EMUSIC-L Digest Volume 64, Issue 08 This issue's topics: MIDI circuits (12 messages) MIDI-to-CV converter (4 messages) Your EMUSIC-L Digest moderator is Joe McMahon . You may subscribe to EMUSIC-L by sending mail to listserv@american.edu with the line "SUB EMUSIC-L your name" as the text. The EMUSIC-L archive is a service of SunSite (sunsite.unc.edu) at the University of North Carolina. ------------------------------------------------------------------------ Date: Tue, 3 May 1994 09:55:43 BST From: Philip Jones Subject: MIDI circuits Hi All ;^) I am trying to build a MIDI to CV converter. I have a good knowledge of electronics and am just finishing my 2nd year of an electronic engineering degree here at Imperial College London. However, I don't know the specification of CV or how exactly to handle the asynchronous MIDI signal before it is processed. That is how to buffer the incoming MIDI signal with the processor circuit. If ANYONE has experience in this area ir can point me in the direction of a book or journal containing detailed information I would be most grateful. The converter will initially be designed to ONLY handle note on and off info, with no program change etc, control messages accepted. I will therefore need to filter these out. I have a good grasp of the MIDI spec so that is not required. Thanks for your help Philip Jones (pj02@ee.ic.ac.uk) ------------------------------ Date: Tue, 3 May 1994 14:47:38 -0500 From: Jon Southwood Subject: Re: MIDI circuits Philip, I know that Electronic Musician magazine had a DIY project which involved making a CV to MIDI converter. I'm unable to recall the issue off-hand, but if you have access to back issues I'm sure you'd be able to find it. My recollection is that it showed schematics, parts list, and general directions. I hope this helps. Jon Southwood ****************************************************************************** "I once played poker with a deck of tarot cards. I got a full house and four people died."-Steven Wright ****************************************************************************** On Tue, 3 May 1994, Philip Jones wrote: > Hi All ;^) > > I am trying to build a MIDI to CV converter. I have a good knowledge of > electronics and am just finishing my 2nd year of an electronic engineering > degree here at Imperial College London. > > However, I don't know the specification of CV or how exactly to handle the > asynchronous MIDI signal before it is processed. That is how to buffer the > incoming MIDI signal with the processor circuit. > > If ANYONE has experience in this area ir can point me in the direction of > a book or journal containing detailed information I would be most grateful. > > The converter will initially be designed to ONLY handle note on and off info, > with no program change etc, control messages accepted. I will therefore need > to filter these out. I have a good grasp of the MIDI spec so that is not > required. > > Thanks for your help > > Philip Jones (pj02@ee.ic.ac.uk) > ------------------------------ Date: Tue, 3 May 1994 16:20:31 EST From: "Cook, Dennis R." Subject: Re: MIDI circuits Philip, MIDI circuits are identical to the RS-232 serial circuits used on computers, except that MIDI runs at 36KBaud and only the data receive (RX) and data transmit (TX) and signal ground are used in the MIDI. Since the voltages and impedances are the same, the serial receiver chips and universal asynchronous receiver transmitter (UART) chips found at Radio Shack (Tandy in the UK) make dandy interfaces with a proper clock. The UART handles all of the serial bit synchronization and lets you concentrate on the byte data. If you get a UART with a built-in first-in-first-out (FIFO) memory, or use a separate chip to do that job, the hardware will easily handle most MIDI command strings which are only about 4-bytes. You can get 2-port UART chips which can be wired to provide for a fully buffered MIDI through on your device. Good Luck. -Dennis ------------------------------ Date: Tue, 3 May 1994 23:47:16 +0200 From: Tommy Anderberg Subject: Re: MIDI circuits I doubt this will be of any use to you, but there is a complete description of a MIDI project in the June 1986 issue of Byte (vol. 11, no. 6). Some high school sophomore (!) built a simple interface and wrote a C/assembler sequencer. A complete description of the hardware is given (he used a Zilog Z-80B DART, an 8253 PIT and (I think) a 2 MHz oscillator which, divided by 64, yielded the correct 31.25 kb/s MIDI rate. However, I thought the problem was with the *** CV *** part of the device, not the MIDI part... - Tommy ------------------------------ Date: Wed, 4 May 1994 08:56:24 BST From: Philip Jones Subject: MIDI Circuits Hi All ;^) Thankyou to everyone who has replied to my queries - hopefully I've got some good leads to follow up now. Cheers ------------------------------ Date: Wed, 4 May 1994 11:07:34 EDT From: Keith Maynard - CCC Subject: Re: MIDI circuits Philip, There's an outfit in Oklahoma called PAIA electronics, they sell kits or complete assemplies to do CV to MIDI and MIDI to CV. The prices for MIDI to CV range from US129.00 for single channel kit to US399.00 for a fully assembled 8 channel unit. Apparently schematics are available but the sales person mentioned that there is some embedded software ( I guess it's some sort of EPROM ). Their address is: PAIA Electronics, 3200 Teakwood Lane, Edmond Oklahoma 73013 USA Voice: 405-340-6300 Fax: 405-340-6378 If you like lots of DIY electronics gizmos you should request their catalogue. Bye Keith Maynard > Hi All ;^) > > I am trying to build a MIDI to CV converter. I have a good knowledge of > electronics and am just finishing my 2nd year of an electronic engineering > degree here at Imperial College London. > > However, I don't know the specification of CV or how exactly to handle the > asynchronous MIDI signal before it is processed. That is how to buffer the > incoming MIDI signal with the processor circuit. > > If ANYONE has experience in this area ir can point me in the direction of > a book or journal containing detailed information I would be most grateful. > > The converter will initially be designed to ONLY handle note on and off info, > with no program change etc, control messages accepted. I will therefore need > to filter these out. I have a good grasp of the MIDI spec so that is not > required. > > Thanks for your help > ------------------------------ Date: Wed, 4 May 1994 17:56:08 +0200 From: Tommy Anderberg Subject: Re: MIDI circuits > Philip, > There's an outfit in Oklahoma called PAIA electronics, they sell kits or > complete assemplies to do CV to MIDI and MIDI to CV. The prices for MIDI to CV > range from US129.00 for single channel kit to US399.00 for a fully assembled > 8 channel unit. Apparently schematics are available but the sales person > mentioned that there is some embedded software ( I guess it's some sort > of EPROM ). > Their address is: > PAIA Electronics, > 3200 Teakwood Lane, > Edmond Oklahoma 73013 > USA > Voice: 405-340-6300 > Fax: 405-340-6378 > If you like lots of DIY electronics gizmos you should request their catalogue. > Bye > Keith Maynard This sounds interesting. Do they also sell kits for more conventional MIDI interfaces, things like serial or parallell printer port to MIDI? If they do, what are prices like? - Tommy Anderberg ------------------------------ Date: Wed, 4 May 1994 09:08:33 -0700 From: Doug Wellington Subject: Re: MIDI circuits Tommy Anderberg wrote: (re: PAiA) >This sounds interesting. Do they also sell kits for more conventional >MIDI interfaces, things like serial or parallell printer port to MIDI? >If they do, what are prices like? I built their MIDI interface board. It is available as a one port or two port model. Strictly a UART board, no MPU-401 emulation. Nice kit and I used it for some time before upgrading to the Music Quest MQX-32M. (I wanted SMPTE support, and also Windows drivers.) You have to make sure that your software has drivers for the PAiA board. Cakewalk Pro for DOS has them, that's what I used. The PAiA board is very easy to program for. It's basically just a serial port that works at 31.25k... Basically all I had to do was to change the I/O port declarations in all my communications routines, and I had MIDI access! The PAiA MIDI board is an internal board. It's not for the serial or parallel port. Unfortunately, I'm not aware of any MS-Windows drivers. I seem to remember that getting the kit for the single port board and adding the second port cost a total of about $65... Took me about an hour to solder it together. Worked the first time. (I have been soldering for a few years as a hobby though.) -Doug ------------------------------ Date: Thu, 5 May 1994 08:18:58 +0100 From: "Steven D. Bramson" Subject: Re: MIDI Circuits Take care when comparing pianos with synths. Although a piano is usually tuned on an equally tempered scale with 1 8ve = x 2 frequency 1 semitone = x 2 ^ (1/12) frequency If think you may find that a synth CV may not do this. If you think carefully, 6 8Ves = a 64 Volt range (assuming the lowest note = 1V). Not really a control voltage. OK, maybe the top note is 1 V and six 8ves below is 1/64th of a volt, but that leaves a pretty fine semitone at the bottom of the range (I make it just under 1 mV). So maybe 1 8ve = 1 V giving 6 8Ves = 6 V and there is a linear step in the CV of 1/12 V = 83 mV. Then the internal circuitry translates the linear steps into logarithmic steps to reflect equal tempered tuning. Please note I do not speak from a knowledge of the CV convention, I'm just trying to reason it out. I guess the easy way is to shove 0 to 1V up your synth and have a listen? Steven Bramson sdb@jet.uk ------------------------------ Date: Thu, 5 May 1994 09:12:25 +0100 From: Nick Rothwell Subject: Re: MIDI circuits >Since the voltages and >impedances are the same, Er, no way. Nick Rothwell CASSIEL Contemporary Music/Dance nick@cassiel.com ------------------------------ Date: Thu, 5 May 1994 11:31:36 GMT0BST From: Martin Rootes Subject: Re: MIDI Circuits > Take care when comparing pianos with synths. Although a piano is usually > tuned on an equally tempered scale with > > 1 8ve = x 2 frequency > > 1 semitone = x 2 ^ (1/12) frequency > > If think you may find that a synth CV may not do this. If you think > carefully, 6 8Ves = a 64 Volt range (assuming the lowest note = 1V). Not > really a control voltage. OK, maybe the top note is 1 V and six 8ves below > is 1/64th of a volt, but that leaves a pretty fine semitone at the bottom > of the range (I make it just under 1 mV). > > So maybe 1 8ve = 1 V giving 6 8Ves = 6 V and there is a linear step in the > CV of 1/12 V = 83 mV. Then the internal circuitry translates the linear > steps into logarithmic steps to reflect equal tempered tuning. Nearly all analogue synthesisers on the market use oscilators that have a logarithmic voltage to frequency relationship so that an increase of X volts at the control input will double the frequency (octave rise), in nearly all of these the value of X is one volt (1 volt per octave). To obtain a semitone rise in note would require the voltage to rise by X/12 V. The advantage of this system is that the voltage rise required to produce a particular interval is constant regardless of the starting note, hence you can transpose every note by an octave just by adding 1 volt to the control voltage.. There are some synths around that do require a logarithmic control voltage (the oscilators are linear i.e X volts raises frequency by Y hertz), I think one of the Korgs (the one where you could patch things around) used this system. If you think about it this system is very difficult to use musically as the amount of voltage needed to raise the note a particular interval is dependant on the starting position, for instance assume a 1 volt per 1000 hertz control ratio, the voltages to raise the note one octave are: 250Hz - 0.25V 500Hz - 0.5V 1000Hz - 1V 2000Hz - 2V As you can see it is impossible to easily transpose notes as the amount of volatge that would need to be added is dependant on the note, you could of course use an analogue multiplication circuit, but this is far more difficult to implement than simply adding X volts. Martin. ------------------------------------------------------------------------------ Martin Rootes - Senior Systems Programmer/Analyst, Sheffield Hallam University Email : M.Rootes@shu.ac.uk ------------------------------------------------------------------------------ ------------------------------ Date: Fri, 6 May 1994 00:30:53 LCL From: Dave Amels Subject: MIDI CIRCUITS To: EMUSIC-L@AMERICAN.EDU E > E >I am trying to build a MIDI to CV converter. I have a good knowledge o ......................... E >Thanks for your help E > E >Philip Jones (pj02@ee.ic.ac.uk) E > I believe the easiest way is to use a single chip microcomputer like an Intel 8751. It has the required serial port for MIDI and can easily be interfaced to a D/A for control voltage generation. It's probably best to scavenge up an Intel data book from the school library. Cheers, Dave Amels Voce Inc. --- ß WinQwk 1.30 #0 ß Unregistered Evaluation Copy ------------------------------ Date: Wed, 4 May 1994 09:13:23 +0500 From: Diane Kubarek Subject: Re: CV converter At 7:50 AM 5/4/94 -0400, Joseph Fosco wrote: >>Further to my previous mail, I know that one CV implementaion used >>1v per octave but does anyone how the frequencies are split over the >>octave range. What I mean is that I know that a doubling in frequency >>results in an octave up but what is the frequency increase from C-C#-D-D#- >>E-F-F#-G-G#-A-A#-B-C is it just the octave frequency range divided by 12 or is >>it a non-linear relationship. I need to know this so that the CV I have >>generated from MIDI information can be used to directly control a VCO to >>give the right (correct) tuning. When I say tuning I mean the ordinary >>tuning that a piano say would have. > >When using 1 volt per octave control voltages each half step is 1/12 of a volt. > >Joseph Fosco >Joseph Fosco >fosco@ils.nwu.edu >From what I remember from my Physics class in college, a piano is not tuned in a linear fashion, each interval being 1/12 of an octave. Instead, a piano is tuned with the circle of fifths (C -> G -> D -> A -> E -> B -> F# -> etc. back to C), and if you make each interval in the circle a perfect fifth, you will not end up exactly where you started - you'll be off by a few Hz. However, I remember that I tried an experiment where I programmed a keyboard using the "correct" piano-tuning scheme, and a keyboard that used the 1/12 of an octave scheme, and it was difficult to hear the difference. So, I doubt very highly that anyone will hear the difference if you use control voltages of 1/12 of a volt. A whole lot easier to deal with! Good luck - Diane Kubarek dmk2@cornell.edu ------------------------------ Date: Wed, 4 May 1994 09:15:02 BST From: Philip Jones Subject: CV converter Further to my previous mail, I know that one CV implementaion used 1v per octave but does anyone how the frequencies are split over the octave range. What I mean is that I know that a doubling in frequency results in an octave up but what is the frequency increase from C-C#-D-D#- E-F-F#-G-G#-A-A#-B-C is it just the octave frequency range divided by 12 or is it a non-linear relationship. I need to know this so that the CV I have generated from MIDI information can be used to directly control a VCO to give the right (correct) tuning. When I say tuning I mean the ordinary tuning that a piano say would have. Thanks for your help Phil Jones (pj02@ee.ic.ac.uk) ------------------------------ Date: Wed, 4 May 1994 06:50:37 CDT From: Joseph Fosco Subject: Re: CV converter >Further to my previous mail, I know that one CV implementaion used >1v per octave but does anyone how the frequencies are split over the >octave range. What I mean is that I know that a doubling in frequency >results in an octave up but what is the frequency increase from C-C#-D-D#- >E-F-F#-G-G#-A-A#-B-C is it just the octave frequency range divided by 12 or is >it a non-linear relationship. I need to know this so that the CV I have >generated from MIDI information can be used to directly control a VCO to >give the right (correct) tuning. When I say tuning I mean the ordinary >tuning that a piano say would have. When using 1 volt per octave control voltages each half step is 1/12 of a volt. Joseph Fosco Joseph Fosco fosco@ils.nwu.edu ------------------------------ Date: Wed, 4 May 1994 09:02:15 EST From: "Cook, Dennis R." Subject: Re: CV converter Phill, In the tempered scale, the frequency ratio of one note to the half step below it is 2^(1/12) [i.e., C#/C = F/E = 2^(1/12) ]. Thus the ratio of an octave is 2^(12/12), or 2. This is an accepted (by most) approximation to the pure Pathagorean ratios achievable on non-fretted string instruments, slide whistles, trombones, theramins, and probably CV synthesizers. (Sorry if I left your pet instrument out.) Dennis ------------------------------ End of the EMUSIC-L Digest ******************************