Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sun-barr!newstop!exodus!oobleck.Eng.Sun.COM!bender From: bender@oobleck.Eng.Sun.COM (Michael Bender) Newsgroups: sci.electronics Subject: Re: MIDI Interface Keywords: MIDI Message-ID: <2973@exodus.Eng.Sun.COM> Date: 16 Nov 90 21:05:36 GMT References: <3757.658193903@lemsys.UUCP> <10268@uhccux.uhcc.Hawaii.Edu> <16585@mentor.cc.purdue.edu> <1138@tsdiag.ccur.com> Sender: news@exodus.Eng.Sun.COM Organization: Sun Microsystems, Mt. View, Ca. Lines: 30 In article <1138@tsdiag.ccur.com> davet@tsdiag.ccur.com (Dave Tiller N2KAU) writes: ->In article <16585@mentor.cc.purdue.edu> lairdkb@mentor.cc.purdue.edu (Kyler Laird) writes: ->-In article <10268@uhccux.uhcc.Hawaii.Edu> whinery@hale.ifa.hawaii.edu (Alan Whinery) writes: ->->Whoa! Although the serial port and serial device on the mighty Amiga can ->->handle the 31,250 baud necessary for MIDI, most other machines (for ->->instance, IBM & compatibles generally cannot. You will find that the ->->serial handler in MS-DOS is not capable of higher than 19200 baud, not ->- ->-1) Some programs push the serial port to 56K baud, no? -> ->The problem here is not whether the serial port can run _fast_, it's ->whether it can run 31.25 Kbps. The typical 8250 uarts use a guzinta ->circuit (a programmable divider) to set the bit rate. Unfortunately ->the accuracy at high speeds just isn't there. As I recall they use a ->1.8432 MHz xtal to drive the divider and the fixed div-16 prescaler in the ->uart. Thus, 1.8432 MHz / 31250 / 16 = 3.6864. Using 3, we'd get 38.4 Kbps, ->using 4, we'd get 28.8 Kbps. Other values: ->115200 = 1, 19200 = 6, 9600 = 12, 4800 = 24, 2400 = 48, etc.. ->This number get stuck in a register on the serial card. that's true if you use the stock oscillator, but if you replace it with either a 1 or 2 MhZ oscillator, it works fine. I believe some of the early MIDI implementations in keyboards/drum machines used the motorola 6850 clocked at 1 MhZ. mike -- Won't look like rain, Won't look like snow, | DOD #000007 Won't look like fog, That's all we know! | AMA #511250 We just can't tell you anymore, We've never made oobleck before! | MSC #298726