So why haven't we got a Moog transistor ladder envelope filter project then?

[R.G.]

Stephen, why haven't you mentioned this before? 

I believe that's an implementation of the "Steiner" filter; certainly the diffamp driving diodes as a central cell is.

[SeanCostello]

The Korg-35 filter might be a good candidate for stompbox use:

http://www.aleph.co.jp/~takeda/radio/MS20clone2E.html

I know that a variant of this runs off of a battery in the Gakken SX-150:

http://matrixsynth.blogspot.com/2008/08/gakken-sx-150-schematic.html

It is a 2-pole filter, but it has a nice nonlinear behavior (I haven't tried it in person, but the sound examples are RAUNCHY). Most of the desirable synth filters have nonlinearities within a feedback loop, which results in various cool effects when the resonance is turned up or the input gain is increased. The Moog filter and the 4-pole OTA filter are identical from a linear perspective (4 1-pole filters of identical cutoff with negative feedback) - it is the different nonlinearities that result in different sounds.

A few other ideas for filters:

- A 4th order OTA-based phase shifter design is a good candidate for conversion to a 4-pole lowpass filter.
- By staggering the capacitor values of the different 1-pole stages to unequal values, a higher amount of feedback is needed to achieve the same perceived resonance as a 4-pole filter with equal capacitor values. The diode ladder filters require a lot more feedback to get their sound compared to the Moog ladder, as there is no buffering between the diode filter stages.
- You can insert nonlinearities as needed. The SH101 filter, for example, has back-to-back diodes going to ground in the feedback path. It sounds way more aggressive than the same OTA chip in the Juno-60, which has a fairly linear feedback path.
- The Minimoog and Moog Modular filters used AC coupling for the feedback differential amplifier. This results in highpass filtering of the feedback signal. My theory is that this results in less cancellation of the low frequencies when a lot of feedback is applied, hence the good reputation of these synths for bass.

Sean Costello

[frequencycentral]

The Korg-35 looks interesting, never saw that one before. Some very good observations in your post Sean, almost tempts me to take a bunch of diodes to my Juno-60.

[SeanCostello]

The Korg-35 looks interesting, never saw that one before.

I know that there is a 9V version of the OTA-powered MS-20 filter out there. The schematic I have seen leaves out the frequency to voltage converter, which would be useful when summing several control voltages together. I doubt that the Korg-35 has a particularly accurate sweep range, but if you are doing envelope filtering and/or a CV pedal, this doesn't really matter.

I am wondering which version of the MS-20 filter will be in the Korg Monotron. My guess is that it is a surface mount device. The Korg-35 filter would require a few surface mount transistors and diodes, and would probably be dirt cheap to manufacture in quantity.

Some very good observations in your post Sean, almost tempts me to take a bunch of diodes to my Juno-60.

My long-departed Juno-60 was a great synth, but the filter sound was always weaker than my SH-101. Same IR3109 used in both. It would be interesting to see what the addition of a few silicon diodes (12) would do to the sound of the Juno-60. For that matter, adding a bunch of germanium diodes might be useful as well.

Sean Costello

[Ry]

Hmmm, I have a Juno 60 and a 106...I've been meaning to sell the 106, but it sounds just a little bit more interesting than the 60.  Maybe I could bring myself to part with the 106 if I did this mode to the 60.

Thanks!

[spectraljulian]

Slightly OT:


Why hasn't anyone come up with a pitch tracking filter for guitar yet?  

One of my favorite things about synthesizers

[StephenGiles]

Slightly OT:


Why hasn't anyone come up with a pitch tracking filter for guitar yet? 

One of my favorite things about synthesizers

I think EH did in their Rackmount Guitar Synth.

[spectraljulian]

Ah yes, designed by David %^&*erell.  I want one.  And a VSC-3.  And everything else he designed that I don't have.  Well I guess not the akai stuff, I'm not into that. 

I've looked at that P2V converter on Mark Hammer's site, and man. . . that thing scares me.  It looks like a breadboarding nightmare. 

The pitch tracking is also possible to do with an arduino, but the math right now is scaring me.  It's been far too long since I've sat in a calculus class. 

[space_ryerson]

I've been pondering some form of synth filter adapted for pedal use for a while, but haven't gotten it off the ground yet, partially because I stumbled onto a used EH microsynth. I'm not sure if I like the standard ADSR implementation of the filter envelope for guitar playing versus the microsynth's method. I'd almost want the cutoff to be controlled by an expression pedal, but that quickly will become a pedalboard hog!

I don't know if it helps, but the soundlab runs on 9V, and there's a guitar trigger circuit on the site as well. The filter is different, but perhaps there is some information that can be gleaned from there.

Speaking of filters, have any of you ever used a Polivoks?

[Eb7+9]

http://isp.lynx.net/~jc/pedalsOSH.html

after all this (not to mention some fun control circuit tricks) I realized there's one filter topology that rises above the rest,
one that doesn't spike the ear (variable-state) or isn't lame sounding (ladder) when it comes to gtr

it's not an easy hunt - keep looking! ...

[StephenGiles]

I've been pondering some form of synth filter adapted for pedal use for a while, but haven't gotten it off the ground yet, partially because I stumbled onto a used EH microsynth. I'm not sure if I like the standard ADSR implementation of the filter envelope for guitar playing versus the microsynth's method. I'd almost want the cutoff to be controlled by an expression pedal, but that quickly will become a pedalboard hog!

I don't know if it helps, but the soundlab runs on 9V, and there's a guitar trigger circuit on the site as well. The filter is different, but perhaps there is some information that can be gleaned from there.

Speaking of filters, have any of you ever used a Polivoks?

The EH Microsynth sweep generator is OK as far as it goes, but I once again suggest that a considerable improvement would be if the initial start frequency was controlled by picking strength. One of these days I'll get round to designing this. The sweep generator as is always sweeps to the point set by the start frequency pot when energised by a note played, no matter how hard the note is played, so the filter has no note dynamics imparted on the sweep. The decay is fine - no ripple as far as I can hear, so leave that alone. So effectively, the the start frequency point needs to be voltage controlled - there may already be elements within the control circuitry which would largely do the job, with the addition of a sample and hold.

Any takers??

[StephenGiles]

Something like this perhaps, AX would be a sample & hold - just a sketch


http://www.4shared.com/photo/tk3Z2yJi/EH_sweep_mod.html

[HarryBBD]

Stephen...

that EH schematic looks an awful lot like the Stiener VCF, which is a two pole (basically) Sallen Key
filter. There is lots of info on this one at Synth-diy (check the archives).  I wonder where EH got the
schematic... oh yeah ! it was published as a design idea, or idea for design in EDN or similar by its
daddy, Neil Stiener

H^)

[StephenGiles]

Stephen...

that EH schematic looks an awful lot like the Stiener VCF, which is a two pole (basically) Sallen Key
filter. There is lots of info on this one at Synth-diy (check the archives).  I wonder where EH got the
schematic... oh yeah ! it was published as a design idea, or idea for design in EDN or similar by its
daddy, Neil Stiener

H^)

That must have been a long time ago, I should think EH changed to this sweep generator in the mid 80s.

[slacker]

Something like this perhaps, AX would be a sample & hold - just a sketch

http://www.4shared.com/photo/tk3Z2yJi/EH_sweep_mod.html

That looks like it should work to me. I've got the other version of the Microsynth, but basically the same thing should work for that as well. It would be easy enough to try out, for a quick test in only needs one wire to the start slider disconnecting.

[HarryBBD]

OOps newbie mistake. I'm used to top posting so I was referring to the original schematic for the filter,
which has been identified as the Steiner by others already.  I need to get up to date here

H^)

[Mark Hammer]

You know, Harry, ever since I had the pleasure of trying out your guitar synth in my basement, some years back (and the envy has not subsided), I have lauded it constantly.  At the same time, I think I'm just abut the only person among the 18,303 members today, other than yourself of course, who has heard it.  Any possibility for a soundclip that might give people a sense of how it tracks?

[slacker]

Apologies to Rick for swinging off at a tangent

AX would be a sample & hold - just a sketch

Thinking out loud here, but do you even need the sample and hold? Once the trigger from the "attack detector" has kicked off the filter sweep does it matter if the start voltage changes?

[SeanCostello]

It looks like the Korg Monotron has a Korg-35 filter variant, running on +5 volts:

http://www.timstinchcombe.co.uk/synth/korg.html

No schematic yet, but I'm sure one will emerge in the next few months.

Sean Costello

[StephenGiles]

Apologies to Rick for swinging off at a tangent

AX would be a sample & hold - just a sketch

Thinking out loud here, but do you even need the sample and hold? Once the trigger from the "attack detector" has kicked off the filter sweep does it matter if the start voltage changes?

I don't know, I just whan the start voltage to be dependent upon the note picking level, which would then acheive a more or less perfect sweep generator as the sweep decay has no stutter at all.

I don't have a Microsynth so I would need to breadboard most of the circuit to try this out at some point.