Path: utzoo!mnetor!uunet!husc6!mailrus!umix!metavax!fmsrl7!mibte!jbh From: jbh@mibte.UUCP (James Harvey) Newsgroups: comp.sys.cbm Subject: Re: 1541 disk alignment (1541 Physical Exam) Message-ID: <2511@mibte.UUCP> Date: 11 Feb 88 15:02:56 GMT References: Organization: Michigan Bell Telephone Company Lines: 66 Summary: another way to align the drive In article , prindle@NADC.ARPA (Frank Prindle) writes: > One of the best apparent 1541 alignment programs is called "1541 Physical > Exam". The disk supplied contains several tracks on which some sectors have > been recorded exactly on track, some off center to the inside, and some off > center to the outside. -- description deleted - > > This method of alignment closely approximates a *real* alignment disk which > contains an analog sinusoidal waveform recorded in an eliptical track. I've aligned many 1541's. The technique I use is to monitor the AC voltage at the output of the second analog head amplifyer, and just tune the stepper motor for maximum signal with a commercial disk in the drive. I have a Hewlett Packard 400D AC Voltmeter that is perfect for the job. The going rate for a used 400D is about $15 around here at the Ham Radio swaps. Any AC voltmeter that will display a half volt full scale should work, the frequency is around a megahertz though. One thing the commercial alignment disks (Cardinal et al.) don't tell you is that it is possible to align the 1541 exactly a whole or even exactly a half track off. This is because the stepper actually moves two steps per track and the software in the drive will hunt for the best spot to read the disk. It will quite happily settle on a stepper phase that is not correct. Strange things will happen when aligned this way. The drive may refuse to read at first, then settle down and read perfectly. A disk formatted in the drive may be missing track 1 (this is a dead giveaway). Protected software laughs at you. You can read the stepper phase by doing a M-R from $1C00 in the drive (I think it's 1C00, I'm doing this from memory here). Mask off the lowest two bits by ANDing the byte with 3. These two bits control the stepper motor and should be either 0 or two depending if the track number is odd or even (again, from memory, I may have this backwords). I have a simple program that positions the head on any track and does continuous reads so I can tune the stepper up. It also can step the head back and forth a half step to check for sticky rails and can print the stepper phase. By the way, spindle speed is critical. It must be within a couple of RPM of 300 RPM. I was able to make a dandy strobe disk by doing a pie chart of 48 spreadsheet cells all of which had the same number. I used Lotus123 but any pie chart should work. I then imported the Lotus pie chart into a graphics program and black filled every other piece of the pie. The result is a 5 1/4 disk (I put it into an old floppy jacket) with 24 black and white segments. This will stand still when viewed under 60 HZ flourescent illumination if turning at exactly 300 rpm. I just put my strobe disk in the drive (sideways so as not to get xerox ink on the head) and adjust the speed pot. I find this method to be much more accurate than the software speed checkers. I've not found two of those that give the same results yet. -- Jim Harvey | "Ask not for whom the bell Michigan Bell Telephone | tolls and you will only pay 29777 Telegraph | Station-to-Station rates." Southfield, Mich. 48034 | ihnp4!mibte!jbh or try ulysses!gamma!mibte!jbh