From Bruce Duncan Sent Fri, Nov 10th 2006, 19:06
From Bruce Duncan Sent Fri, Nov 10th 2006, 19:06
Actually Paul the Modcan filter that you used for testing was the Low Pass 02a which is a clone of the Mini Moog filter. Hence it not sounding exactly like the 904A. The MOTM-490 is the *only* commercial VCF that mimics *exactly* the 904A responses. Surely part of the sound character of the 904A comes from the 3 sets of ladder caps that can be switched in with the 904a range switch. How can the MOTM-490 be the *only* commercial VCF that mimics *exactly* the 904A responses when you don't include this feature. I would say that my clone of the 904A would more likely hold this position as I followed the 904a schematic to the letter including the range switching. Bruce At 11:01 AM 11/10/2006, you wrote: >When I was designing the MOTM-490 Moog VCF, I made many accurate >measurements with my Audio Precision analyzer. I compared the following: > >a) 5 different 904As, spanning 8 years of production >b) 2 MF-101s >c) Modcan VCF >d) .com VCF >e) Doepfer VCF > >The following measurements were run: > >a) Frequency response >b) THD versus frequency response (at 1V pk-pk input) >c) THD versus input level (how does the VCF respond at different >levels, and at overload?) >d) Multi-tone IMD (inter-modulation distortion, a measure of how >different frequencies are treated as they are beating against each other) > >Results were as follows: > >1 - all the 904As were close to each other (within say 4%) in the tests >2 - the MF-101 was the *least* closest >3 - the 3 commercial VCFs varied widely from each other, but none >was close to the 904As >4 - the "smoking gun" in the question "Why does a 904A sound the way >it does?" has to do with test B. The 904As have a very distinct >"bump" in the THD curve when plotted against frequency (ie the >distortion is a funtion of the input frequency). I thought this was >odd, and I even called Bob Moog on the phone and we discussed this >at length. It turns out that certain.....errrrr....parts decisions >were used based on the parts he had availble at the time (he liked >to buy surplus parts from the Long Island electronic people). And >the combination that he used created this "bump" *by accident*. But >it was a happy accident :) In 1967, there was no test equipment to >make the measurement. > >So, after...errrrr...'reverse engineering' the other designs, I saw >that none used the same technique/impedances that the 904A did. So, >I did :) The MOTM-490 is the *only* commercial VCF that mimics >*exactly* the 904A responses. > >Also remember: the 904A is *NOT* the same as a Minimoog VCF. It will >*NOT* sound like a Minimoog *exactly*. In order to get that sound, >you need both the VCF and the Minimoog VCA interconnected, because >the input impedance of the VCA loads the current sources of the VCF >and that also causes a THD term. > >Paul S.