FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 

First VCO - Help needed
MUFF WIGGLER Forum Index -> Music Tech DIY  
Author First VCO - Help needed
Hi all,

So I'm doing some elementary circuit building and settled on making a very classic triangle-core VCO. I made a Falstad circuit online so you guys can see exactly what I made.

Link to the circuit simulation.

What's not shown in the simulation is that I buffered both the outputs of the two OPAMPS with another set of TL072's used as a buffer. After that I put trimmers to adjust the output voltage and thus the volume (so I could listen to it). All hooked up to two thonkiconn connectors in which I can plug my oscilloscope or headset.

First of all, when plugging in my headphones, no signal at all. I'm thinking the headset is just too low-impedance that it pulls it to 0V right away. Didn't double check that with a multi-meter. That's strange since the signals have been buffered. Got me puzzled.

When checking the signal from the supposed TRIANGLE I come across a strange square-like pattern with an odd duty-cycle. Here's a picture of that.

The SQUARE wave is based off of that and of course shares the same duty-cycle, but inverted. Here's the picture of that shape. I think that fixing the triangle will fix the square wave to have a even duty cycle.

I started from this and made my own adjustments.

I also noticed that changing the input voltage (which would normally come from the 1V/OCT linear/exponential converter) doesn't change the frequency as dramatically as I expected it would at first. I arranged a trimmer potentiometer attached to +12V and GND, and the wiper would then be the input voltage. Trimming this barely changes the frequency. I'm really puzzled as to why this is. I would expect a small change to really zoom in/out the entire wave on the scope, but instead it barely moves and even glitches at some points. I checked the resistance on the trimmer and it effectively goes full range (22k).

I could share you a picture of the breadboard, but I don't think that will help a lot finding the problem. But I can post one if any of you wants to have a look. Basically I quadruple-checked the circuit to resemble the simulation I made.

Right now I'm thinking the capacitor is charging way faster than I thought, and there's a delay out of nowhere before it saturates and the square wave switches the transistor to unload it again, making the odd-duty cycle happen in the form of a square-like pattern where it should be triangular. It's also annoying that I can't test it at its full supposed bandwidth.

I really hope I can figure this one out...
So I just measured the bandwidth and seems like the lowest possible frequency starts at 7 kHz, and goes all the way to above 80 kHz, my previous claim that the frequency barely changes is false. Because this is not an exponential trimmer pot, there's just an area where it doesn't seem to have a lot of effect, and an area in which is starts changing rapidly.

I don't know what would cause this higher-than-expected bandwidth. I selected the resistors and capacitor according to the best fit in the simulation. I'm kind of puzzled...
1) In the original schematic look at what the left end of R13 is connected too.
2) Look at the ratio of your R13/R14 compared to the original.
3) Simulations are useful but unless you really know how to simulate things and how to interpret the results you may not be getting real world results.
the opamps wont be able to drive headphones. if you put 1k resistors at the outputs, you might be able to hear some sounds without pulling the outputs down too much.

both scope shots you posted were of the same thing. is the scale right on that? 0.2V/div? thats a really small signal.

this is a linear VCO, so the sweep will be pretty small, maybe 1:100, and most of that change will happen at the low end.

a picture of the breadboard might help.
also, on that scope shot, the horiontal scale makes it look like your frequency is up at 50kHz, which would be inaudible. maybe put a bigger capacitor in.
Thanks for the replies.

As for the case of R13 in the original schematic. Please take note that the LM358 OPAMP has its rails set to GND & Vcc, not Vcc- & Vcc+. So the Vref in that schematic is set to be half the voltage, and it thus oscillates with a DC-component. This isn't the case in my schematic, and is therefore connected to GND instead. Obviously the simulation seems to reflect this.

As I thought, the headphone was pulling the output down cry. I might want to calculate the charging time of that capacitor/integrator. As for the scope shots and the signal scale, I've scaled them down from 12V. Normally they go 24V pp. So that's really a rail-to-rail signal. The horizontal scale is another problem. I should try using another cap. Will post any other findings if make progress!

So I was going to the store to change out some capacitors, even though calculations showed it should have perfect range with the values I initially chose.

Took another look at the setup before swapping the cap out. Turned out my cap was going from the non-inverting input on the opamp to the output instead of the inverting input going to the output. This of course explains the weird behaviour. This did trigger the Schmitt trigger at a much higher rate (the slopes on the waves are effectively the slew rate of the opamp, wownice), and in turn swtiched the input on and off at a much higher frequency.

I was so happy to finally see and listen to a perfect square and triangle wave I made for the first time! Such a glorious moment!

BUT! But I did notice that the trimmer I used as a voltage divider for the input CV was very unstable! It kind of changed the pitch from the very least you touched it. Is this normal? Am I forgetting something here? I really do feel like it would make tuning this a pain! I'm using those basic small blue trimmer pots you can get everywhere, and I was planning to use those trimmers in the exponential converter I'm going to build next. Any thoughts?

Thanks so much guys!
the coarse tuning knob is coarse. trimmers are trimmers. it depends on the trimmer. a 10 turn trimmer will do better than a cheap open frame single turn trimmer. I think it is working correctly. when you have any value trimmer hanging from +12v to 0v you have a 12v range. when you put a 100 Ohms trimmer for 1v/oct trimming it is a lot less sensitive to small adjustments. you need to use the right tool for the job. a lot of people have coarse and fine tuning knobs on the front panel for DIY VCO.

one point of language that I think is important to learn are not anywhere near the maximum slew rate of the opamp. the triangle frequency is exactly a result of the speed of integration using an opamp and a cap. it is called an opamp integrator circuit. you could say this is a function of the square wave getting slewed or the slew rate of the square wave. this is affected by the maximum open loop gain of the opamp, the capacitor value, and the miller effect. there are other factors but this covers the most important things. you might also consider leakage through the cap internally, any currents sourcing or sinking from external connections. there is probably something else I am forgetting.
glad its working. sounds like you had it oscillating, just not under your control! without the negative feedback, the amps were just saturating back and forth.

as for the trimmer, do you have a drawing of what youve built? its hard to say without knowing the values of the trimmer, and how its connected to the circuit.
MUFF WIGGLER Forum Index -> Music Tech DIY  
Page 1 of 1
Powered by phpBB © phpBB Group