Here is a close-up of the circuit cards:
Each one has a different prototyping method used, because I am still working out what my preferred prototype method is. PCBs can be made from PCBexpress or other online houses for fairly cheap but they take a lot of time to lay out and get made etc. so it's sort of against the concept of quick prototyping circuits unless you know what you want in the end. I am pretty much against etching PCBs at home because the chemicals involved make it the least green method for DIY'ing circuits.
From the left, the dual LFO circuit is a 2-opamp integrator/comparator with variable pulsewidth and ramp/square outputs. It's on pad-per hole board with solder blob interconnects. The pad per hole method works out fairly well for me but the blobs leave a bit to be desired when you want to connect neighboring pads. The next module is the acoustic noise source, in the back of the pic is the fan and a compaq electret mic capsule, the preamp is built breadboard style, on a block of maple with little copper or brass nails that the components are soldered to. The circuit is a discrete 'JLH 1969' type of amp, based on the old John Lindsey Hood topology. I am pretty happy with the breadboard method, you can saw off a block of wood and pound a few nails into it pretty quickly so this one is great for working out new ideas. The next board is the sample and hold amplifier, a JFET circuit with an opamp output buffer. This is on unplated fiberglass board, so you need to use the leads of the component to make the connections, which only works with thru-hole components. I had been using this method for a while, but reworking components is a b1tch because they are usually wrapped into a knot then soldered, not so good for tweaking components or changing topologies. The next card is a the trigger section of the sample an hold, it is on a piece of ash veneer and done as an eyelet board, as were nearly all Fender amps, but they use a waxed fiberboard as the substrate, not veneer. I got the brass eyelets and the tool to round them off online at a place that does bookbinding for pretty cheap ~$10 so it was worth a shot. This is a pretty nice method to use, you can rework components and add new holes pretty easily. The blue board in the foreground is the start of the joystick interface, it's an old piece of vector pad-per hole board. I think I am going to put the amplitude sampler on that too, there should be enough room left over after the joystick circuit is done. At first I put all of the boards parallel to the floor of the rack, but I ran out of space pretty quickly so I made a makeshift card frame out of cardboard, which is working surprisingly well. The profile looks like this: |\____/| and the corners of the cards are inserted into slits in the angled part of the base. It is holding pretty well for now, but I am pretty sure that it will stretch out and I'll have to do something more permanent to hold the cards....
This is the second rack, the mod processing modules:
It's not as far along as the first rack, but when it's done it will have a phaser, a voltage controlled panner, two filters set up for stereo, and an asymmetrical waveshaper. It's got some old modules from an ARP Omni 2 that was busted, the top board is the phaser/flanger module and the board in the center bottom is the 4075 4-pole filter from it. The bottom blue board is the support interface , the audio and cv in sections and the output buffer. The small board on the right is a chopped up lowpass filter from of an EFM mad mouse board that never worked for me, so I hacked it up into parts that actually worked. /flame on: this is one of the crappiest boards I have ever had the misfortune of working on. You have exactly 0 tries to rework anything on it and the unplated holes all peel up when you put a hot iron on them. I worked on this board for years until I got so p1ssed I literally cut it to pieces - /flame off. Other than that, it is a decent sounding filter but the sharpness of the cutoff is a bit weak. I am planning on replacing both filters with CEM 3372 chips that I pulled out of an old Chroma Polaris, which have a VCA and a VCF section. The waveshaper hasn't been started yet, but I am planning on doing either a diode breakpoint or a log shaper with adjustable positive and negative slopes. Asymmetric and/or nonlinear distortion is one of the areas I'm looking into for a new type of sound, because it hasn't really been done before circuit-wise and it has the potential to generate a new type of harmonic distortion based on the asymmetry of the transfer function. I have been doing it in DSP land (Adobe Audition and VST plugs) for a while and you can always get a different, non-traditional sound, but AFAIK it hasn't been implemented in an analog circuit. I would _really_ love to do a spline based arbitrary transfer function module but splines take a lot of analog circuitry to implement because of the multiplication, they are a lot easier to do with a DSP.Anyway, my modular racks as they stand in May 2008. Hopefully I can get them finished off this year
No comments:
Post a Comment