[sdiy] Analysis of Buchla 259 principal oscillator...

[Aaron Lanterman]

...or, a (failed) attempt therein. I seek the wisdom of the gurus.

IC22 is a 311 comparitor; it appears as though a couple of resistors form
a "pullup resistor" effect to 5V. (15V through 10K to output of 
comparator, and 4.99K from output of comparator to ground).

Yet, the notation shows a square wave with a "4.29 V" by it.

Question #1: I'm guessing there's stuff happening in the caparitor that 
cause it to not pull up all the way to 5V? How do I find out what the 
difference is? It is not something I can seem to find on the 
datasheet.

Cap 20A, a 22p, runs from the + input of the comparitor to ground; its 
purpose seems to just be to suck off Random Unwanted Goo. Assuming an 
ideal comparitor (dubious, yes), R164A, a 10K resistor, has little effect, 
and either 0 or 5 (or 0 or 4.29, but I've yet to explain the 4.29) V goes 
to the + input of the comparitor.

The triangle output is taken from the output of IC20, which is a trusty 
TL082 that forms the integrator along with a 0.0047 (no explicit units 
given) polysomething cap.

Let's call the triangle output vtri, and the input to the - terminal of 
the comparitor v-.

There's a resistor setup...

                   15 V
                     |
                   20.7K
                     |
                     |
vtri ---- 4.99K ----+-- v-

Let's find out what values of vo will trip the comparitor.

By the usual resistive divider setup:

v- = (vtri 20.7 + 15 * 4.99) / (20.7 + 4.99)

v- = vtri 0.8058 + 2.9136

vtri = (v- - 2.9136) / 0.8058

Let's suppose the comparitor trips at 0 and 5 V. This corresponds to 
tripping at

vtri = -2.9136 / 0.8058 = -3.6158 and
vtri = (5 - 2.9136) / 0.8058 = 2.5692

If we use 4.29 instead, that's

vtri = (5 - 4.29) / 0.8058 = 1.7081...

but either way, that's hardly a triangle centered around zero! There's 
just a line with a triangle on it, suggesting that's the output - I don't 
see any "zeroing" circuitry elsewhere...

Question #2A: Is there some DC shifting circuitry for the triangle in the 
259 I'm missing? It's a typically Buchlaesque diagram, drawn in Buchlaese, 
so that is one possibility.

Question #2B: Is there something daft in my analysis? What is it that I'm 
missing?

Oh, wait, there MUST be something daft in my analysis. Notice the 
resistive divider of two 100K resistors, R140 and R141, which is feeding 
7.5 volts to + terminal on IC20, the integrating op amp - so vo, it looks 
like, is supposed to be centerd around 7.5???

Help! I've fallen in a Buchla design and can't get out!

- Aaron

P.S. I thought the 20.7K might be 26.7 or something, as it's very smudgy - 
but if you use that, you get, vtri = -2.8 for v- = 0, vtri = 3.1313 for v- 
= 5, and vtri = 2.2285 for v- = 4.29, again not really centered)

-----------------------------------------------------------------------------

Dr. Aaron Lanterman, Asst. Prof.       Voice:  404-385-2548
School of Electrical and Comp. Eng.    Fax:    404-894-8363
Georgia Institute of Technology        E-mail: lanterma at ece.gatech.edu
Mail Code 0250                         Web:    users.ece.gatech.edu/~lanterma
Atlanta, GA 30332                      Office: GCATT 334B