Buchla 258b Clone Project

Peake · Posted: Fri Aug 15, 2008 2:07 pm Post subject: Buchla 258b Clone Project

Seeing this posted on matrixsynth shows me that there is interest, even only to see whether I can pull this off or not.

http://electro-music.com/forum/topic-26796.html

As I've said, there are two cloning projects going on at the moment aside from my own; one of them appears to be intended for public consumption. That will be a 1V/Octave hybrid with no expensive uA726s. Mine is an attempt to come reasonably close to the original, using original parts (and 2V/Octave CV response, as in the original). If I can pull that off, I'll likely do 1V/Octave versions as well, eschewing the uA726 (not a bad idea, that.."THAT" being a pun, see further in this post for new info on replacing the 726 with the extremely accurate, inexpensive THAT 300 quad PNP IC.).

Remember- No hard or soft sync, no PWM, no animated waveforms..."just" a Buchla dual oscillator. For those not interested in 2V/Octave tracking, limited choice of imperfect waveforms, and no other niceties aside from the finest exponential FM I've ever found...you'll definitely want the 1V/Octave clone, or Mark Verbos' DIY 3080 version on simple-answer.com/DIY , or the Plan B Model 15, which has PWM, sync, and 1V/Octave control. I wouldn't be surprised if zero people take me up on this project offer- which is okay by me- I'm in it for myself, and if others might like it, I hope they enjoy the results as much as myself.

http://ebolatone.blogspot.com/2011/05/buchla-258b-remake-project.html

I'm reworking 258c artwork into 258b artwork (Thank You again kind sir!), as I prefer the b's exponential FM to the c's linear FM (although of course, either can be made to behave like the other; I once owned a pair of b's and they produce excellent non-chromatic effects unlike any other oscillator that I've used over the period of more than two decades of searching and sound design). Anyone interested in this could mod theirs for linear FM based upon the many 258c schematics which are available on the web. The c may be considered a slightly superior design; you might want to deal with that if you are a perfectionist, and if it is publicly available.

The 258 series fully-analog oscillators run on 15V bipolar power supplies. There are 741 opamps in the circuit, for CV summing as well as to buffer the waveform outputs. Remember that these are 0.5V/uS designs, and cause the resulting waveforms to be darker than their contemporaries. An easy mod would be to replace them with LF351s or other pin-equivalent single types, for much brighter saw/square waveforms as well as much brighter FM, or improved tracking/summing, if that makes you grin. I've never seen an LF involved in CV, so this might not be a good idea for that area. I'm certain that others will chime in with ideas and data. (Oh, the 351 data sheet says: "The LF351 is pin-compatable with the standard LM741 and uses the same offset voltage adjustment circuitry. This feature allows designers to immediately upgrade the overall performance of existing LM741 designs.")

My intent is NOT to produce anything commercial, not having received a reply from Don Buchla regarding permission to produce publicly-available etch-ready artwork based directly upon his PCBs...If someone else is doing so or planning to do so, I hope that it's both approved and also out of love for this gear and its musical characteristics and tactile pleasure.

I've managed to source all of the components for the design excepting the uA726HC heated expo pair. If this works I might replace that entirely, having a limited supply of my own...they tend to go from around $12.00 to over $25.00 each.

So far, the cost of the rare components, just the trannies, is coming to around $18.80 per (dual) oscillator. Quite a difference in price from the current vintage market value. You might be able to source these for even less than I have; I'll share my sources once functionality is confirmed, and if you have better sources, I hope that you'd share them as well.

Components for a 258 dual oscillator:

4) 2N4248 (original gold-legged, .20) =$.80
2) 2N4341 (original metal can, $1.80) =$3.60
4) 2N4339 (one should be selected; I don't know how to do so. Original metal can, $1.50, but deals can be had for quantity. Perhaps you know where to find PN versions for less...) =$6.00
14) 2N or PN3565 (.10 for PNs, .45 for original gold-legged) =$1.40
2) 2N3802 dual TO78 ($3.50) =$7.00

Additional:

4) 1N457 (1N457A, $0.05) = $0.20
2) 1N5233
8 LM, uA, etc. 741
2) uA726C (or HC) $12->35+ =$24->70+

Note that ebay appear to not have any deals on these sorts of parts. Good on you if you come across any. I've been watching for a lonnnng time and nothing...

For those who abhor the concept of spending that much on a single IC, although you're still getting a dual oscillator for something less than $200.00, perhaps much less depending upon panels, etc., here are some options:

Note that it would be easy to mod this design to use the SSM2210 and LM394, for example. I haven't researched the LM394 but that is a possibility, as is the THAT 300 quad NPN, which could be used across both oscillators, if you wish the maximum accuracy. Remember to ground the substrate.

I really recommend the THAT IC. It costs as much as one of the other types, and has enough trannies for both oscillators, which are on the same board anyway. You'll also receive excellent response from KA-Electronics who sell them in small quantity. I note that the 3046, another candidate for single-IC-solves both oscillator's needs, still has a price hovering around $1.00 or less, but it's a double-edged sword. Matching between devices is only 5Mv (Moog ladder filter matching is 1Mv). You can of course do the Moog Prodigy trick and use the extra trannies as heaters for greater stability, but the accuracy won't be amazing. I note that the THAT specs are amazing:

# Low Noise:
0.75 nV / ÖHz (PNP)
0.8 nV / ÖHz (NPN)
# High Speed:
fT = 350 MHz (NPN)
fT = 325 MHz (PNP)
# Excellent Matching: 500 µV typical between devices of same gender".

There should be artwork ready to press'n'peel early next week. I'll update if it produces a usable PCB. If not, the basic artwork would be usable for anyone interested in doing veroboards.

If Don didn't get my email and does NOT want anyone to share anything like this, based directly upon his PCBs, I'm happy to respect his wishes on the matter, first and foremost. Thank you sir for such excellent, inspirational musical instruments.

Update.

Build page:

http://ebolatone.blogspot.com/2013/10/258b-build-etch.html

258 Schemo w troubleshooting notes.pdf

Description:

Download

Filename:

258 Schemo w troubleshooting notes.pdf

Filesize:

293.66 KB

Downloaded:

1354 Time(s)

numbernone · Joined: Aug 16, 2006 Posts: 477 Location: new york city

This is great news.

GREAT NEWS

Peake · Posted: Fri Aug 15, 2008 5:49 pm Post subject:

Even if others come out with clones, this will have made me far stronger in the processes involved. There will be 258s, one way or the other.

We who build them should arrange some sort of donation to Don, as it's his work and design, and this would be the only honorable route.

Luka · Joined: Jun 29, 2007 Posts: 1003 Location: Melb.

Yup im keen
I keep telling myself i dont need to save up for an AFG or PlanB15 VCO.
If this eventuates i'll be very happy peake

hehe i might still succom to the AFG tho. :S

Peake · Posted: Sat Aug 16, 2008 9:27 am Post subject:

Upon re-reading my old email, it would appear that one of the clones will be public, either at the PCB or at the complete module level, or perhaps both. At this point, only absolute diehards need follow me any further. The other project will be a much better solution for those who don't need the closest thing to the original that is possible. It's based upon the c, and eschews the uA726, for good reason. It'll also likely have 1V/8Va response. I'm going after the original 2V/8Va for at least the first batch for myself. Once I can confirm these things, I'll look into 1V/8Va response..

You are aware that what I'm doing, and the original, are 2V per octave, right? At minimum I've shown that the majority of original parts are obtainable..

Edit: I'm including the 258b schematic, if anyone wants to point out how to change it to provide 1V/8Va response.

http://www.flickr.com/photos/11165691@N03/2767671021/sizes/l/

metasonix · Joined: Dec 28, 2003 Posts: 34 Location: CALIFORNIA

Hi, Mike. Jeez, you must be feeling really confident to be using uA726s in this VCO. They are DAMN difficult to find. If there's a way to use a regular matched pair and a thermistor (or an external heater, whatever), I'd advise it.

There are some good reasons why Don doesn't want to reproduce his old modules anymore.

(Did you know that futurlec.com still has LM394Ns for $1.40 apiece in quantity? Get some fast, the LM394 was discontinued by National recently.)

Peake · Posted: Sat Aug 16, 2008 3:18 pm Post subject:

uA726? I've already got some Laughing

And since when is "having" to use a $35.00 chip in a $200.00 or so clone of a $1500+ vintage oscillator a problem to anyone? Snerk...and we're both fans of the Western 300B- go figure about rarity and price Wink

I was -just- looking into setting the 726 footprint into a 10-pin DIP form, to see if other options could drop-in- They cannot. The SSM2210 is the closest, and the pinout is different. Turning the 726 one pin clockwise to allow the two trannies to be at the "top" is fine, but again, the pinout is different. Turning the 726 clockwise another two or three pins doesn't help to allow for a 2SA798 (edit: mistook the 798 for the 394, neither of which I have any interest in using) to drop in, either. Close, but no cigar. You'd have to do some ugly twisting to get the two compatable in the same SIP 5-pin footprint. Using a CA3046 between the two oscillators would work, using the Prodigy "heater" trick, but it's pretty hard to find as well, although far cheaper than the Fairchild, but it's also only matched to 5Mv- I don't see the matching spec on the Fairchild data sheet, but I'm fairly certain that it's superior (but not TOO superior).

Remember, I'm just a lowly sound designer and interface consultant with good ears and a passion for what sounds good, for coming up on three decades...I'm not an EE, don't claim to be, from observation wouldn't want to be, and am using the caveman tools available to me to pursue what I know to sound better than anything else, for the specific reasons they do. Pretty much no one else will, or even cares, so I have to build my own.

metasonix · Joined: Dec 28, 2003 Posts: 34 Location: CALIFORNIA

>both fans of the Western 300B- go figure about rarity and price

Did you know that the 300B is one of the lowest-distortion amplifying devices ever made? I know this from personal experience--having tested hundreds of different tubes for distortion.....anyone who wants to read about it is welcome to pick up back issues of Vacuum Tube Valley.
http://vacuumtube.com/vtv.htm
(The 300B tests are in issue 3, and listening tests in issue 8.)

Besides....there are several NEW 300Bs, currently IN PRODUCTION, at SIX different factories.... Wink

Dual matched transistors? Not so many. Apparently the MAT-02 and SSM2210 are still available. For how long, and WHY, who knows?

I'm thinking of this VCO being more than a DIY project. If you can get Don to sign off on it, you might make and sell finished versions. I seriously doubt he wants to do it.

If Cynthia can make a Buchla Zeroscillator, you can make vintage Buchla modules. Just don't listen to the whiners when it comes to pricing. Wink

dammit, this version of phpBB must have bugs or something. It won't accept my HTML or bb code...

metasonix · Joined: Dec 28, 2003 Posts: 34 Location: CALIFORNIA

>.I'm not an EE, don't claim to be, from observation wouldn't want to be, and am using the caveman tools available to me to pursue what I know to sound better than anything else, for the specific reasons they do.

Heh heh. I know one "high-end amp designer" who is a former high-school janitor. And I know another guy who makes high-end amps AND publishes a magazine about audiophile matters.....who is a former pastry chef. The sum total of their "engineering education" consists of hobbyist mucking-about, and reading Sound Practices magazine back in the day.

(And oddly enough, they both make and sell single-ended 300B amps. Behold the 300B! The Tube For Janitors!)

synthi · Joined: Jul 20, 2007 Posts: 29 Location: Spain

good news Mike!

Peake · Posted: Sat Aug 16, 2008 5:11 pm Post subject:

Peake · Posted: Sat Aug 16, 2008 5:14 pm Post subject:

synthi · Joined: Jul 20, 2007 Posts: 29 Location: Spain

Never heard a b or a c, so i can´t choose only based in the linear vs exp fm... it must be heard for choosing... also i think would be modificable for lin-exp fm input.

For the 1v vs 2v/8va is a matter of input scaling...

I`m with you for a "real" 258, if parts are unobtainable it could be modified for acepting substs later... Wink

Peake · Posted: Sat Aug 16, 2008 7:26 pm Post subject:

Don't buy anything just yet until either you decide to go with the other project or I can confirm the veracity of this one. Only the 726 is a problem.

synthi · Joined: Jul 20, 2007 Posts: 29 Location: Spain

Mike, do you have a link for the "other" 258 project? I`d like to see what is changed.

Synthi

Peake · Posted: Sun Aug 17, 2008 9:43 pm Post subject:

It has not yet been announced. If it's going to be for public sale at any level, I hope that this has helped raise awareness and interest, and hopefully hasn't been in bad taste or to be problematic in the least, which is not my intent.

I'm not snobby about my gear choices. It's not my fault that excellence tends to command amazing vintage pricing...my attempt to provide artwork for anyone to use is a response to the vintage pricing versus actual parts cost and desirability. Buchla gear has indeed received a greater awareness and respect over the last half-decade or so, and that's a good thing, so long as those offering various versions do so out of reverence and love for the gear. Anything else is an insult to the gear and to the community.

I don't think that anyone will want my version over the potential clone. Does anyone have uA726s lying around beside myself? I've updated the original post with this, for your consideration:

Remember- No hard or soft sync, no PWM, no animated waveforms..."just" a Buchla dual oscillator. For those not interested in 2V/Octave tracking, limited choice of imperfect waveforms, and no other niceties aside from the finest exponential FM I've ever found...you'll definitely want the 1V/Octave clone, or Mark Verbos' DIY 3080 version on simple-answer.com/DIY , or the Plan B Model 15, which has PWM, sync, and 1V/Octave control. I wouldn't be surprised if zero people take me up on this project offer- which is okay by me- I'm in it for myself, and if others might like it, I hope they enjoy the results as much as myself.

I'm looking to produce an "actual" Buchla bit, in the interest of using it in actual Buchla context. This might not be so attractive to non-Buchla owners.

ericcoleridge · Joined: Jan 16, 2007 Posts: 889 Location: NYC

3vcos · Joined: Oct 26, 2006 Posts: 106 Location: California

There are currently quite a few good and (relatively) cheap sources for the UA726.

SoundDaveUK and TechnologyTransplant are both well know and reputable sellers on Ebay. Their '726s have been tested good by numerous buyers/groups. Search for them on Ebay.

Here is a current link...

Beware HonkKorgSuperSelleer / PartsPipe on Ebay. Don't buy UA726s from them!

[Blue Hell: shortened link]

norman phay · Posted: Mon Aug 18, 2008 8:37 am Post subject:

Good luck with this project, Mike. I'll certainly be interested in seeing what you come up with, & will be following you progress as you report it.

Peake · Posted: Mon Aug 18, 2008 9:28 am Post subject:

Waveshaping on a 258 consists of a crossfade (VC as well) between the sine and the other waveform- one of the two being square, and the other sawtooth. It's not a morph, and there is a lot of extra sine wave fundamental in the saw/square- looks terrible on an oscilliscope, but sounds alright. Remember per the OP that due to the use of 741s on the output buffers, that the waveforms are darker than contemporary oscillators, most especially something like the Plan B. Drop in an LF351 or LF356 for brightness.

Tim- Would you please be so kind as to tinyurl.com your link? It's thrown the margins. Haha.

3vcos · Joined: Oct 26, 2006 Posts: 106 Location: California

just search for 'ua726' on ebay. Be sure to search worldwide.

ericcoleridge · Joined: Jan 16, 2007 Posts: 889 Location: NYC

Peake · Posted: Mon Aug 18, 2008 1:20 pm Post subject:

funkyfarm · Posted: Mon Aug 18, 2008 1:37 pm Post subject:

Peake · Posted: Mon Aug 18, 2008 1:56 pm Post subject:

Yes, both the pot and CV cause the waveform to crossfade between either sine and saw, or sine and square, depending upon how you stuff the board (on the original, one is saw, one is square).

Mark's DIY hybrid looks to be quite nice. I've gone over it and the waveshaper is 99% the Buchla waveshaper. Mark is someone I trust with the Buchla gear, at the functional, form, and musical level. Although I haven't built one of his DIY osc's, I've talked to him enough and seen enough to have faith.

The artwork I'll release has the advantange of having the pair of oscillators on a single PCB, but uses the uA726, is 2.0V/Octave, has no PCB-mount pots, etc. The original requires a bit of panel-mount wiring, including resistors etc., using a pair of terminal strips hot glued to the back of the panel. Don later started using control boards with panel-mount jacks, pots, and switches for good reason, looking at the difference between the 292b and 292c...all you have to do is hook up a couple of ribbon cables between the main and control board. Same with the 281 and 259, and doubtless others.