roland funny cat advice?

[pinkjimiphoton]

hi guys, building a funny cat, waiting on some 23k30a's to come in... but i DO have j201, mpf102, 2n5457, 2n5458, j112's kicking around...
would any of them be a reasonable substitute? or should i just be patient and wait for the proper fets to arrive?

and... any cool circuit perturbations/molestations/mods worth trying on the beast?

it was a "kit" containing a build doc and two pcbs that appear to be based on RG's project on geo, by 3pdt.com which no longer bothers with anything online or anything. he suggested adding a buffer before the circuit to hold oscillations down...

anyone got a decent suggestion for a buffer circuit for this monstrosity? i figure i may as well just build it in, maybe with a switch rather than find room for yet another pedal on my live board if i diggit.

thanks peeps!!!!!

[Kevin Mitchell]

I've used the 23k30 interchangeably with BF245A, BF245B, 2N5457, J201 and 2N4303 without any trouble. Not for this circuit but for others (glides and distortions)

[pinkjimiphoton]

cool, thanks kev! i'll ust socket the sucker and have a play with what i've got.

[rutabaga bob]

Sockets would let you get on with it.  Might sub, might not.  In the Q-Tripper, the subs won't function, you have to use the real deal.  Hope this isn't that way.

[italianguy63]

Guessing.. start with the 2N5457 or the J201.. as they have different gains....

[pinkjimiphoton]

or i COULD develop some PATIENCE and wait a couple days

[italianguy63]

or i COULD develop some PATIENCE and wait a couple days

Screw that.

[pinkjimiphoton]

hahahahahah marc, you know me too well , brother

[Mark Hammer]

The K30 is acting the same way a FET does with the EA tremolo: as a variable resistance that sets gain.  Only in this case, the drain-source resistance is varying at the speed of the ripple in the envelope coming through the diode.  And as is often the case, when there is too much ripple in a gate, compressor, or filter, people often report it as sounding distorted.  Roland/Boss put the ripple to productive use here.  You can vary the audibility of that ripple-driven "soft distortion" with pretty much any FET whose drain-source resistance responds to what the very simple rectifier is feeding the gate.  If you can't hear any distortion, then consider upping the value of the 470k feedback resistor in that stage, or maybe place a 1M fixed resistor in parallel with the drain and source pins.

Having built a few of these, I recommend separating the functions of the rotary switch.  One set of contacts simply selects between three different gain settings for the rectifier that feeds the filter.  Replace the 47k and 22k resistors in that part of the switch with a 15k fixed resistor in series with a 100k variable resistor.  That'll get you all the drive range you need.

The other part of the switch selects between a minimally smoothed and better-smoothed version of the envelope, which you can do with a toggle.  Really, the unit is more complicated than it needs to be.

[pinkjimiphoton]

thanks mark!
i WOULD try that, but already installed the rotary switch to the board, and as i've learned from past outings, un-soldering them can be  a real bummer!
i'll have a play with it once i get it up n running.. i have no prob snipping leads and soldering to them if i have to, i'm fairly old school and grew up reading jack darr

[Rob Strand]

My notes show R10 should be 15k (not 47k).   
Someone confirmed this at one point.   
The main thing to notice is there is a 15k on the original PCB but no 15k on the schematic.

[pinkjimiphoton]

thanks rob, there's no schematic with this thing, of course
i'll have to take a look and see, i'll try and scan the build doc to upload it
man, i musta been half way to zombie-woof land last nite, i stuck the rotary switch in, but never got around to solderin.....ooh, shiney!

[Rob Strand]

A group of us on this group traced through this thing 15 or so years ago,

Schematic and Layout,
http://geofex.com/       entry date 2/3/03
http://geofex.com/PCB_layouts/Layouts/funnycat.pdf

The original unit has funky epoxied modules we worked out what was inside and RG put up  a modernized version with opamps.  Inside the epoxied modules were real transistors and resistors.

http://moosapotamus.net/ideas/roland-funny-cat-pictures/

Just found this one with no epoxy (real or copy/repair?),
http://www.effectsdatabase.com/model/roland/funnycat#
http://files.effectsdatabase.com/gear/pics/roland_funnycat_005.jpg


FYI,  on the non-epoxied modules I think I can see 33k when originally we thought it might be 100k.
Maybe that explains why R10 was left at 47k?
It's a long time ago!

[Mark Hammer]

I hadn't realized that Roland had been using discrete op-amps long before the BD-2.

[Rob Strand]

I hadn't realized that Roland had been using discrete op-amps long before the BD-2.

Another one was the "ROD-10 Roland Overdrive Distortion" from the mid 80's.  It used discrete opamps but not modules. 

The Funny Cat was '73, quite early,
https://www.soundonsound.com/music-business/history-roland-part-1

These things are logic gates from the 60's made my Philips.  They were used in industry.  Check out the size of those things.

When you cut away the epoxy there were BC107's (TO-18 metal cans) and real resistors inside.

[pinkjimiphoton]

of COURSE, the "us distributor" that would have my parts by "friday the 22nd" ...which is the whole reason i chose the damn vendor... is now saying the damn fets won't be here til march sometime.

went back to bitch, they changed the listing. ebay has gotta lay off the damn molly!!

grrrrr....
so still i wait to mess with it. i DO have to scan the docs that came with it and upload, but i am <of course> fairly lazy.

gonna start on either tom's flangelicious flanger tonite, or nick's foxx phaser i guess.
i need a break from that ovation for a couple days, but am so close to finishing up the last details i really should focus on it and get it done so i can close it up and get some video of both it and the standalone fuzzular.

but... see above lol

[Mark Hammer]

I hadn't realized that Roland had been using discrete op-amps long before the BD-2.

Another one was the "ROD-10 Roland Overdrive Distortion" from the mid 80's.  It used discrete opamps but not modules. 

I have an ROD-10, and although it does use discrete o-amps similar to what is used in the BD-2, they are not epoxied modules like the Funny Cat.

[pinkjimiphoton]

ok, STILL waiting for the dang fets. so i tried what's here. it works with 2n5457, 2n5458, j113, bs170.... but none super well.

the volume is pretty low. ya gotta really whack the bejesus out of it to get it to really fire.

i tried a bunch of different fets. also tried them backwards. in every single case, it made the unit come to life fuzz wise, and oscillate like crazy. but right way in, it has no balls at all.

the fuzz side is about halfway to unity.

the filter side is kinda touchy to get it to really quack well. of course, the guy who made the board closed his shit down didn't include a schematic and obviously messed the whole project up... stuff backwards from how it should hook up, etc

he says to use only two transistors, not three... there's pads for the other one but cautions not to use it.

i ended up using a 5532 for the second stage opamp, and a lower gain one left over from the ibanez echo clone i built... name escapes me at the moment.

1458 has better distortion, but won't quack. i tried every dual oa i had here, nothing really special stood out.

looking at rg's schem doesn't help a whole lot, as the guy (3pdt.com) seems to have taken some liberties with the circuit, tho it appears to be based on rg's project

so.....

it needs to be louder... any suggestions?

it needs the fuzz to not suck so badly....

any suggestions for a good voltage range to shoot for with the fet?

will swapping out the 1n914's for ge make the sensitivity work better?

is there a preferable opamp for the fuzz side?

i will scan the build doc and upload it here. its a dual sided board, so some of it isn't much help unfortunately.  as shown in the doc, pots are attached with flying leads. i just soldered on solid core wire to the pins, added shrink wrap and then used the extended 9mm pots to reach the holes in the enclosure. the led's were marked in the build doc as oriented backwards, which, of course, they weren't. so i had to take it apart to fix that part, which was precarious as hell cuz i used ribbon cable which always breaks if ya move it more than a couple times.

but it IS working, and sounds pretty much like shit. the autowah side is ok, nothing spectacular, but the fuzz side doesn't sustain as i expected and sounds a lot weaker and shittier than the demos i've heard.

advice appreciated!!!!

[Mark Hammer]

Keep in mind that different FETs will provide different "on" drain-source resistances.  And since the d-s resistance is part of what sets the gain of the distortion stage (and here I am referring to RG's op-amp based translation of the circuit), if the d-s resistance of the one you're using is high, then the gain will be low, unless you increase the feedback resistance as well.

[pinkjimiphoton]

mark,
is that what ya meant by making the d/g resistor 1m?

or should i go for the weakest fets i have?

remember, for all the crap i've done and messed with, for the most part, i know nothing about electronics, so i appreciate you explaining stuff to my moronic arse!