MIDI-CV converters

DocteurU at info.polymtl.ca DocteurU at info.polymtl.ca
Wed Feb 7 22:22:44 CET 1996

	Last week I thought about what has been said about my project 
(Building my first synth) and I decided to take more time to build it, and 
modify it a little (well a lot)

	About the part of building the MIDI-CV converter I don't really
like using EPROM's since it's more difficult to upgrade the system if I want
to change it... 

 	I wondered how it was possible to make the thing the more versatile
possible. I thought of building a small computer based on (maybe a 80186 or 
a 68000) with a bunch of ports (MIDI IN OUT THRU + CV outputs inputs and many
8 bit buses) some with interrupt capabilities. Of course RAM and a EPROM for
startup. The advantage of doing this is while it does'nt know a thing about
MIDI, I can load anything into it. (Actually, on power up, it would wait
until the MIDI port receives data and it would put it in his RAM, after that
it would put it's adress counter at the beggining of the first RAM chip). 

	The first part of the program fed in it would be the info on 
configurating the ports (input output, info on interrupts etc). After that,
the machine waits for other inputs, which could come from the midi port, or
any of it's other ports that has been configured as input. (of course, many
of the ports are hardware configured, by configuring, I mean what I expect
to put in a port or receive from a port)

	The other big part of this is doing a program that will generate the
assembly code for it. The interesting part is that it could be done in C and 
be very portable. The other interesting part is that thingy could be used
to control almost anything, from VCO to VCF to a light show! The other good thing
is that it could be used to save sounds from the analog synth built after that.
The really bad part is that everytime it's powered up, you have to load a 
program into it.  

	After the program has been loaded into it, you start your MIDI sequencing
program. And if the assembly code in the thing is well writen, the thing should
understand the MIDI messages. Since everything is assembly writen it would be
fast, the slow part is feeding it with code. Actually, if there's enough memory,
I could load a sample in it and use the sample to control a VCF... (now what 
that would do?) One of the port could a 16 bit entry + an interrupt request bit
that could be attached to an ADC. The thing would be used as a sample processor
even a digital effect processor! If the processor is fast enough and 
since everything would work with interrupts, that would mean real time.

	I am dreaming or is this feasable? Well I do know conception-building of this will be LONG... and writing the assembly code for it will be even longer 
(but it would fit a computer engeneering project nicely (I am thinking of 6
credits now)) .

	Is it really more difficult to use a processor and RAM instead of a
UART(8051) and an EPROM? Maybe slower... Interupt request handeling is
complicated, but not impossible. Also the clocking/timing of the signals/events
involved. Well, all of this could be programmed into a real computer with a card 
that has all the ports I am talking of. But the computer has a lot of things to 
do... it has the sequencer program to worry about... it has Windows 95 to worry 
about!!!!! The program would have to be run parallel to the sequencer program...
memory resident? All sorts of conflicts are to be expected. 

	That thing would do a nice (and big) MIDI-CV converter though.


* This message was sent from The Aircrash Bureau...             	    *	
* Signe Benoit Piette ou si on veut, Le Docteur U                           *     *  DocteurU at info.polymtl.ca                                                 *     *=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*


More information about the Synth-diy mailing list