904a question

[EricK]

Can anyone comment on the decision to put the Frequency Range switch on the 904a as opposed to just sweeping the entire cutoff frequency with the pot? Does this make it sound better or is this to facilitate more precision like when one has the LFO range split between high and low?

Thanks,
Eric

[CZ Rider]

Not sure why Bob Moog decided to switch in/out capacitor values to get the different ranges? The original 904 bandpass was quoted as having a three octave range and Bob's first filter. That one made sense as the three octave range could be shifted and had 4 range settings via switched capacitors. The 904A has a much wider range so shifting the range doesn't quite have the same need. However there is a difference in sound/tone and it is not just voltage shifting into a higher or lower range.
The setting 1 is perhaps the most drastic as the filter frequency response is 1Hz to 5,000Hz. Very noticable that the filter will not go much above that 5,000Hz wall. It is a soft wall and can be pushed a tiny bit, but no matter how much control voltage applied, it will never go much higher. Easy to hear with white noise, not as much sizzle as in the 2nd and 3rd range setting. Not important if you approach this as a standard filter and the low range could be eliminated. But if you are after a particular sound you may have heard on an album recorded with the 904A, it may be difficult to replicate without that 5,000Hz wall. I know a few Berlin school tones I have listened to that seem to have the range 1 selected.
The second range is more like what they dialed in on the Minimoog at 4Hz to 20,000Hz with full hearing range and the third setting with 16Hz to 80,000Hz that dogs can hear. Never noticed either setting not sounding deep enough, but the high end trim on the first setting is very noticable and a classic tone. Depends on what you like, but the 904A gives you a choice.

[MC]

Like the Minimoog, the 904A had non-linear resonance. Make the filter self-oscillate and you will notice that it does not oscillate the full range of cutoff. On the Minimoog, the oscillation disappears on the low end. The reason for that was the excess resistance in the recovery amplifier at the top of the ladder. The switching ranges on the 904A was to get oscillation in any range. For musicians exploring FM or using the 904A as a tone source that was important.

[EricK]

Thanks very much, guys.

[CZ Rider]

The switching ranges on the 904A was to get oscillation in any range.

But the original 60's made 904A's did not self oscillate. Wasn't untill a later revision change they decided the sine oscillation was a good thing and lowered the feedback resistor. My stock summer of 1969 built 904A did not self oscillate untill I modified it to effectively change the 1.8K feedback resistor to 1K. Have a few patches from a 1967 book where the filter is constantly set at 10, and if that filter was self oscillating, none of those patches would sound right. Guess it was around 1970 or later when that self oscillation revision was implemented. Could always patch in a negative feedback loop to make the 904A oscillate though if you wanted on those early ones.

[MC]

Yes the original 904A did not self oscillate, but the resonance was not linear across the range of the frequency pot.

Switch to the high range, crank the resonance (not to self-oscillation), adjust the frequency control ("fixed control voltage") almost CCW, then switch to low range and raise the frequency control to the same cutoff point and you will find that the resonance isn't the same. The 904A does not maintain the same resonance as you approach the low end of the cutoff control.

[CZ Rider]

Absolutely, and a good point. Not only is switching the caps changing the frequency range on the filter, but also the regeneration response/range.
I'm still in the planing stage for a Minimoog filter clone that switches caps like the 904A does. Interested to hear if there is a difference between the various brands used on the original Minimoogs over the years, along with different values. So far I have matched sets of the .068uF in the Mullard tropical fish type like R.A.Moog Minis used, El-Mencos like the modulars used and the white box type the latter Minis used. Along with some larger values to try like .67uF in the Mini ladder. Eyeing up a Centralab 9 pole, 11 position rotary to do this filter cap switching for both the four ladder caps and the two .22uF buffer caps. Should be an interesting experiment to see how this all effects the Minimoog filter. Might even be fun to tap the filter for different polls too. Have to get busy with that project.

The various sets of filter caps I plan on swithching in/out.