Sound Quality Variation?

[antto]

guest: my nose tells me the difference is mainly in the feedback DC filter
this is what makes sub-peak

and my oppinion on the subject: i'm don't have a mania for the filter that much..
the only important things you need are two:
enough resonance (make the reso-boost mod if needed)
decreasing level of resonance at lower freqs (this is done by the feedback DC filter already)

any filter which has these will work just fine ;]
even a 12dB StateVariable! yes.. it's still gonna make acid ;]

[textile]

Cool Don!
I am doing the same but with THAT300 as these will share thermals together. The MAT series have a good reputation and would go good in the VCO.
Although a 300p would house all the NPNs in the VCO.
The stock out there of the 2291 and 1583 are shocking... there are nice ones left but I think its reached the bottom.

I think that with precise matching of to-92 transistors one could replace the dual NPNs with to-92.
I stress precise , as the story goes Dr Moog only changed to monolithic transistor ICs for increased production and the early model D and Modular used only hand matched, the schematics say so as well.
By using the method of matching transistors as stated in the service manuals from Moog one could very well have an authentic Transistor Ladder VCF. Free you mind ,and you know Yo' ass will follow.
I dislike buying those matched transistors but people wont buy with-out..



Is the VCF 18db due to the caps ? The structure is 4 stage but as with the Moog the caps are equal .
Rule : Low Pass Filters require Capacitors to function. Put a larger cap in there and the frequency changes , like in the Moog Modular VCF (i can't recall the Model#) frequency switching is made by switching caps - equal ones at that.


I am a little mania induced about VCFs I have made several types (and sold 3 different types), all the main topologies and the sound variation with them all is different. Some filter faster some filter hard and some more wet and flabby - kind of that sweaty guy who hangs out at the greasy chipy.

The TB vcf sounds more like a 2 pole than a 4 pole to me.
3 , 3 and a half .... almost 4 ? It is interesting as to what it exactly is.
4 stages yes , to my ears 4 pole ? ..... no. 

[antto]

it's 4 pole, with the first pole ~1 octave higher than the others
so it varies between 18 and 24dB

[Acrobatics]

it's 4 pole, with the first pole ~1 octave higher than the others
so it varies between 18 and 24dB

sorry, does it varies with amplitude or frequency of the incoming signal?

[Altitude]

THAT300 is an excellent modern replacement. My TM3030s have those and they are very good

[antto]

Acrobatics: varies with frequency a little (or more)
but it's more complicated than that.. there is a strong DC filter in the feedback path
when there is no resonance at all - the DC filter doesn't make any difference because there is no feedback
so you get kinda between 18 to 24dB lowpass filter (which sounds pretty good)
then, when you turn up the resonance - phases change, you get less resonance on the low frequencies, more on the highs (unlike a moog ladder, where the resonance level is almost equal)

on the other hand.. if you replace the x0x filter with a moog ladder, and modify the ladder so there is a similar DC filter in the feedback path (with a frequency somewhere from 72 to 159Hz i can't really say)
then you wouldn't be able to really tell if there is something wrong with the x0x or not, it will still be doing acid
the only big difference is when you turn resonance off
the moog filter doesn't lose so much of the gain with high resonance, compared to the 303 filter, which needs more feedback and thus - the gain at high reso drops more

so, the 18 vs 24dB banned doesn't matter (at all) .. what matters is the resonance level vs frequency ;]

[Brassteacher]

Acrobatics: varies with frequency a little (or more)
but it's more complicated than that.. there is a strong DC filter in the feedback path
when there is no resonance at all - the DC filter doesn't make any difference because there is no feedback
so you get kinda between 18 to 24dB lowpass filter (which sounds pretty good)
then, when you turn up the resonance - phases change, you get less resonance on the low frequencies, more on the highs (unlike a moog ladder, where the resonance level is almost equal)

on the other hand.. if you replace the x0x filter with a moog ladder, and modify the ladder so there is a similar DC filter in the feedback path (with a frequency somewhere from 72 to 159Hz i can't really say)
then you wouldn't be able to really tell if there is something wrong with the x0x or not, it will still be doing acid
the only big difference is when you turn resonance off
the moog filter doesn't lose so much of the gain with high resonance, compared to the 303 filter, which needs more feedback and thus - the gain at high reso drops more

so, the 18 vs 24dB banned doesn't matter (at all) .. what matters is the resonance level vs frequency ;]

antto, what do you mean by "DC filter"? Could be a semantics issue, such as when audio people refer to a "crossover", synth people say "filter". Or when synth people refer to "oscillator", audio people say "blowed-up amp" .

[antto]

DC-Filter = 1st order Resistor/Capacitor High-Pass filter
it's role is to take of the frequencies near DC (well, you know what a HighPass does)
but, since it's frequency is pretty high (72Hz) it has a greater effect in the filter

as far as i know, the moog ladder also has a DC filter in the feedback path.. i guess this is needed technically
but it probably is at a very very low frequency, because the moog ladder doesn't sound like losing so much resonance on the low cutoff frequencies (around and below 100Hz)

on the 303 ladder filter, the resonance at ~100Hz is much less and it goes away on lower freqs too
this is one reason why the 303 sounds so good with a distortion ;]

[Brassteacher]

DC-Filter = 1st order Resistor/Capacitor High-Pass filter
it's role is to take of the frequencies near DC (well, you know what a HighPass does)
but, since it's frequency is pretty high (72Hz) it has a greater effect in the filter

as far as i know, the moog ladder also has a DC filter in the feedback path.. i guess this is needed technically
but it probably is at a very very low frequency, because the moog ladder doesn't sound like losing so much resonance on the low cutoff frequencies (around and below 100Hz)

on the 303 ladder filter, the resonance at ~100Hz is much less and it goes away on lower freqs too
this is one reason why the 303 sounds so good with a distortion ;]

Ah, I'm used to calling such a device a "coupling cap". Sometimes a resistor is present, sometimes it isn't. No resistor usually requires a much bigger cap to get the job done. High-end audio circuits almost always use no resistor to avoid adding more Johnson noise to the circuit. However, circuits that use tubes use the resistor much more often. Their real main purpose is to remove a DC offset from a signal path when any DC could either damage or negatively affect the next stage of a circuit. They do serve a dual function as a high-pass filter for AC signals.

The Moog ladder sometimes has a coupling cap, sometimes not. The earlier versions, such as in the Minimoog, do not. I haven't studied the schematics for all the different Moog synths yet (but I have studied quite a few), but the presence of a coupling cap seems to correlate with the buffer/differential amp being discrete transistors or opamps. Discrete=no cap, opamps=cap, in general. Of course, we all know how good most generalizations are...

[CoolColJ]

My unit, one of the original batches, no. 173. A friend built it for me. still going strong after all these years

Has hand picked high beta transistors, and PSU mod via trimpot - sound clip set at zero, half and full on the trimpot
Makes the resonance sound different at each setting. I think this is the key to the 303 sound of the original.
Also my unit has more bass than a standard 303, removed the low end cut
http://members.optusnet.com.au/coolcolj ... f_full.mp3

Filter input mod allows one to feedback the headphone input back into the unit for more resonance
http://members.optusnet.com.au/coolcolj ... edback.mp3



PSU mod trimpot next to power switch at the back (another mod), plus filter input, switch to turn oscillator off, and another switch for extended envelope modulation, and the filter can go lower than an original 303