Friday, February 18, 2011

HARTMAN ENVELOPE FILTER MOD JOB


 


this is a commission i took on about a month and a half ago. the idea was simple really. a friend of a friend wanted to add an expression pedal to his Hartman envelope filter. after doing some quick
research on the Hartman, i found that it is a clone of the famous mutronIII pedal. rather than just agreeing to add a simple expression pedal to open and close the filter, i suggested that he(the owner) have me add as many mods as possible. i suggested adding the three missing filter modes highpass, bandpass and notch filter outs. i also suggested adding attack/release controls for the envelope, and a multi-shape lfo to modulate the filter. i gave the owner a pretty reasonable quote for the commission and he agreed to go for it. initially i thought that this project would not take more than a couple of weeks to finish. as always, i was sorely mistaken. first of all drawing out the schematic alone took an entire day. the pcb is double sided. i thought it would be identical to the mutronIII schematic, and later i found that for the most part it was, but the way the circuit was laid out, the only thing i could recognize was the filter. i had to redraw the rest of the circuit point to point only to find that the envelope generator was the same too. the only thing Hartman seemed to add was a mixing buffer to be able to mix between dry and wet, and some kind of power conditioner that i don't fully understand. see my original drawing  for more into that...
 
 another issue i had with the Hartman circuit is that it just seemed to be very cheaply constructed. apparently this particular model was a prototype. it had numerous extra parts and wires that went nowhere or were redundant. i thought that was kind of cool. i however didn't think it was very cool that the solder used in this thing seemed to be either lead free or flux free. either way, that shit was awful. it was all dry and crusty and it would "infect" any fresh solder i would add. another thing was that the traces on the pcb were extremely thin and fragile. they could be scratched right off with no effort. the wires coming off of the pcb to the hardware broke and had to be reattached so many time that eventually i just replaced the worst ones with solid copper wire. despite all my gripes with having to work around someone Else's work, the circuit sounded pretty great. i came up with a plan that i thought might work, and drew up a schematic for the mods that would be added. i picked up this vco chip years ago and I've been meaning to use it. its called the icl8038 and from what i understood, it is a pretty nice function generator chip with sine, pulse,and ramp outs. i thought it might be cool to be able to modulate the filter with the lfo and then use the envelope from the Hartman to modulate the depth, rate or duty ratio of the oscillator, as well as being able to control all of those parameters in addition to the wah filter via the expression pedal. i had planned to put all of the additional circuitry in the expression pedal itself. in order to get those 16 connections that i would need from the expression pedal to the Hartman, i would need my old friend the MDR cable. fortunately too i was actually able to find one at goodwill for only $5. the only catch was that it was attached to a floppy disk drive. an MDR cable is a Mini D Ribbon cable and they are awesome! they can carry as many as 40 leads and they have these wonderful squeeze-lock connectors  that are super strong. i would love to know where i can find more of these. i found them on mouser but they are non-stocked. they are super rare  to find in thrift stores too. everything is usb now so the old floppy drives that employed them are probably all off to the recycling. if you know where to find MDR cables with the quick release connectors, please let me know.
one thought that crossed my mind is that in order to reroute mods from the Hartman to the expression pedal, the original functions would be disabled when the expression pedal is disconnected from the hartman. i came up with a plan to have all of the mods switched to the the expression pedal through a cd4066 digital switch that would activate when the connector is in. it was a pretty slick feature. unplugged the Hartman would be stock, plugged in, the expression pedal would take over. sadly though, one of the leads that the 4066 was responsible for switching was the -9volts which the 4066 didn't appreciate very much seeing as how it was running off +9 and ground, and it died on me. i considered running the 4066 off ground and -9, but the input for the switch was the ground. i figured no matter what solution i came up with to be able to use 4066, it would inevitably die anyway. they always do. so i ripped out all my work and replaced it with 4 pole switch. its not nearly as slick, but just in case something awful happens to the exp. pedal rendering it useless, the Hartman can still be used as it was originally intended.
 
coming into this project i have to say, i didn't know that much about Vactrols. the Hartman filter is modulated via matched vactrols. vactrols are basically an led stuck to a light dependent resistor encapsulated in a light proof box. the more light, the less resistance. i picked up quite a few vactrols for a good deal a while back and i thought i would use them for this project. in my original schematic i had planned to to implement all of my mods with their own respective vactrols. i soon found that not all vactrols are the same. all vacrtols exhibit some latency, even the fastest ones. the vactrols take noticeable amount of time to peak and disengage when the current to the led is cut off. the vactrols in the Hartman are pretty fast but the vactrols i have are painfully slow to disengage. i had to go back and reconfigure my schematic to use only the vactrols already in the Hartman to modulate the filter with the expression pedal and lfo. the other vactrols and optoisolators i used for modulating the lfo parameters are not as important as the response from the lfo or pedal to the Hartman filter. although the lfo modulates the filter via the original vactrols, they still exhibit latency too. i noticed that at higher lfo frequencies the filter would stay closed because the vactrols wouldn't have enough time to disengage. and the frequencies weren't even that high... i had no choice but to literally rectify the situation. the vactrol cathodes go to ground and not -9 volts. this is how it has to be in order to be able to invert the envelope with the direction switch. the signal from the lfo would have to be half wave rectified. i tried to run the lfo off of 9v and ground but it wouldn't oscillate. below is a diagram of whats coming out of the lfo(left), and what is modulating the filter(right). those shapes are probably distorted too because of the latency of the vactrols(i can only assume without plugging it into an oscope).

so yeah, the wave shape names on the side of the expression pedal are totally false but still quite usable. who knows, maybe the solution will come to me before i ship it out on monday. oh yeah, did i mention that the icl8038 didn't work out? yeah, that was supposed to be before all the vactrol nonsense. so the icl8083 is basically a piece of shit that is worthless in more ways then i care to mention... actually its not that bad. I'll just never use one for anything. instead i found this super simple tri-shape lfo schematic built around a quad op-amp. it works great and i am sure i will be using it much more in the future.
ok, where am i going with this? i guess now is where i describe the finished pedal. i painted the base of the roland EV-5 pedal orange and applied white lettering for the parameter labels. i then finished it with clear lacquer.
the PED switch assigns the expression pedal to the lfo rate, lfo depth, filter or off. the MODE switch selects the filter type highpass, bandpass, lowpass, or notchpass. the silver knob sets the maximum depth of the expression pedal.
 
 the ENV switch assigns the Hartman envelope to the lfo rate, lfo depth, filter or off. the the LFO switch selects ramp, square,sine, or none waveshapes. the toggle switch under the expression pedal switches between stock Hartman envelope and custom envelope. the two pots control the attack and release of the custom envelope when engaged.
 
 the DEPTH and RATE knobs control depth and rate or initial depth and rate of the lfo.



 this is a sort of diagram for how it all works


 here are some pictures of the inside of the finished pedal prior to some extensive debugging.









 her is a crappy video i threw together. its too hard to try and play guitar, shoot a video and tweak a pedal at the same time so i just ran my omnichord through it. I'll try to make a better demo if i have time.
thanks for reading/watching.
-tanner


7 comments:

  1. Masterfull.

    A goodwill in town is dedicated to computer stuff. I went yesterday and no MDR cable. But Ill check again.

    POE

    ReplyDelete
  2. in portland!? where? i found a few mdr cables at freegeek too.

    ReplyDelete
  3. This sore in Austin, where I live.

    http://www.yelp.com/biz/goodwill-computer-works-austin

    I'll make a visit in the coming days and look again.

    ReplyDelete
  4. Hey Tanner,
    Good to see you today, it's been a long time eh?
    I checked out your blog, crazy stuff; I don't understand a lick of it, but I'm so impressed that you do and are having some fun.
    Keep it up, and take care.
    Uncle Dwight

    ReplyDelete
  5. trying to build something with a super simple LFO, require sine, whatschem did you use? any chance of linking? nice build BTW

    ReplyDelete