1581 DISK DRIVE DIAGNOSTICS 03-25-2012.


           1581 DISK DRIVE DIAGNOSTICS         
             latest updates and corrections 03-25-2012


To properly diagnose a problem with your 1581, you need to know what 
should happen during normal operation of the drive. A visual inspection 
of the stepper and spindle activity is helpful in troubleshooting. To 
see the drive operate, you will need to remove the top cover from the 
mechanism, and that requires removal of the drive from the case. NOTE: 
It is especially important to note and write down the orientation of 
the four wire drive power cable at the PC board. This cable carries the 
+5 and +12 volt power sources from the PC board to the drive mechanism. 
Note that the red wire goes to pin 1. If that drive cable is accidently 
reversed on the PC board, it will destroy the drive instantly when the 
drive is powered up! Make sure you get it back the way it came off. The 
34 wire flat data cable also needs to be oriented properly. If put on 
backwards, the drive will not work but that mistake shouldn't cause any 
damage. I've done it a few times without a problem during installation 
of alternate mechanisms.
     One difference between the 1581 and earlier drive types such as the 
1541 and 1571 is the functioning of the drive activity light. In the 1581, 
the green activity light comes on during the power-up sequence (drive 
self-test) and during disk accesses, i.e. a read or write in progress, 
but disk errors are displayed by a flashing -power- (red LED) light. It 
will alternately flash dim and bright at about a 2 per second rate if a 
disk error is encountered. This is different than in a 1541 or 1571 where 
the drive activity light functions as an error indicator.

Normal operation:
     When powered up without a disk in the drive, the 1581 power (red) 
and drive activity (green) lights will come on. The stepper motor will 
pull the heads back to track 1 if it is not already there. To see if the 
stepper is working, turn off the drive and push the head mechanism 
forward, then turn it back on. If it is already over track one, it will 
move forward -slightly-, then move back in one quick sequence. If you 
listen closely, you can hear it "burp" as the stepper operates. Then, as 
the drive completes the internal self test and initializition (about 2 
seconds), the green activity light will go out and the stepper will 
"burp" again and remain in the "track zero" area. Note that the spindle 
motor does not turn unless there is a disk in the drive. 
     If there is an unformatted (blank) disk in the drive at power up, 
the spindle will turn and the stepper will pull the heads back to track 
1 (if not already there). As the initial disk read fails, the heads will 
remain at track one and the stepper may burp again. The spindle motor 
will continue to run for about 15 seconds, then stop. Exception: if the 
computer is connected but turned off, the spindle will run continuously 
because one or more serial lines are held at a logic low level by the 
computer. If the computer accesses the drive by trying to read a blank 
disk, the drive error will flash the red (power) LED. If you access and 
read the error channel, it will indicate: 74,DRIVE NOT READY,00,00. The
same thing will happen if the drive heads are clogged. 
     If there is a formatted disk in the drive, but the drive cannot 
read the data for some reason (example: an IBM formatted disk), the 
stepper will move the head back and forth seeking a good track to read. 
Depending on what the drive reads, the heads may remain in the track 1 
area or move to the directory track (40) area, and the spindle will 
continue to turn for about 15 seconds after the green activity light has 
gone out. (See Note 1) If a computer is connected and a disk with 
unreadable (corrupted or incorrect) data is accessed, the 1581 red drive 
power light will flash. If the error channel is read, it will say: 
20, READ ERROR,40,00 because the drive can't read the disk directory at 
track 40.   
     Lastly, if there is a correctly formatted and readable CBM disk in 
the drive at power up (or if connected, a computer reset), the stepper 
pulls the heads back to track 1 (if not already there), then the stepper 
will move the heads forward to track 40 (the directory) and "park" the 
heads over that track, and the green activity light will go out after 
about two seconds. 
     Note that during disk accesses, the green activity light will blink 
or stay on as the drive reads a good disk, and the spindle motor will 
stop about 10 seconds later. Long directories with many entries may take 
several seconds to load. If the same disk directory is accessed again by 
the user, note that the drive spindle doesn't move but the activity light 
turns on for a short time. That's because the BAM and directory 
information for that disk is stored in drive memory until the disk is 
changed. A disk change prompts the drive to reread the BAM from the disk 
the next time the drive is accessed by the computer. Therefore, after a 
disk change, the spindle will turn and the drive will read the disk 
again... even if the same disk is reinserted. There is a tiny switch just
inside the drive door on the left side that closes when a disk is 
inserted. That's what the drive electronics uses to sense when a disk is 
changed. Another switch (or photo sensor in some mechanisms) is used for 
write protection. It checks to see if the hole in the disk is open 
(OK to write) or closed (protected). Note: that is exactly opposite to 
the way a 5.25" disk is write protected. 
     If the 1581 is connected to a C128 and the computer is turned on or 
reset in the 128 mode, the 1581 (like the 1571) will attempt to autoboot 
from the disk. If the disk contains a special USR file with the name 
COPYRIGHT CBM 86, the drive will read and execute that file and 
automatically run the program at the specified load address. 

Note 1: An IBM formatted disk has the same MFM encoding as a Commodore 
formatted disk, but other differences normally make IBM disks unreadable 
in a 1581. However, because of the similarities, with the proper 
software (such as Big Blue Reader), an IBM double density (720K) disk 
can be read or written to in a 1581. Likewise, a 1581 formatted disk 
can be read in an IBM type PC if the proper software is used. Of course 
programs from either system are not cross-compatible, but such things as 
text files can be accessed. Note that newer IBM 1.44M (high density) 
disks are not compatible and will not work in a 1581. Also, some PC
systems will not automatically recognise 720K disks properly. 

Drive problems:
     The 1581 is a very reliable drive compared to other CBM types and 
actual breakdowns are rare. The most common drive failure is actually a 
fault in the external power supply. If the +5 volt regulated source 
"sags" (drops below 5 volts) during drive accesses, the drive will throw 
random errors or will not work at all. That supply problem is usually 
intermittant and it may work when cold but fail after warmup. If you can 
borrow another supply, that's the easiest way to check for a bad one. 
An alternative diagnostic is to monitor the +5 volt line (the red wire 
on the four wire bundle to the drive mechanical deck) inside the drive 
with a voltmeter while the drive is being used. It should remain steady 
at 5 volts DC. If the voltage drops at any time, the external power pack 
is probably failing. The supply must be checked with a load on it (drive 
connected) for voltage checks to be valid. Unloaded, the voltage may
measure a normal level. 
     One drive problem shows up as a continuous green activity light (the 
LED doesn't go out two seconds after power up) and there is no drive 
access. Most of the time, it's the DOS ROM chip U2 (CBM 318045-01) that's 
bad. That chip can be a ROM or an EPROM and normally is the only socketed 
chip in the drive, unless someone has added other sockets during repairs 
or upgrades. Other chips that can produce that symptom are a bad VIA chip 
U5, a CBM 8520, (also used in the Amiga) or U9, a 74LS14 (a standard TTL 
IC). Note that if the drive doesn't complete its initialization sequence 
at powerup or reset (for whatever reason), two of the serial lines (pins 
4 and 5 of the DIN connectors) will remain logic low instead of going high 
as they should. 
     Another thing that concerns some 1581 owners is the 1770 vs 1772 U4
drive controller IC dilemma. Some early drives failed because of a bad 
run of 1770 disk controller ICs from the factory. Due to that one-time
problem, some users routinely changed U4 to the later 1772. 
That "upgrade" requires a jumper change on the drive PC board. If the 
original chip was a WD1770, there may not be a jumper installed at J1 on 
the PC board. If a WD1772 is installed, a jumper wire or 47 ohm resistor 
should be installed at J1. It changes the step rate to 6mS. That jumper
goes from U5 (8520) pin 25 (PHI-2) to pin 19 (TOD). Note that later
drives may already have a WD1772 installed by the factory. If so, the
board has no marking for J1 as it has already been hardwired elsewhere. 
The original fault with a failing 1770 was directory corruption. I had
one 1772 that would not format a disk although everything else worked. 
I used a good 1770 to replace it and the drive now works fine. Note: the 
1571 drive uses the same two controller ICs interchangeably, but has no 
jumper. 
     Another fault that Commodore found and fixed in some early 1581 
drives was a bad connection between IC U10 pin 10 to ground. Some drives 
apparently had a bad connection at that point, and it should be checked 
and resoldered if necessary. That fault caused intermittant "device not 
present" errors. 
     One user reported a U13 (74LS241) failure that I'll include here: 
"Symptoms: reading anything (directory, file, error channel) from the 
drive results in the mechanism spinning, green light on and the system 
locked up. Initialization at boot or reset is always ok. Format and save 
commands seem to be executed, although after saving, the error LED is 
blinking. Correct execution of the format and save commands was 
confirmed with a working 1581."
     Lastly, a fault that is not really a drive problem but that looks 
like one shows up when you connect a 1581 to a C64 (or VIC20) with a 
datassette also connected. When accessed, the drive will lock up the 
serial bus and the green activity LED on the drive will stay on. The 
1581 uses one serial line for burst mode (128 mode only) that the C64 
doesn't support, and that line is also used by the datassette for its 
read function. A 1581 pulls the \SRQ (pin 1) line low when the drive is 
accessed, and that locks the bus. Note that this problem doesn't happen 
with a 1571 or on a C128 (64 or 128 modes) with a 1581 and datassette 
connected. If it is necessary to use a datassette with a 1581 on a C64, 
one workaround would be to cut the computers \SRQ line to the serial 
port (it's never used anyway), or make up a special serial cable with 
that line disconnected. 

Ray Carlsen CET
Carlsen Electronics... a leader in trailing-edge technology