Page 5 of 6

Re: Paul_C's 2019 builds (having fun!)

PostPosted: Fri Mar 22, 2019 7:20 pm
by BetterOffShred
MPSA18 goes up to around 1k, most of mine are like 800.. bc550s will be 500 to 800..

Most stuff is below 1k aside from Darlington. You can make Darlington pairs using 2 transistors, but I suspect not much to gain there ( pun intended).

You either built the board wrong or got a bad part at this point on the OK, and it looked OK to me (pun intended). I gotta build it now to see if that layout is broken :P

Re: Paul_C's 2019 builds (having fun!)

PostPosted: Fri Mar 22, 2019 7:59 pm
by Paul_C
If the circuit is expecting to see an hFE of say, 800+ from the two MPSA18s, might it be an idea to order some more 18s and see what I get, or would it make more sense to change one of the other components to compensate for a lower hFE? Maybe swap the 3M3 resistor? (feel free to correct me if this is a ridiculous idea - I'm here to learn :) )

I tested all my 18s with the tester you suggested I get, I assume you test your transistors with the same device ?

Based on my very limited understanding of what I'm doing, could the lower hFE of my transistors be why I could only get a signal by hitting the strings hard ?

Re: Paul_C's 2019 builds (having fun!)

PostPosted: Fri Mar 22, 2019 8:09 pm
by BetterOffShred
Most Devi circuits sound better with lower gain. She just got into the MPSA18 and never looked back (IMO) I usually swap at least one for a 550C in the 600 range and it cuts out the hiss and crackly piss noises. I doubt very highly the circuit needs 800+ hfe

Yes buy that tester and it'll work great for everything but Germanium because it does not calculate leakage.

I don't have the schematic in front of me because I'm on an airplane, but I'd wager something is amiss with a component.

Re: Paul_C's 2019 builds (having fun!)

PostPosted: Sat Mar 23, 2019 11:45 am
by BetterOffShred
Yeah so I built the OK using the same layout you did and it works. With the control all the way to the 10n it's a little splatty but not by any means low output and it transitions up to full roar gain towards the 100n .. I was too lazy to move everything and get out my 2907s so I used a 4403 I had sitting there, maybe a little lower gain in that one. I'll check the 2907 later, but if anything it will probably be higher output. Used 18s where they are called for..

Not sure what's going on with your build dude, wish I could be of more help

Re: Paul_C's 2019 builds (having fun!)

PostPosted: Sat Mar 23, 2019 1:18 pm
by Paul_C
Not to worry !

I'm tempted to build half of the circuit to see what happens (removing the section that has the 3M3 resistor, one of the MPSA18s and a 100nF cap).

When I was testing transistors the other day I discovered it still works without that section (with Q1 removed from its socket) so it might be interesting to see what it sounds like with all the components taken out.

I won't give up though, working through different things until I get it working will help me learn what's doing what, and why something isn't doing what it should. I'll breadboard it next time so as to take the risk of unwanted connections away.

I'll record a clip of what I've turned mine into so you can hear it :)

Re: Paul_C's 2019 builds (having fun!)

PostPosted: Sun Mar 24, 2019 5:09 pm
by Paul_C
Here's a clip of the Ok-ish circuit.

It's running through my pedal chain but it's the only one on - you can hear it reacting to the input, speeding up with a properly hit string and slowing down as the note decays.

I don't mind what anyone else thinks of it, I really like it :)

Re: Paul_C's 2019 builds (having fun!)

PostPosted: Sun Mar 31, 2019 5:37 pm
by Paul_C
Update: I've finally settled on four transistors in my Random Number Generator which make it work :)

It's been a bit of a struggle because it worked from the start, but not very well, so I've been going back to it from time to time and trying random combinations to see what happens and occasionally my lack of proper knowledge took me down some interesting but unsuccessful paths (putting J-FETs and MOSFETs in because they all look the same ;) for example ).

Now I've tested all my transistors and noted their hFE it's given me a little more idea of what I'm doing (or at least what I THINK I'm doing) and gave me the knowledge that some of what I assumed would be similar (due to them being the same identification code) could be quite unexpectedly different, for example I have two sets of 2N4401s one set the hFE measures 169, the others are 430.
It turns out the lower hFE works far better as Q1 than the higher number (which I think I used first) and with that in place it all came together quite quickly.

As I've said before, while it's been a bit of a pain to get it working it has taught me a lot more than a pedal which worked first time :)

Re: Paul_C's 2019 builds (having fun!)

PostPosted: Mon Apr 08, 2019 5:12 pm
by Paul_C
I've still been playing with the transistors in the RND and "finally" settled - I should write them down sometime.

My Parallel Universe is working, but not properly so it needs looking at - I read on the Parasit page that they can be sensitive to which LM386 is in (I tried one that came with something else and it sounded worse, which suggests that may well be true) so I've ordered 10 to try swapping.

After all that fun, and with a build (wish)list of around 20, I spent the last hour building another OK board (I couldn't walk away without making one work) using one of the first two vero boards that failed to work. The only thing left on it was one jumper and two resistors and I used three sockets for transistor swapping.

It worked !

I'm going to play with some more transistors (Q1 currently has an hFE of 50 instead of the 430 I took out and it's still making a decent noise) to see which position has the most effect on the circuit.

Either way, I can rest easy now I've produced a successful board - I'll have to try the original transistors to see if it then fails to work, so I can get a better idea of why it didn't work first time.

Re: Paul_C's 2019 builds (having fun!)

PostPosted: Fri Apr 12, 2019 9:41 am
by Dowi
The OK is one of the first things on my list too and your experience with it is gonna be super useful. :thumb:
I m gonna have the parts here in a couple of weeks (at least, I hope so, eventually I m gonna find a European seller so I don't have to wait this long).
Meanwhile I started to try a couple basic things and mess around with the breadboard, just to see if I am understanding what I'm studying. :whateva:

Re: Paul_C's 2019 builds (having fun!)

PostPosted: Wed May 01, 2019 5:58 am
by Paul_C
Update: my pedal chain now has seven self-built pedals on it, with another one (something "completely" new) waiting to be boxed up.

I'm completely happy with the OK now, which is where it all started, plus an Electric Brown, Parallel Universe and Hyperion 2.

I've also got the Random Number Generator and Finish Line built, but I'm still tinkering with them as I feel they're not quite there yet, though they both work ok.

Last of all is the Happy Little Accident which is a spin off of the OK, which I still like a lot.

The new one is a further exploration of a cut-down OK circuit which I called a Teeny Tiny Fuzz - it wasn't anything spectacular but it worked :)

I made another Teeny Tiny Fuzz board, but without the first pot and its two capacitors - cutting things down even more, then connected that to the original board. I've drawn up a vero layout for it (on paper, which is the best I can manage for now) which looks like this:


I started with both of the transistor pairs identical, but there was a lot of noise so after some swapping settled on the pairings above.
The "Fat" control goes from bright and a little less full through "a bit of a starved sound if you play lightly" to a fatter fuller sound when full up.
The big thing about this is that no matter where the Fat control is, it needs a reasonable amount of signal to work so chugging chords can easily be separated from each other, which is the bit I like about it. Once the signal drops the output cuts out completely so it has its own noise gate built in.

It's not the greatest thing in the world, but I like the sound of it and as my aim was to build something with a fairly small number of components I consider it a success :) I'll post some sound clips once I've boxed it up.

Re: Paul_C's 2019 builds (having fun!)

PostPosted: Wed May 01, 2019 6:08 am
by Paul_C
I've called it Big Brother Fuzz until/unless I can think of something better as it's the bigger brother of the Teeny Tiny ;)

Re: Paul_C's 2019 builds (having fun!)

PostPosted: Fri May 03, 2019 9:01 am
by Paul_C
The above picture is "nearly" right - the lettering for the two 2N2907s is correct, but as they're PNPs the flat faces should be facing the opposite way :)

Here's a sound clip of me playing it into my Boss Katana on the clean channel, starting with the Fat control fully CCW and ending with it fully CW.

I like it a lot, and it sounds way more interesting than the Finish Line I built ;)

Re: Paul_C's 2019 builds (having fun!)

PostPosted: Fri May 03, 2019 10:12 am
by BetterOffShred
Just fyi, pnp/npn doesn't really dictate the pinout, so make sure you always consult the data sheet unless you know for sure.

Re: Paul_C's 2019 builds (having fun!)

PostPosted: Fri May 03, 2019 11:04 am
by Paul_C
BetterOffShred wrote:Just fyi, pnp/npn doesn't really dictate the pinout, so make sure you always consult the data sheet unless you know for sure.

well that's annoying - is it done deliberately just to confuse me ?


Re: Paul_C's 2019 builds (having fun!)

PostPosted: Fri May 03, 2019 12:00 pm
by BetterOffShred
I am not sure of the actual design intent behind it, but it's just one of those things I've learned to check every time. Nothing worse than thinking you did something wrong the whole time and finding out you just had your pinout wrong.