Progressive build of MFOS VCO - can I test in stages?

[kjackman]

Hi folks,

I'm breadboarding the MFOS VCO. Rather than building the whole thing at once and then troubleshooting the inevitable mistakes, I decided it might be better to build this thing in stages, testing after each stage to make sure all is well before moving on.

Stage 1 is the summer circuitry, this was easy. Stage 2 is the expo converter. This is where I'm having difficulty. In the diagram below, point "B" is not connected to anything (stages 3 and 4 not having been built yet), and no matter what voltage enters stage 2 from the summer, I get 0V on point "B". I realize it's probably current I need to measure here, not voltage, but I figured there'd be some voltage there, no?

1) Have I divided the stages in a reasonable way?

2) if I want to test at point "B" before stage 3 is built, do I need to attach some kind of load, or temporarily tie "B" to ground or something with a resistor while I test?

3) Is this "progressive build" approach even feasible?

I'm still pretty new at this stuff, although I've built simple chipamps and baby 10 sequencers and so forth.


MFOSVCO by kjackman, on Flickr

Thanks!

[jordroid]

Oooh, i like the way you are thinking, i try to build things in small stages as well, it lessens the chance of spending hours on something only to find it doesn't work and you have no idea why.

It looks to me like you have divided the circuit very logically, you have your inverting summer, expo converter, integrator, and then comparator.

In a circuit like this, i would personally not break it down quite as much as you are proposing. Like you said you can easily verify the functioning of the inverting summer on its own, but i would just go ahead and breadboard the rest all together. It's not that large of a circuit, and it's easier to verify the expo converter, integrator and comparator working together than it is on their own.

I think you are correct in thinking that the expo converter output is a current, maybe someone will chime in with a clever way of verifying this circuit chunk on it's own. Since it's on a breadboard you don't have to desolder anything to get a current reading, which is nice.

You can leave off the high frequency compensation circuit of D1, R9 and R11, and the sync circuit of R68, C5 and R34 at first to keep things simple if you want.

You could probably verify the comparator section with a voltage divider feeding a voltage into the non-inverting input i suppose, the output should go high and low as you cross the threshold by turning the pot. You could use the coarse tuner for this i bet, maybe with a little tweaking. (i'd just build sections 2, 3 and 4 together though myself)

cheers

jordan

[kjackman]

Thanks for the advice! I went ahead and built the whole thing, minus the sync input, the HF compensation, and the linear input.


MFOS VCO attempt no. 1 by kjackman, on Flickr

It outputs a ramp wave... at 8000-ish Hz. A noisy-looking ramp wave. Whose frequency doesn't budge no matter how I twiddle the coarse and fine tune knobs. And it draws a ton of current apparently - the LEDs by my voltage regulators go dim when it's on. But hey, ramp wave, woo-hoo! Does this mean I have it half right?

I guess I'll have to re-check all the connections next.

[jordroid]

It's hard to see but it looks like the collectors of your transistors may be tied together, instead of the emitters? I've done that before and it caused the circuit to suck a lot of current and heat up the chips as well, in a different type of VCO.

Nice breadboarding

[kjackman]

[kjackman]

I got sidetracked by some mistakes. Troubleshooting details in this thread. I've successfully built the first three "stages" of this VCO, all the way up to the "raw" ramp wave output. So it looks like this "progressive build" idea is do-able, but I'll need to figure out each stage the hard way. I'm really appreciating Ray's notes, though.

I'll post the step-by-step procedures when I've had a chance to compile them properly.