Hugo Holden
Veteran Member
I struggled all day to find the fault in a new old stock YD-580 5.25" floppy disk drive.
I thought I would report it in case this problem (or similar) happens to somebody else trying to run a pair of drives. I set up the drives with power supplies and a case and a virtual sector generator (as I had done before) for use with a SOL-20.
When I tried it, drive B wouldn't work. But I quickly found that drive B itself was normal and that drive A, while seemingly working perfectly, would selfishly not allow any other drive to operate while plugged on the same data cable. Yet, on its own, the drive tested perfectly. If I installed this drive assigned to B, it would inhibit the A drive, there would be a few motions of the head in drive A, then nothing, also, occasionally a screen of random data/symbols appeared.
After running around in circles wondering what on earth was going on I decided to make a sacrificial ribbon cable and started cutting wires between the drives having a known good drive as A and the "defective" or selfish one as B, when I cut the wire corresponding to pin 30 (read data), the defective drive stopped the inhibiting effect and drive A functioned normally.
Unfortunately the schematic I could find didn't match the revision of the pcb. But I found that pin 30 connected to the output of an open collector Nand gate 7438 and with the scope there were pulses( like clipped nose) on that pin even when the drive was doing nothing. But no pulses there on a known good drive unit. These pulses were stopping any another drive on the same cable from working, yet the problem did not affect this selfish drive's ability to work on its own. This might be how it slipped out of the factory passing its tests I would guess.
Tracing back with the scope, it turned out that a Hitachi brand 74LS02 gate, which fed one input of the 7438 had lost its ability to sink current on one output, this allowed the 7438 to output data pulses onto pin 30 when it shouldn't be doing that.
It is pretty rare that a nos drive would have a dud IC in it, but it just goes to show it can happen, the IC was probably dud from new and it would have never been noticed if this drive was installed as a single drive as it appeared to work just fine for that.
I thought I would report it in case this problem (or similar) happens to somebody else trying to run a pair of drives. I set up the drives with power supplies and a case and a virtual sector generator (as I had done before) for use with a SOL-20.
When I tried it, drive B wouldn't work. But I quickly found that drive B itself was normal and that drive A, while seemingly working perfectly, would selfishly not allow any other drive to operate while plugged on the same data cable. Yet, on its own, the drive tested perfectly. If I installed this drive assigned to B, it would inhibit the A drive, there would be a few motions of the head in drive A, then nothing, also, occasionally a screen of random data/symbols appeared.
After running around in circles wondering what on earth was going on I decided to make a sacrificial ribbon cable and started cutting wires between the drives having a known good drive as A and the "defective" or selfish one as B, when I cut the wire corresponding to pin 30 (read data), the defective drive stopped the inhibiting effect and drive A functioned normally.
Unfortunately the schematic I could find didn't match the revision of the pcb. But I found that pin 30 connected to the output of an open collector Nand gate 7438 and with the scope there were pulses( like clipped nose) on that pin even when the drive was doing nothing. But no pulses there on a known good drive unit. These pulses were stopping any another drive on the same cable from working, yet the problem did not affect this selfish drive's ability to work on its own. This might be how it slipped out of the factory passing its tests I would guess.
Tracing back with the scope, it turned out that a Hitachi brand 74LS02 gate, which fed one input of the 7438 had lost its ability to sink current on one output, this allowed the 7438 to output data pulses onto pin 30 when it shouldn't be doing that.
It is pretty rare that a nos drive would have a dud IC in it, but it just goes to show it can happen, the IC was probably dud from new and it would have never been noticed if this drive was installed as a single drive as it appeared to work just fine for that.