After swapping the heads around, Head 1 still refuses to write (tested with IMD), even though it's being controlled by Head 0 circuitry. To swap heads around it was enough to just switch CT and E contacts places in the header, because R/W coils of both heads are essentially connected in parallel. If control board is fine, then it must be the Head 1 that is faulty. I think that the magnetic permeability of the ferrite core may have deteriorated somehow, an internal crack possibly?

Increasing the write current from 4 mA to 6 mA improved the overall picture at least somewhat, but for some reason Head 1 refused to write to the last sectors of the floppy disk (higher coercivity at the innermost sectors?). I hacked up a degaussing wand on a 6 mm diameter nail, which can produce a magnetic field strength of about 375 Gauss (on a 9V AC adapter) and tried to demagnetize the head assembly with it, but that made no difference for the head's performance.

Conclusion - Head 1 can't produce strong enough magnetic field to write the disk.

These are two separate 2D floppy disks with an image of DOS 3.30 Disk 1, written and then read back on the faulty FDD using an IMD.

Have you measured in circuit resistances at both inner and outer positions of the head? I am wondering if a pull or change in position on the cable is opening a solder crack or parting broken ends somewhere.

Edit: another thought is that if the motor for the head position or the sled assembly is stiff or binding, maybe motor current goes really high, starving another part of the circuit.

This was one of the first things I checked. I tried to find a break in the cable, by connecting the upper head's R/W coils to a multimeter and then moving the cable around, but nothing came of it, cable turned out to be just fine.

Doubt it. If that were the cause of the problem, it would also affect Head 0. Though initially the sled portion of the head assembly was stiff as the grease on the sled rod became too tacky, but I didn't leave it in that condition for long.

I assume you've given it a good cleaning,

How about pressure.. perhaps the spring has weakened. Even this technically applies to both heads, there can be noticible differences beteen fixed(lower) and floating(upper - you can try tacking a small weight to the upper head to slightly increase the pressure.

Believe me, I've cleaned that head so many times with a cotton swab dipped in isopropyl alcohol that it should be polished like a mirror by now if it wasn't before. The only strange thing I noticed was that these heads don't have an air gap that could be felt by the tip of the needle between the R/W ferrite strip, which usually gets dirty and there's none of that here.

I tested this pressure idea by gluing a 50 g weight to the top of the head with double-sided tape. The weight was well balanced and didn't wobble, but alas, it didn't help the head with writing.

As far as I understand, the design of FDD heads, at least the read and write part, is basically the same as that of tape recorders, that is, a coil wound on a C-shaped magnetic core. If even a little bit of this magnetic core is missing or cracked, the magnetic field strength in the air gap of the head will decrease. I will have to completely dismantle the head holder and check if there is any damage done to the ferrite core under the top protective metal cover, because it's possible to access the top of the head through the gaps of that cover. I don't know if the ferrite core is even exposed there, should I even bother dismantling the whole thing just to find out?

I remember some commodore people hunting for floppy heads to restore 1541s without success, I might even remember someone trying to DIY a replacement. Shouldnt be that difficult considering Jugi Tandon made first batches using stuff from hardware and house appliance stores 😀

Oral History of Jugi Tandon https://www.youtube.com/watch?v=hf7RK8xYwG8 <- this dude invented double sided floppy mechanism everybody adopted, before him IBM had some complicated contraption trying to squeeze disk from both sides instead of just one moving head.

Interesting - it does seem llike you've had an actual head failure - something that I who has used *many* floppy drives, new and very old (creator of ImageDisk) have not encountered. Guess you really do see something new every day! (or at least every few years 😀 )

rasz_pl wrote on 2026-06-22, 07:03:

I remember some commodore people hunting for floppy heads to restore 1541s without success, I might even remember someone trying to DIY a replacement. Shouldnt be that difficult considering Jugi Tandon made first batches using stuff from hardware and house appliance stores 😀

Oral History of Jugi Tandon https://www.youtube.com/watch?v=hf7RK8xYwG8 <- this dude invented double sided floppy mechanism everybody adopted, before him IBM had some complicated contraption trying to squeeze disk from both sides instead of just one moving head.

Thanks for pointing out the C64 disk drive, I did some digging and found out that 1541's made by Mitsumi were notorious for open RW coils because the ferrite core encapsulating epoxy resin tends to absorb moisture. Some repair statistics are shown on this web page, together with internals of the 1541 disk drive head. It's a straddle erase type head, but I think the tunnel erase head should be built similarly, at least the RW part. This guy here managed to rewind the RW coil for one of these heads, just wow.

DaveDDS wrote on 2026-06-22, 07:22:

Interesting - it does seem llike you've had an actual head failure - something that I who has used *many* floppy drives, new and very old (creator of ImageDisk) have not encountered. Guess you really do see something new every day! (or at least every few years 😀 )

You are insanely lucky in this regard, because this is the second drive I encounter with a faulty head in the past 8 years, the first one had an open RW coil in one of its heads. Thank you for creating and releasing this superb program, it makes life so much easier for us retro enthusiasts 🫡

As soon as I can dedicate at least 3 hours of uninterrupted work one afternoon, I'll take this head apart.