Happens on mine, too.
But there's not really a way to avoid it, with analog VCOs.
The popping comes from when the PWM causes a jump in the waveform that isn't in sync with the jumps present in the square waveform itself. For there to be no pops or snaps, the changes in duty cycle would have to occur exactly in sync with the frequency of the square wave, at exactly the right moment in the oscillator's waveform, which just isn't going to happen.
AFAIK, the pops aren't actually peaking above the square wave itself, so it won't mess up a mix or anything. But we hear it as pops because it's jumps in the waveform that aren't harmonically related to the original square wave.
I'm just speculating, this could be wrong.