Ideal VCO for mass production

[jorgen.bergfors at idg.se]

>> I have built the Doepfer PMS quad VCO, which is rather similar to the ones Tomg has on his site. It uses CA3086s for heaters and expo converters. I'm not quite satisfied with it (...)

>Could you elaborate a little bit and tell us why ?

Well, I suppose it is kinda OK, but the ASM-1 VCO definitely is better.

* The PMS use 555 timers, which has slightly varying threshold voltages. This means that the pulse width modulation will not be the same from VCO to VCO. This is of course most noticable when you use really narrow pulses.

* I already mentioned power consumption. I think my quad module consumes 500 mA from the positive rail and 300 mA from the negative rail. And that is after the heater temperature has stabilized. While heating it is around 700 mA from the positive supply.  I'm building a very compact portable synth and I don't have room for huge heatsinks or maybe dual power supplies.

* The sawtooth wave is not quite linear at low frequencies. The ASM-1 also has a slightly quicker reset time, which is important when you add waveshapers like saw to triangle.

* The fact that the ASM-1 doesn't have a high range trimmer simplifies adjustment a lot. The PMS is rather sensitive to adjust. That's fixable by adding resistors in series with the trimmers, but it is difficult to find room for them on the board.

* The oscillators sync slightly even when you don't want them to. They don't lock completely, but the beating is jerky when they are almost in sync and the frequency difference is less than approx . 1 Hz.  I think this is the most serious problem and I haven't found a fix for it. Any VCO design must be tested running several oscillators at almost the same frequency. Before that the design is not finished. The beating between oscillators is very important to get that lush analog sound.

/Jorgen