Author |
Message |
rumpofsteelskin
Joined: Apr 22, 2009 Posts: 52 Location: brighton, uk
|
Posted: Mon Jun 01, 2009 9:36 am Post subject:
VCO-1 and general log converter questions Subject description: PNP, NPN, SMOKE |
|
|
I'm currently building the VCO-1 on veroboard, actually a modified version as it's only required to produce a constant tone under control from a pot. Because of this I have taken out the summing for the control voltages, and have a dc signal fed straight into the log converter.
Now, im no expert...
but, if you're inverting your input signal and then using PNPs in the converter, isn't this the same as not inverting the input and using NPNs? (I did try this and got some fried transistors).
My ears may be lying to me, but when I use the PNPs as the original design states, the exponential seems to start at the high frequencies, where it changes slowly, and then increase alot faster at the low frequencies - the log curve seems to be inverted. Is this an effect of omitting the inverting stage? |
|
Back to top
|
|
|
Tim Servo
Joined: Jul 16, 2006 Posts: 924 Location: Silicon Valley
Audio files: 11
|
Posted: Mon Jun 01, 2009 11:46 am Post subject:
VCO-1 and general log converter questions |
|
|
Hey Rump,
I don't have the VCO-1 schematics in front of me, but from other similar TH designs, I'd say that the functions of the summing stage (and associated op amp) are to invert AND to reduce the range of the CV.
Okay, NOW I'm looking at the VCO-1 schemos. Yup, note the 2K resistor in the feedback loop of the summing op-amp? That reduces that gain of that stage a LOT. Between that and the 820 Ohm resistor to ground, the range of current going to the expo converter is very small. I'd say that's a big part of what you're missing by not using the summing stage. I've also been told that the difference between using an NPN or a PNP to drive the VCO core is whether the core needs a current source or a current sink. That may explain why you fried some transistors when you changed them.
If you really don't need a VCO and all you want is a variable pitch tone source, maybe you can try an LFO design and just use a smaller cap to put the frequency up in the audio range. I have a nice LFO design with a wide range (over 8,000:1) that outputs a triangle and square waves, but can also use typical waveshapers to get saw, pulse and sine waves. I'll post it here if you're interested, or just search E-M for the 'can anyone recommend an LFO' thread.
Tim (mmm... fried transistors) Servo |
|
Back to top
|
|
|
rumpofsteelskin
Joined: Apr 22, 2009 Posts: 52 Location: brighton, uk
|
Posted: Mon Jun 01, 2009 12:19 pm Post subject:
|
|
|
Hi Tim, thanks for the reply
I forgot to say I scaled the voltage down a long way to compensate for the effect of the summing stage. Concerning the transistors, I changed them round so the current was running in the same direction, so where on the design the current is running from emitter to collector, i changed it to a NPN with the current running from collector to emitter so they are still sourcing current.
So the same thing, but turned on by the opposite polarity - I thought - my transistors had other ideas.
What Im building is actually to go inside something else which will take more than one control input to the vco, but i was investigating the possibility of just summing them with the input transistor as in Rene Schmitz MS-20 filter clone. Here the log converter is very similar but there is no inverting stage |
|
Back to top
|
|
|
|