Page 1 of 1

Index position in 3.5" drives

Posted: Fri Nov 11, 2011 4:20 pm
by TeaRex

this is not directly KryoFlux related but maybe somebody here knows anyways; I've got a 3.5" drive which I assume to have an index position adjustment issue - the glue is gone from the spindle screw and the screw seems to have been turned:

1.) In a 3.5" drive - am I correct to assume that the index signal is supposed to fire at the moment in which the driving pin (the off-center pin that spins the platter) is exactly in between the spindle's center and the center of the magnetic heads? That's what looking at other drives seems to suggest, but of course you can't be completely certain just from looks.

2.) How exact is the index position supposed to be (in terms of degrees of angle, or time, or number of bit cells)? Especially I'm wondering if careful mechanical turning of the screw to optically re-align things, and subsequent re-glueing, can make the drive usable again, or do I need some kind of special alignment disk (which maybe has a special bit pattern at the exact index position)?

3.) If an alignment disk is needed: Can I use KryoFlux to write such an index alignment disk on a known good drive and/or to then do the alignment on the problematic drive?


Re: Index position in 3.5" drives

Posted: Sat Dec 31, 2011 9:01 pm
by philpem
Bit late but I think I might have something to add...

What I would do in this situation (if replacing the drive wasn't an option -- 3.5in drives are basically two a penny these days) is use a known good drive to DC-erase an entire track, then on the second pass write a burst of data when the index signal is detected. Take this disc and write protect it. Put it in the drive you need to align.

Now hook up an oscilloscope to monitor the INDEX and READ DATA lines. You'll need a drive exerciser or similar to do the seeks (unless the Kryoflux can be used as a disc drive tester -- activating the drive selects and seeking to user-specified tracks?). Your goal is to get the start of the READ DATA burst to coincide with the INDEX line going low.

Ideally you'd use an alignment disc, but this is a mostly-acceptable hack.

Best estimate, this'll take an hour or two *if you've done it before*. As a first time run, probably an entire evening. The point I'm trying to make here is.. you can do it, but it probably isn't worth your time.

Now if this were an Amstrad 3-inch drive (which are actually fairly rare in working order, especially the double sided ones) I'd say "do what you must". I've gone as far as building entirely new controller boards for these using surplus head-amplifier chips (MC3470 rings a bell) and Xilinx CPLDs... lots of fun!