's trip and arcane analog are both right.  I've built several of these and trimmers can definitely be useful in dialing in a good sound, especially if you have a limited selection of transistors to choose from.  I also feel that you can go with the stock values and  if you have a good selection of transistors you can play around with it a bit and find a combination that will sound good as well.  This is a notoriously finicky circuit and will probably require some fine tuning no matter how you build it.

What happens if you have a device that falls well within the established HFE and leakage parameters and biases up where Luciferbsays it should land yet still sounds shitty? This happens alot. This is why trimmers will just prolong your aggony and drive you nuts. Some transistors will not work no matter what you do. Voltages be damned. You can set the voltages this way and that for an eternity and get nowhere simply becausenthe device you are working with sounds like trash in the circuit. Trim the bastard to death - it wont help you.

Preselect a pile of transistors and swap until you are happy.

LucifersTrip

#56
QuoteSome transistors will not work no matter what you do. Voltages be damned.

Of course...that's true with many fuzzes.  Did I ever say that you could use any transistors? you have to get close! then, tweak the gating, noise, sizzle, decay, etc...

"The bottom line here is not to depend on trimmers too much, but to use them to tweak to exactly where you want the sound. In other words, you shouldn't have to tweak a trimmer too far from the stock values."

Quote
Preselect a pile of transistors and swap until you are happy.

if you repeat that enough times, it will be true

http://www.youtube.com/watch?v=jkn29p1aLdI&list=UUqQDtCG7rijf_O408zPaqpA&index=42
[1:22]
always think outside the box

Arcane Analog

#57
Still no list Lucy? Lets here a video of those excellent sound builds you have. After all there are so many.

LucifersTrip

the discussion is whether a trim is helpful.  you have (1) a clear video above showing a perfect use to kill noise and as you mentioned earlier (2) Jimmy Behan uses an external bias...

there's also (3) ghosteffects:
http://www.youtube.com/watch?v=p9mT-61E4uY

and of course, there's (4) fuzzfaceless, who uses 2 bias controls:
http://www.worthpoint.com/worthopedia/tonebender-mk1-fuzz-pedal-166580510

and how many builders use trimmers on the breadboard, then sub a fixed resistor?

whether you choose to put those controls on the inside as a trim or outside is up to you. I do both or neither...
always think outside the box

Arcane Analog

#59
I have already awknowledged people use them for fetaures including jimmy b :icon_mad:y name. Those are not necessary uses though. They are features.

Funny. Graham is using bags of AC125s and OC75s. I bet he didnt even need the trimmer.

Strange that the original poster has trimmers and cannot get his dialed in.

Anyway. Everyone knows you are the self proclaimed fuzz king. All hail Lucy.