MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

Buchla 208r V2.1 (Blue) Build Thread
MUFF WIGGLER Forum Index -> Music Tech DIY Goto page 1, 2, 3, 4, 5, 6, 7, 8, 9, 10  Next [all]
Author Buchla 208r V2.1 (Blue) Build Thread
Don T
Roman's newest version of the 208r contains quite a few changes, and many mistakes from the older versions have been corrected, so I think it may be worth starting a new build thread for this version. Due to the many corrections, there will/should be less modifications and other tweaks made, and this should hopefully turn out to be a much shorter thread.

To get the ball rolling, here goes:

First, if your boards look like this, you indeed have a 208r V2.1:



Here's the link to the updated BOM and parts layout diagrams:

Stored-Program-Sound-Source-Model-208r V2.1

The first thing to note: I at first thought that the missing trace from Pin 5 of Card 1 to Pin 7 of the programming card slot had been added. This is not the case, the trace is still missing. The correction that sduck posted in the previous thread will still need to be made:



The first notable correction is on Card 3, the Envelope Generator. Previously, two of the pins of the 2N2907A in the 13.5V supply had to be swapped, and a flying 100K load resistor had to be added to get the correct output voltage. This has been fixed, so insert the 2N2907A without crossing any legs:



And the formerly flying 100K resistor has a spot on the board, and it goes here (Be glad I didn't go with making this an animated .gif where the word "Here" blinked like I wanted to at first twisted ):



A couple other suggestions for further along in the build that may save some people some trouble (I'm a visual thinker, so have a few hopefully helpful pics):

Card 6 - Match the two 2N3565 transistors, Q2 and Q3, for Vbe. These two are the core of the Modulation Oscillator, which is basically 1/2 a Buchla 258 without the sine shaper circuit:



Card 8 - Match the components enclosed in the like-colored rectangles:



Match D1 and D2 (1N457) with the diode checker on your DMM, which should display a Voltage reading. R36 and R37 are 330K, and R38 and R39 are 680 Ohms. Even if using 1% resistors, I would still match the closest pairs I could. These components, along with Q1, are the sine shaper circuit from the Buchla 258. The better they match, the easier time you'll have getting a clean sine wave. Also, Q1 is a 2N4339 that should have an Idss between .7 and 1.2mA.

[EDIT]: In case people need these for chasing bugs, attached are the schematics for the version of the 208 that Roman's boards are based upon. The schematics on Magnus' most awesome site are for a later version of the 208.

[EDIT - February 1, 2017]: Even more help for bug-chasing: I cannot remember where online I found this, but I'm attaching a pdf file that shows the entire 208 Mother Board schematic as one continuous image. I've found it very helpful. If anyone knows who should get the credit for this image, let me know!

[EDIT - January 30, 2017]:
In case something goes wrong, here are a few general troubleshooting tips, plus a short example (presented at the risk of making even more a fool of myself)

If something doesn't work, or work as expected (Go through this list In Order!) -
1. Check the power supply Voltages!!!!
2. Go back and make sure you put all the parts in (more on this in a bit!) Good suggestion from Peake: Hold the (hopefully completed) board up to a bright light, and look for any unexpected holes.
2A. If you put your chips in sockets, make sure none of the pins got folded under and are not making contact
3. Double-check the board against the BOM and the parts placement pics to make sure you put the correct parts in the correct places, and with the correct polarity. This means all diodes, vactrols, chips, electrolytic and tantalum caps facing the correct direction. This is an easy mistake to make, so the more sure you are that you didn't make this mistake, the more you probably should make yourself double-check Mr. Green
4. Check all the solder joints on the card. It is easy to miss soldering a lead or two sometimes. An incomplete, broken, or bad solder joint can cause all sorts of problems. If in doubt, re-solder it, but don't be sloppy.
4A. Check for solder bridges between component pins that are not meant to be connected, or between adjacent traces that are not supposed to touch. This can happen easily if too much solder is applied to a connector, or a chip socket/chip. The photos of the boards on davebr's site are a wonderful reference if you're not sure a connection is bridged (Note that this is V2, not V2.1):

http://modularsynthesis.com/roman/buchla208v2/208spss.htm

5. Try swapping chips, assuming you put them in sockets.
6. If you didn't put the chips in sockets, or you suspect a permanently soldered-in component, get the DMM out and start checking for reference voltages against the voltages that are indicated on the schematics.
7. Ask for help!! Because if still no joy at this point, the problem could be complicated!

Here's a short story for you related to the above:
Finished up Card 2 (Random Voltage Generator) the other night. Went to test. Nothing, nada. I had recently checked the power supply, and everything else was working normally, so...
I skipped to Step Number 5 on the list above. Swapped one chip, tested, swapped another chip, tested, swapped another... until I had swapped them all out. Still nothing. So, after wasting all that time, I got marginally wiser (but not much), and skipped back up to Step Number 4. All the solder joints looked great, until I got to a place where there were two holes I wasn't expecting to see. I flipped the board over and:



Yes, that 1M resistor was still missing after everything else was installed and I thought the card was complete. So, no matter your level of experience (and believe me, I'm still no expert), go through the general troubleshooting list In. The. Order. Above. There's a reason Step Number 2 is Step Number 2.

After adding that one missing resistor, and kicking myself a few times, the Random Voltage Generator worked as expected!
Don T
Just moving this bit of information from the other thread for the benefit of those who are looking into buying vactrols for this build:

Don T wrote:
muncky wrote:
- now fretting about excelitas b xvive, but other than very excited...!


I'll offer my observations and opinion on this, and it's only my opinion, and based on a small sample, but here goes:

I built a Card 10 for the 208 (Gate 1) that had the vactrols socketed. I tested Excelitas VTL5C3, Xvive 5C3, and Xvive 5C1. With the attack, duration, and decay sliders all the way up, here's how they compared: The Excelitas 5C3 was as we would all expect, while the Xvive 5C3 had a barely perceptibly slower attack than the Excelitas, but the decay was much longer, and it took what I thought was an unacceptable length of time to finally settle at the bottom of its curve. If it wasn't for the long tail at the bottom of the curve, the difference would not have been so pronounced. As it is, fast, short notes repeated rapidly would start bleeding together long before the triggers reached audio rate with the Xvive 5C3.

The Xvive 5C1 was an entirely different beast. It is fast. Very fast, on both attack and decay. Compared to the Excelitas 5C3, it was basically almost too fast. Adding just a little duration and decay slider got the same response you would expect from the Excelitas 5C3.

The good news is that between the three, I noticed no real difference in tone on sustained sounds. And for those that don't know me, I make my living dealing with issues of tone quality.

The bottom line: If it was me building a 208 (And I'm just starting another one right now), I would go about vactrol preferences in this manner, in order of preference starting from the top:

1. Complete the build with Excelitas VTL5C3 in all positions.
2. Complete the build with Excelitas VTL5C3 in the balanced modulator and both Lopass Gates, with Xvive 5C1 in all other positions.
3. Complete the build with Excelitas VTL5C3 in both Lopass Gates, Xvive 5C1 in all other positions (This actually may improve the balanced modulator, but perhaps not as "authentic" as option 2).
4. Complete the build with Xvive 5C1 in all positions.

Like I said, this is just my opinion and observation. I know a couple other wigglers have been playing around with this as well, and they may choose to add to this, or not.
muncky
Thanks so much for starting this thread - can wait to start building this, and this kind of info is super helpful thumbs up
Reality Checkpoint
I am seriously considering building this, so this is great. Need to source some UA726. Are they ridiculously expensive, and can they be bought from a reliable source?
Don T
Reality Checkpoint wrote:
I am seriously considering building this, so this is great. Need to source some UA726. Are they ridiculously expensive, and can they be bought from a reliable source?


IF (Big if, get it?) you can find a genuine uA726, they will probably set you back around $100 each. There are many fakes out there, especially on eBay from Asian sources, so beware.

Thankfully, a couple uA726 alternatives are out there. papz has a small uA726 replacement board which I believe is based on the expo converter from the Moog Prodigy, and jhulk has a uA726 replacement board in the prototype stage that is based on a different circuit that also uses a heater. I will be testing jhulk's board in this build (As soon as I get them etched). I also have developed a replacement that uses a matched transistor pair and tempco that is in the prototype stage. My boards replace Card 6 and Card 7 entirely. I have a pair built, and another wiggler has built a set and is testing them in his 208. Eventually, I'll be able to give pitch drift tests between a 208 with real uA726, my tempco boards, and jhulk's replacement.
Peake
I've got a working 258r V1 with sockets for the 726 I've been waiting to test jhulk's replacement...would be happy to do so. Thanks for the post and information Don T!
Reality Checkpoint
Many thanks for that Don T, most helpful

I reckon that this build will expend my complete SDIY budget for the year so I must weigh that carefully against other projects that might come up. I shall follow this thread very carefully.
jhulk
mike if you want the foil files for self etch i can send you them thats what don t has

was waiting for final tests before doing a run or you can wait for the production run.
Peake
jhulk wrote:
mike if you want the foil files for self etch i can send you them thats what don t has

was waiting for final tests before doing a run or you can wait for the production run.


Drop them on my email please, I have a double-sided etch coming up and I'll add it in. With something that tiny I might not pull it off. I've done okay on a couple of 258R V2 builds using the 857 and 847 but they're at the limits of my soldering ability without something like a microscope! Thanks!
jhulk
they are soic and through hole parts 2 soic chips which are the largest of the smd types

don t is sending them you
jhulk
here is the size the picture of it



gives you an idea of the size of it its nearly 1" square
Peake
Oh, this is very different from what I'd seen before. Thanks to you and to Don!
davebr
My notes also say there was a missing run between Card3 pin 15 and Card4 pin 10. Also Trimmer TR3 on the motherboard should connect to pin 12 of Card8, not pin 11. This was on an early version and I have not kept track of revisions. Can you verify these on the blue?

Dave
Don T
davebr wrote:
My notes also say there was a missing run between Card3 pin 15 and Card4 pin 10.


I haven't started on the motherboard yet, so this was easy to check. I get continuity between those two pins, so that issue has been corrected.

davebr wrote:

Also Trimmer TR3 on the motherboard should connect to pin 12 of Card8, not pin 11.


Alas, Pin 12 is not connected to anything. Pin 11 is still connected to the wiper of TR3.

Dave (or anyone else), if you think of anything else to check while my motherboard is still blank, please let me know. I also emailed you a decoupling cap question related to this project in case you didn't see it.
Don T
This is not exactly a requirement, but is strongly suggested for long-term reliability, and can apply to other similar circuit boards.

For double-sided circuit boards that do not have a solder mask, i.e., the traces are not insulated, when installing resistors it is suggested to mount the resistor slightly above the board if one of the end caps would come into contact with an uninsulated trace. For example:



In the photo above, all three of the visible resistors would have their end cap area contacting a trace if they were pulled flush against the board. Yes, the resistor has a very thin insulating coating in that area, but over time that coating could flake off due to vibration, corrosion, or expansion/contraction due to heating/cooling cycles.

As I said, this is strictly not necessary, but anyone who has ever repaired a beast known as the "Polymoog" would probably back me up on this...

Carry on! cool

[EDIT]: Just for clarification, all those words above boil down to: Keep the body of the resistor above the board if the body is going to be contacting a trace. I get too wordy sometimes.
davebr
Don T wrote:
This is not exactly a requirement, but is strongly suggested for long-term reliability, and can apply to other similar circuit boards.

For double-sided circuit boards that do not have a solder mask, i.e., the traces are not insulated, when installing resistors it is suggested to mount the resistor slightly above the board if one of the end caps would come into contact with an uninsulated trace.

Another option is to use a "christmas tree" to bend the leads of all components. That way the bends are 90 degree and have clearance to neighboring PCB traces. The 208 is one of the few boards that actually uses a 0.400" resistor lead spacing so I can use my vintage "christmas tree". It's 40+ years old so has fewer standard spacings than this one at Sparkfun

Mine looks like just the bottom half of this one Thinkiverse

Good tip, though about shorting to runs. That is why even with solder mask I make sure all my component leads are perpendicular to the PCB. I never solder leads bent over.

Dave
Don T
Troubleshooting timesavers-

Tip 1 -
Should anyone encounter the following symptoms all at once:

1. Sequencer LEDs do not light.
2. LPG LEDs do not light.
3. CO Sine Wave is inaudible.

Your 208 is not getting +5V from your power supply. (Do NOT ask how I know this very frustrating )

A 208r V2 or V2.1 will work 95% correctly with NO +5V power present (Again, do NOT ask how I know this angry )

Tip 2 -
If your CO Sine Wave is audible, and both LPG LEDs work, but the Sequencer LEDs do not light, try replacing IC9 on Card 1, a CD4009
Don T
Some progress, and perhaps a solution to a common problem should anyone else want to confirm.

I have worked on many different pieces of gear, and have noticed almost all of them that use CMOS chips have a decoupling cap across the Vdd and Vss (or ground) pins of most, if not all the CMOS chips. The original 208, as well as the 208r V2 and V2.1, have NONE. Most of the stability issues in the 208r are CMOS-chip-related, and I happened to run across a deal on these really nice machined-pin sockets that have decoupling caps built-in. I figured it may be worth a try, and at the very least it should do no harm...

Since I'm not an expert (at all) on CMOS-related things, I ran the idea by davebr. He gave me a thorough and educational response (Dave, feel free to jump in, or I can share your response here if you'd like) that basically said it may be a good idea in the sequencer, but perhaps putting them on every CMOS chip would be overkill. He mentioned that he tends to put one decoupling cap for every three CMOS chips. Dave also cautioned to make sure that any additional caps would not add enough capacitance to allow a charge to be stored in a voltage regulated circuit large enough that it would reverse-bias a voltage regulator, unless the regulator has protection diodes installed. Think of it as a "back flow" if you will, especially when powering down.

So, with basically a "go for it", I installed the sockets on my sequencer board. The decoupling cap can be removed without too much trouble, so I pulled it out of the socket for the LM3900:



In my previous post, I mentioned the trouble I had been having with my first 208's sequencer LEDs, and the solution, which involved replacing what I assumed was a faulty CD4009, which is the LED driver for this circuit. As an experiment, I placed the new 4009 in one of those decoupled sockets to see if there would be a result, plus you can see above that I didn't use sockets at all on my first build, which tends to slow things down a bit if you need to replace a chip...

Here's that part you've been waiting patiently for, hoping I would shut up and just get on with it:

Before installing the decoupling socket, my first sequencer card displayed the common behavior of only doing 2 steps with the Stages switch set to 3. It would do 4 and 5 steps normally, but not 3. Well, after installing that socket with the decoupling cap, and no other modifications done, that same sequencer card now reliably switches between 3, 4, and 5 stages as designed! nanners

The new, blue sequencer card, which you can see above has decoupling caps everywhere except the LM3900, and no other modifications, also switches reliably between 3, 4, and 5 steps as designed.

Being a firm believer in "Correlation Does Not Imply Causality" (unless confirmed), I am not going to declare the decoupling solution as the fix for the 2, 4, 5 step sequencer bug until someone else confirms. All one needs to do is solder a 0.1uf cap (ceramic will do nicely) on the back of IC9 on their sequencer (Card 1) between pins 8 and 16. Please share your results here!

Right now, all I'm going to tell you guys is that I have two otherwise-unmodified sequencers that behave as intended!



And yes, the sockets just happened to be blue! nanners
Peake
Don T wrote:
Alas, Pin 12 is not connected to anything. Pin 11 is still connected to the wiper of TR3.


Anyone have a visual on what to cut and what to jump to solve for this please?
davebr
Peake wrote:
Don T wrote:
Alas, Pin 12 is not connected to anything. Pin 11 is still connected to the wiper of TR3.


Anyone have a visual on what to cut and what to jump to solve for this please?

This is on Card 8. Pin 11 is not used, so the easiest solution is simply to jumper pin 11 to pin 12. I do show that on card 8 on my 208 V2 Stored Program Sound Source Module page.

Dave
Don T
[EDIT]: Text deleted because davebr beat me to the answer cool
Peake
It can be done on card 8 and not the motherboard? Good news, thank you both! This -is- my first 208 build so thanks for the patience. I have your page open in my browser, Dave. Superb work.
Don T
Just to clarify, the recently found error on Card 2 that applies to some older versions of the 208r V2 does not apply to this version, which is correct.

Carry on...

davebr
Don T wrote:
Just to clarify, the recently found error on Card 2 that applies to some older versions of the 208r V2 does not apply to this version, which is correct.

Carry on...

Ah ... what recently found issue on card 2? I am currently building an older version so am interested ...

Dave
thereminsynth1234
Thanks for the post on this build I just started getting parts for the 2.1. I'll try to post some pics on progress. thumbs up Guinness ftw!
MUFF WIGGLER Forum Index -> Music Tech DIY Goto page 1, 2, 3, 4, 5, 6, 7, 8, 9, 10  Next [all]
Page 1 of 10
Powered by phpBB © phpBB Group