[sdiy] AS3320 VCF chip - some queries
tom at electricdruid.net
Mon Nov 12 10:45:33 CET 2018
I’ve done some more experiments and had some success.
I noticed that the Sequential Pro-One uses a buffer between the output and the resonance VCA. In fact, it’s not just a buffer, it has a gain of x3.4.
I was thinking of adding a differential mixer to add in some of the input signal to the signal fed back to the Resonance VCA to provide a volume-drop compensation. So I did, and gave this mixer a little bit of gain (x2 in mine).
Big improvement! Not only do I now not have a volume drop with increasing resonance, but the oscillation signal level is constant across the whole range. Checking the Pro-One again, its resonance is also very stable in level across the whole frequency range, and this must be why they added that buffer. I’ve always wondered what it’s for!
Synth & Stompbox DIY
> On 12 Nov 2018, at 02:13, Andrew Simper <andy at cytomic.com> wrote:
> Good point about the main filter caps David. Two other things to watch out for to make low frequency self oscillation happen with filters:
> (*) make sure any DC blocking caps in the resonance signal path are well out of range of where you want the filter to self oscillate down to, otherwise the phase and gain changes can throw off normal resonance
> (*) make sure you use good buffers between stages to not drain out the current from your caps.
> With the 3320 the buffers are on chip so this makes things more difficult if they are causing the issue. Do you know what they are? If they are a part of the problem then one way to mitigate it somewhat would to use larger valued caps, and then likewise increase the cutoff current so the relative drain of the buffers is lower, but this may impact the highest cutoff you can get.
> On Mon, 12 Nov 2018 at 02:17, David G Dixon <dixon at mail.ubc.ca> wrote:
> You might find that the filter will oscillate easier if you use good film caps. I had problems getting a four-pole filter to oscillate at low frequencies with monolithic ceramic caps, but it oscillated strongly down to the minimum frequencies with WIMA polyester caps.
> I just built an SVF with silvered mica caps (which I usually use in oscillators) and it is being finicky with self-oscillation below about 50Hz. Maybe I'll swap those out for WIMA polyester and see.
> From: Synth-diy [mailto:synth-diy-bounces at synth-diy.org] On Behalf Of Pete Hartman
> Sent: Sunday, November 11, 2018 9:20 AM
> To: Tom Wiltshire
> Cc: synth-diy at synth-diy.org DIY
> Subject: Re: [sdiy] AS3320 VCF chip - some queries
> On Sun, Nov 11, 2018 at 10:32 AM Tom Wiltshire <tom at electricdruid.net> wrote:
> 2) The resonance response is tilted to the high end.
> It seems to require a lot more resonance CV (ok, current) to make the chip oscillate at the bass end than at the treble end. Similarly, the amplitude of oscillation increases as the frequency rises. Why is this? Is there anything I can do about it? Should I even care?
> If I compare this with the behavior I've seen in other lowpass filters that self oscillate, even without CV over the resonance, it seems consistent. Typically I need the frequency at 50% or higher to get it to start oscillating -- I can then commonly turn it down and get lower frequencies, but my experience is that if the filter is going to oscillate, it's always easier to start the oscillation at a higher frequency. And typically if I turn it "too low" then it stops oscillating again.
> Synth-diy mailing list
> Synth-diy at synth-diy.org
More information about the Synth-diy