Installing a PC mechanism in a 1581 doisk drive  
                  latest updates or corrections 2-21-2014

Someone brought me a 1581 to repair and they didn't know anything of the 
history of it, so I was starting from scratch. The first thing I noticed 
was that the mechanism was not original Commodore (usually a Chinon F-354E 
or Matsushita JU-363-292) but a Teac FD235F 720K PC type. Disks were hard 
to insert/eject because that drive sits lower in the housing than a 
Commodore mechanism. Although the drive could read correctly on the first 
try, subsequent reads after the disk was changed brought up the first disk 
directory. Obviously, disk change was not being detected. I swapped in an 
old Amiga mechanism to make sure the 1581 board was OK, and it worked fine. 
I noticed that the Teac drive did have the drive select jumper on D0 as it 
should be, so the 1581 recognized the mechanism correctly. Someone had 
attempted to install a replacement mechanism but couldn't get it to work. 
     The misaligned height of the drive in the case was easy to remedy. I 
just installed a few shims to raise it up so it matched the opening. The 
eject lever, fortunately, was a perfect match. The disk change problem was 
more of a challenge. I searched some old newsgroup files I saved on the 
subject of modding a 1581 with a PC drive and found one that described 
swapping some of the pins on the drive cable to make one work in a 1581. 
However, I couldn't get it to work. It couldn't even read the disks with 
those mods installed. Back to square one. 
     In the Teac drive, the disk-in and write protect switches sit on the 
left side, just inside the drive door. When a disk is inserted, the disk-in 
switch opens and the voltage on it goes to +5VDC. I traced that line to an 
IC and the logic level changed there, but didn't anywhere else on the board. 
I concluded that either the chip is defective (no output pulse for disk 
changes) or something else in the mechanism is incompatible with the 1581 
board. Since everything else seemed to work, I decided to try a modification 
of my own to make it see disk changes. 
     Experimentally, grounding pin 2 of the 34 pin (dual in-line header that 
connects the mechanism to the 1581 board) after disk changes would result in 
a re-read of the BAM when the drive was next accessed. However, keeping that 
pin grounded was not the answer... then it couldn't read disks at all. 
     I pulled the DC (disk change) jumper on the Teac mechanism board to 
isolate their circuitry from pin 2 of the 34 pin connector. Then I built up 
a little pulse generator circuit with a single NPN transistor, one capacitor, 
one resistor and one diode. This circuit is wired between the disk-in switch 
(used as the "trigger") and pin 2 of the 34 pin header on the drive rear. 
     When a disk is inserted, the switch line goes high. That positive going 
voltage is coupled through a 1uF capacitor to the base of a transistor and 
turns it on for a fraction of a second. After the capacitor charges, the 
voltage on the transistor base goes low again and the transistor turns off. 
A diode from base to ground is used to discharge the capacitor when a disk 
is removed and the line goes low again. The transistor collector lead is 
wired to pin 2 of the 34 pin header and a 1K ohm resistor is tied from that 
point to +5VDC to pull the line high when the transistor is not conducting. 
This circuit supplies a low-going pulse every time a disk is changed and 
satisfies the requirements of the 1581 board. The circuit modification 
looks like this:

                             | +5VDC
                             |
                             \
                             /  1K ohm resistor
                             \
                             /
                             |
                             +------> to pin 2 of 34 pin header
                             |
        1uF 16V capacitor   /
              -    +      |/ C
       >--------||---+----|     NPN transistor
  from disk-in       |  B |\ E
  switch             |      \
                  K__|__     |
           diode    /|\      |
                  A  |       |
                     |_______|
                             |
                             | ground
                              
     One last problem had nothing to do with the modifications. I 
discovered that the new drive would not reliably read all disks although
the head alignment checked OK. As it turned out, the spindle speed was 
off by quite a bit. Since the Teac mechanism didn't have an adjustment 
for spindle speed, I looked for a circuit problem on the motor board and
found an open capacitor. My customer now has a working 1581!

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