VCO that cannot be scaled! WEIRD

[numbernone]

Recently put together a couple VCO boards I had laying around (dont know the revision off hand but I think they are perhaps the oldest represented on the website). They function, sound nice and look pretty good on scope. However they CANNOT be scaled for V/O. The trimmer would appear to only offset the "center voltage" or pitch, rather than determine the width of an octave. No matter the setting 1V roughly spans about an octave and a half with no difference at either extreme setting.

This is super strange. I had to button the unit up for a gig that did not require such accuracy, but I intend to get back in there and want some clue as to this behaviour.

[jordroid]

I'm not sure, i was never *totally thrilled* with the accuracy of the MFOS VCOs that i built, but they did track nicely over three octaves or so. Maybe check and see what is different about the older revision you worked from as compared to the newest version and see if you can kludge any improvements in? (If it was me i would breadboard it first before chopping anything up)

It's not something i would really recommend that anyone else do, but i kludged the Ian Fritz bare saw VCO from this thread http://www.electro-music.com/forum/topic-28713.html onto a couple of MFOS VCO boards, after being unhappy with the tuning difficulties i was having. I bring that VCO up all the time because it's not much more expensive or harder to build than most VCOs but it tracks waaaaaaaaaay better than others i have tried, if wide range accurate tracking is what you want. It required lots of trace cutting and a little daughter board though.

There is probably some error though if you are getting zero nice octaves out of them, my procedure is almost always the "breadboard it until it works and then compare it to the broken one" approach, it's not a huge circuit if you leave off the waveshaping and just do the VCO core.

A few easy possibles-

-for the CV summing stage the input resistors must be 100K and the feedback resistor must be 2K, preferably a tempco, but it should still scale fine with a plain old resistor here

-the 1V/oct trimmer must be 100 ohms and the resistor from the bottom leg of the trimmer to ground must be small, like 330 ohms to 475 ohms or so depending on the design

if one of those was accidentally off by a factor it would probably do what you are describing

Let us know how you get along,

jordan

[Skrog Productions]

Hello there .

I have built 6 mfos vco's (2 old ones & 4 new ones) , i modded the old versions up to the 2009/2010 standard by comparing the circuit layout differences between the 2 documents , my old vco's used to track up to 4 octaves reasonably well , then after upgrading to the latest versions i am able to get a good 6 octaves , ideal for my melodic music .

It took me a while to learn how to stretch the octaves using the offset trimmer & high freq compensator trimmer when calibrating and i found that Ray's vco tuner helper pcb circuit really helped in the process of calibrating , good luck with solving this


Dave.

[numbernone]

I have built another 15 or so oscillators including MFOS, and am familiar with the fun of getting the V/O in order. I am way better now than I used to be, but these two are just not right.

I am trying with a frequency counter at the moment, and this makes it clear how twisted they are. Regardless of the V/O trim setting the frequency increase per input volt is approx 40%. If 1v=100hz 2v=140hz and so on 200/280 400/560.

So very strange. I am finding the right values around the trimmer. Will keep poking.

[jordroid]

Hmmm, since it's so consistent, i would still guess that it's the CV summing/trimming area, or maybe the high frequency compensation circuit. It really seems like the scaling is circuitry is to blame, there is not a whole lot else it could be.

You could try a smaller resistor between the 1v/oct trimmer and ground, just to see what happens.

It could be further down the line as well i guess, like the cap discharging weirdly or the FET not behaving, but since you say it looks good on a scope i would bet the integrator and comparator are both working ok, which points back to the front end.

I dunno, just thinking aloud with you...

[numbernone]

I feel the same about the cv summing zone. I actually stuffed these boards 5 years ago when I just started and made lots of dumb mistakes, only getting them pulsing recently. Since they both do pretty much exactly the same thing, I must have stuck the wrong thing in the wrong place 2x.

This is where I think out loud too. Often it seems I only need post and something pops out that hadnt for the preceding 6 hours. Not yet this time tho!

[haxster]

If you have tried everything and still no dice, perhaps one of the IC's is faulty. Try swapping for an entire new set of IC's. I recently had your problem and with mine it turned out to be a faulty IC. For the life of me I could not get it to track correct either until I did that, after that all was good.

[numbernone]

Any idea which IC it was? Please dont say the transistor pair.

I bet you're gonna say the transistor pair...

[haxster]

No it probably is not the transistor pair. It was not that with mine. On mine it was a custom build, so I could not tell you exactly which IC it was but I did notice when I would turn my Pulse Width it would also change the Pitch not just a little but way to much. That gave me clue. So I changed a bad TL074 which on my design has to do with the Square wave. That solved my problem.

Also another time I was using nothing but Hi-end IC's and for the life of me they would not track right. I switched over to TL072 and TL074 and (2) 2N2222A transistors and it was tracking very good.