platzhalter back to start platzhalter
platzhalter about platzhalter standard gear platzhalter analog diy gear platzhalter digital diy gear platzhalter music platzhalter contact platzhalter
 

hardware programmer - midi data translator

With this simple wizards, Rolands MC909 with its nice sliders and knobs can be used to comfortly edit a Korg DW8000 or a Kawai K3m. Both hybrid synths are quite cool but the editing is the old 'select parameter, change value by +/-' thing. They dont accept parameterchanges by controllers like most instruments do nowadays. The only chance to change values is by system-exclusive mididata.

Now I like the MC909 (not only) because of its nice sliders on the envelopes. The MC sends out controller-data when fiddling knobs (acually not on all knobs). So why not translate these controllerchnges into sysex-data for altering parameters in external synths?

First i thought of makin ONE box with a dialswitch to set targetdevice but i hopelessly overestimated the flashram of the tiny2313. With 1 kb i didnt come very far. It even didnt reach for all parameters of one synth, therefore i rejected implementing parameters like e.g. 'aftertouch vfc-lfo intensity'. But i got the oscillator waveforms, both envelopes (without this breakpoint/slope), filter, the delayparameters on dw8k (which are VERY cool) and a lot more.

The other problem is, these old synths dont like it very much to get fired with lots of sysexdata. Thats why sometimes some notes tend to hang when changig values while playing notes. Strangely the K3 seems to accept more data after being on for some time and being warm. But well, nothing is perfect :) Its nevertheless very funny to have a real hardwareprogrammer.

UPDATE on the above: it was all my fault :) as a c++-microcontroller-beginner i produced some not so clean codetrunks. thats fixed now. No note-hanging any more (at least very rare). Although some notes sometimes may lag in timing while parametertweaking, a normal behaviour of old synths being feeded with sysex.

In the making, i also added a switch for each of both wizards to merge two midichannels to one. With the RS7000 I often used adjecent tracks to send data on one midichannel. This way you easily can alter/escalate a pattern just by unmuting tracks without swapping to another complete pattern. Afaik the MC909 does NOT have the ability to set the sending midichannels of its sequencer. Track 7 always sends on channel 7, Track 8 on 8 a.s.o. So i added a switch that maps mididata of channel 7 automatically on channel 8 (for the korg, the Kawai merges 2 on 3).

 

 

 The interiour consists of a tiny2313-microcontroller, a cny17-3-optocoupler, some resistors, a quarz, 3 capacitors and a switch. Powersupply comes from the midicable. The tiny2313U still operates with very low voltage, mustnt be exactly 5V. The white cable in front is the InSystemProgrammingInterface-Cable. That is neede to send your code into the controller while developing.  
 Seen from below. I am not a big fan of etching PCBs. Much faster to do this by hand.  
 There they are.  
Here are JPGs of the papertemplates describing which mc-slider alters which parameter on the synths.  

Template K3m

Template DW8000

platzhalter