Slow Gear build report and mods

[Pedro Freitas]

Hi!

Built a Slow Gear from GGG plans and it works. Just removed the
FET bypass and I'm bypassing it with a DPDT.
Event detection (strumming or notes) is poor, just like playing with
a noise gate, it doesn't reinitiate a new cicle with each strum action...
Bummer...
Changed the 5,6Zener/3.3K VR bias voltage source with a 22K
trim pot and lowered the bias voltage to 3,9. That seemed to improve
note picking cicles but it's like polishing a turd  
I guess that it's as good as it'll get regarding event detection but
can anyone tell how to extend attack time? It raises the volume
too fast.

Does anyone know of a similar efect that fades in chords or notes
without damping the strings first???
Thanks for any input!
Pedro

[Travis]

You may want to try the Morley Pik Attack.  It is MUCH more controllable.  The envelope detection is based on making/breaking a circuit to ground with a metal pic and your guitar strings.  It has its own fairly obvious defects, but is significantly more flexible than the Slow Gear.

The speed of regeneration is controlled by a led/ldr footpedal.  This and a warbly pedal can give some pretty convincing reverse tape sounds.

They are running c. $100 on eBay, but you can always contact the guys at http://geocities.com/tel_ray/ for the schematic.

I really need to look inside of mine to see how complex the circuit is.  If I can get into it in the next couple of days, I'll PM you.

[RickL]

The PAiA Gator does pretty much the same thing with about the same tracking. There is also a posted schematic for a String Damper that also does about the same thing with maybe a bit worse tracking.

The best one I've used is the old Morley Pic series. It uses a special metal pick attached to the pedal that retriggers the attack each time you touch the strings. It tracks almost flawlessly, retriggering on each note but the pick is a bit of a pain to use with its attached little cord. The pedal adjusts the attack time which is kind of nice. Unfortunately I don't recall ever seeing a schematic for it. Maybe check Morley's site or ask them directly.

I tried doing the same thing with my Slow Gear with a metal pick attached to the input of the pedal that grounded the input every time it touched the strings and it worked but I had to insulate the pick or I got horrible buzzing. I gave up on the idea but it might work with a little tweaking. Maybe the pick has to go to somewhere other than the input or has to be electrically isolated with a cap or a resistor. I never tried any of those things before I gave up.

[Mark Hammer]

Deciding what you're going to do with a signal when it starts is an age-old problem.  It is even more problematic if the signal itself has enough spurious elements that it needs to be compounded by a decision IF it is a signal.  

In the dying days of vinyl, electronic "click-and-pop" eliminators would substitute small bits of silence for anything it detect as likely vinyl surface noise.  Since that noise was often in the same frequency band as actual music content, the analysis needed to be sophisticated enough to be able to distinguish between clicks and other things.  Of course, if what you're trying to suppress is an even that is MAYBE a couple of milliseconds long, we're talking about one VERY fast noise gate.  What a number of manufacturers did was to literally "buy themselves some time" by running the audio signal through a BBD-based delay section - just for a few milliseconds.  The analysis circuitry used the pre-delay signal, so that it could figure out what it needed to figure out, and turn the gate on or off just in time.

The old E-H Attack Decay pedal used a similar strategy.  If you look at the schematic, you'll see there is a fixed delay section in there which delays the signal to be acted upon until such time as the envelope-control circuitry has done its bit.  This way, the circuit sort of "knows" there is a new note coming (because it has a kind of "advance warning") and readies itself to gradually fade that note in.  In the absence of the delay, the circuit would end up responding with clicks and such at the start olf notes, as if to say "Holy shit, is that a NOTE?  I better get cracking!!".

So, the use of a small pre-delay can be effective in helping analog circuits enact changes in the onset of instantaneous changes in amplitude, whether very brief - like clicks and pops - or somewhat longer, like chords.  Of course, dragging a BBD into it does make it a costlier design, a somewhat more problematic one with respect to layout, and BBD's have their own baggage.  The other thing is that it is one thing to delay both channels of a stereo recording by 2msec in playback, and quite another to delay a single instrument in a group context (whether live or virtual) by 5msec.  The delay time needs to be artfully selected so as not to derail the player's timing.

Finally, I've said it before and I'll say it again.  In the "real world", what changes over time is not just a note's amplitude, but its harmonic content.  If you want things to feel like a reverse-tape swell, they not only have to get louder, but they have to get somewhat brighter.  So, in some respects, the "ideal" slow attack pedal has two things controlled by the envelope circuitry: a VCA or some other controllable attenuator for amplitude changes, and a simple shallow lowpass filter (what amounts to an all electronic version of a RAT-style tone control) for bandwidth/brightness changes.

That being said, I've got my iron-on transfers for a Slow Gear board ready, and despite having a million other things to do, I'd still like a crack at making one.  It may well be possible to have two FETs controlled by the same envelope signal, with a second one being placed in series with the audio signal, perhaps situated between the drain of the first one and the 1uf cap that comes after it.  A small value cap would then go from the source (well, the junction of the 1uf cap and the additional FET) to ground.  Of course the trouble with this is that the amplitude FET needs to go from low resistance to high resistance at the start of each note, while the tone FET has to go from high resistance (= lower treble rolloff) to low resistance.  That would either involve some sort of inverting stage or else ganging an N-channel and P-channel JFET to the same envelope signal.  Frankly, I don't know enough about such matters to know how to do it, or if it could even be done without significant modification to the circuit.

[Pedro Freitas]

Thanks for the replies!
Hmmm, that delay thing to get the detector ready in time is a must.
But hey, I'll be glad to just extend my attack time in the SG right now
Any ideas???

Pedro

[free electron]

just a thought:
how about using dynamic expansion at the input of the envelope detector. Use fast attack and release times, just to expand "pick" moment. With  threshold set properly it gives exactly boost of the attack "pick" and attenuation of the rest of envelope. NE571, or 572 + one OA for expander, with the other half we can do excellent low noise, low distortion variable gain cell... A lot of tweaking, but maybe worth trying..
Regards
Peter