Joined: Mar 09, 2010
Location: United Kingdom
|Posted: Fri Dec 20, 2013 1:46 pm Post subject:
SDIY EB Mods - possible?
Subject description: VCO and/or LFO mods
|Hello all, I posted about this over at Muffs but didn't get a single response, perhaps it was the wrong place for it, or perhaps it is just my essay problem....
I know this is a long post, but I think there will be some cool stuff at the end of it if I can get help, I will share all results and have already been keeping a good build diary for the project which should benefit anyone else
I am building Ray Wilsons Synth DIY Experimenter - I have populated the PCB's and I am working on the front panel. In the mean time I have also been messing around with some op-amp circuits on the breadboard and as a result, I think I can make some nice mods to the SDIY EB, but need some help and advice to make it work:
On the breadboard I was messing with a basic oscillator circuit using an integrator and comparator pair in a feedback loop, you can get a triangle or a square wave from the op-amp outputs. Looking at the schematics, I am pretty sure the VCOs and the LFOs on the SDIY EB use a similar integrator/comparator pair to this.
Ray Wilson did a webcast recently and he was discussing using the integrator/comparator set up, and he dropped the bombshell that you could use a small signal diode to couple the inverting input of the integrator to ground to get a ramp or saw wave (depending on the direction of the diode) - so of course I wanted to try this right away
I found on the breadboard that the diode didn't work at the inverting input of the integrator, but did work perfectly at the output of the comparator (before the feedback resistor that goes to the inverting input of the integrator). The only downside is the ramp or saw is double the duty cycle of the square or triangle, so is exactly an octave lower. However, it is very cool that with just a few components you can have 4 different wave shapes - you could mount two diodes to a switch to get the ramp and saw.
In work the other day I started wondering if it would be easy to do a hard sync with another oscillator using the integrator/comparator pair - at home I tried the idea out and it is!! Using a second breadboard pair, I capacitively coupled the output into the non-inverting input of the comparator on the first oscillator pair, and it works a treat!! You can even attenuate the sync output to get a morphing effect
The schematic below shows these ideas at work:
So - to the point, I am really hoping that I can add these ideas as mods to my SDIY EB project. I drew up the plans (see below) and tried them out today, but neither worked particularly well. The ramp/saw mod doesn't work at all, but by poking around I found that I could get a saw wave by connecting the comparator output to the inverting integrator output with a diode and a 20k resistor - but no ramp wave by reversing the diode. Something weird happens at low frequencies though; the square and triangle waves will normally stop oscillating when you turn the frequency down, the saw tooth at this point however starts going up in frequency again - weird!
I haven't properly tested the sync yet because I have only properly wired one of the oscillators - the other side is oscillating however so I was able to try taking the triangle and square outputs and capacitively couple them with the non-inverting input of the comparator on the other side. This kinda works, but like the saw tooth, weird things happen with the frequency control - the frequency is high at either side of the pot and low(ish) in the middle. Why this is happening is completely boggling my mind!! hmmm..... I am going to wire up the other oscillator when I get back from Christmas with the folks, just because it is meant to run through another inverting buffer before the output gets patched elsewhere - this might make a difference
The are a couple of differences between the SDIY EB oscillators and the pair I built on the breadboard, but the key difference I am worried about is that the LM3900 in the SDIY schematic runs from a single supply (12V to GND) whereas my design works from a bipolar supply, 12V to -12V. I am wondering if this means that the ground bus is working in a different way in the two circuits, at least as far as the math is concerned? - I guess the transistor VCR could have something to do with it as well...
I can post more schematics if it helps - will need to scribble them down first! Well done for reading this far, I would massively appreciate any help - will post any further results I get in the mean time