Head alignment for legacy Sony FDD

Tony Duell ard at p850ug1.demon.co.uk
Sun Jan 20 13:45:10 CST 2008

> Hi,
> I've been trying to create a working Sony MPF50W (as found in the HP1650B

Isd this one of the half-height units using many of the same parts as an 
Apple 800K drive? In other words similar to the drive used in the HP9114B?

> logic analyser) from 2 damaged units.  One unit had a fried board and the

What was fried? I think, given the choice, I'd have prefered to keep the 
mechanics untouched (thus avoiding alignement problems) and chan ged 
components on the PCB.

> other unit had physically damaged heads.  I moved the heads and track 0

In theory if you change the heads, you need to do a radial alignment -- 
what most people mean by 'aligning a floppy drive). That said, I've 
replaced the heads in a full-height Sony (as used in the 9114A), put the 
alignment disk in and found it _well_ within spec. Maybe I was lucky, or 
maybe, as I suuspect knowing Sony, the units are made to a high degree of 
precision anyway.

> sensor from the fried board to the other board and I need to do 2 things: a)
> confirm that the "new" heads are actually working and b) realign the track 0

One thing worth rememebring : Almost all the adjustments are 
'interchenagablity adjustments' to make the disks used in that drive 
compatible with others. If they're off, the drive can still format, read 
and write its own disks. So take a blank disk and try to format it. If it 
formats and verifies OK, or if it fails on the innermost few cylinders 
only (maybe the track 0 sensor is so far off that the innermost cylinders 
are 'off the end of the disk), then the heads are working. If it doens't, 
you've got other faults to find.

> sensor.  I'm reasonably sure that the heads are okay, they show continuity
> on some of the pins, but I'm not 100% certain that is what I should be

If this is one of the drives covered in 'my' diagrams, I'll have shown 
the head connections there. A head winding will test as a very low 
resistance (dead short) on a meter. Just about the only thing that's not 
connected to anything lese is the screen (shield)

The full-height Sony drives have the head switching diodes on the 
flexiprint to the heads, which might confuse your tests. The half-height 
ones have sepearate head tails from each head ans the swtiching diodes on 
the main PCB.

> expecting.  I can't find any obvious sources of information on where to
> probe to see useful signals from the heads or find any procedure to align

If this is the drive I think it is, the analogue circuity is in one big 
ASIC. There's an LC filter circuit hung off that. looking at 2 
'equivalent' pins of that with a differential-input 'scope will show the 
head signal.

The normal way to alignt the track 0 sensor involves the euse of a drive 
exerchiser. What you do is move the sensor right out (to prevent it 
stopping theh eads from stepping out when they get to what thr drive 
things is track 0,), then, with 'standard' formatted disk in the drive, 
step out until the signal is lost (that's one cylinder past track 0), 
step in once, and move the sensor in until it just triggers. Then make 
sure that the track0 signal is asserted on track 0 and not on track 1.


More information about the cctalk mailing list