avatar_Cwiiis

Trouble getting floppy disk drive working on 464+

Started by Cwiiis, 22:05, 16 June 22

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

Cwiiis

I've been upgrading a 464+ to 6128+ spec for a friend, pretty much following the same steps as in this video:
https://www.youtube.com/watch?v=uKaETSxEVMo

Unfortunately, I've had trouble getting the floppy drive cable working with a Gotek. When I try to access the drive, either via selecting it with |a or by running cat, I get a read fail. On the Gotek display, I can see it skips through all the tracks of the disk and then seeks between track 0 and 1 a bit - which kind of feels like it's seeking and failing to find data perhaps, but I've really no idea. On my working 6128+, I can see that the Gotek only seeks track zero, as would be expected doing a directory listing.

As far as I know, I've done everything right - I have a couple of FDC ICs and they both have the exact same behaviour (though one gets noticeably hotter than the other). I've resoldered the floppy cable very, very carefully (which took ages - it's a bummer that doesn't seem to be the problem...) Are there any common issues that would cause this sort of behaviour?

I've attached a couple of pictures of the work - for what it's worth, everything else is working fine - the RAM upgrade went fine and the machine is otherwise operating normally. I have access to a multimeter, scope and logic probe, but am not so well versed as to know what to do to debug this particular situation with them... Any help much appreciated!

d_kef

I had the same problem when I upgraded my 464+ and it turned out to be the SED9420. Luckily I had a second one to replace it with.

d_kef


Cwiiis

Quote from: d_kef on 00:06, 17 June 22I had the same problem when I upgraded my 464+ and it turned out to be the SED9420. Luckily I had a second one to replace it with.

d_kef

Just swapped out the 9420 for a spare, but no change - so unless I'm unlucky and have two bad chips, I guess it's unlikely to be that... I'll triple check all the soldering there, but pretty sure I did a good job on the board...

GUNHED

Take a look at the cable, this can be the source of problems. Maybe shifted.
Good luck anyway!  :) :) :)
http://futureos.de --> Get the revolutionary FutureOS (Update: 2023.11.30)
http://futureos.cpc-live.com/files/LambdaSpeak_RSX_by_TFM.zip --> Get the RSX-ROM for LambdaSpeak :-) (Updated: 2021.12.26)

Cwiiis

Ok, so I tested everything on the cable out, everything went on the board where I expected to go but still I had problems... I've swapped out every IC with spares except for the LS logic (how likely is that to be the culprit really?) and no dice. So I figured, maybe it was the cable, bit the bullet and bought a centronics port so I could test it out with my known-working floppy cable that I use on my 6128+.

The exact same thing is happening :(

I'm really stuck, does anyone have any guidance for debugging this?

I've been testing continuity on the ICs in case it's a bad socket, but everything looks good so far - I'll start cross-referencing the circuit diagram and making sure there aren't any obvious broken traces, but the board is in really good condition, it seems unlikely... I'm shooting in the dark at this point, I have all the tools to debug this, but not the knowledge to do it...

Cwiiis

I happen to have a spare 464+ with a dodgy ASIC, so I used that to verify every connection on the data separator (which I was most suspect of thanks to the tip from @d_kef ) - and wouldn't you know it, the bodge wire that goes from pin 10 of IC15 was connected to the wrong pin on C21. Fixed that and it's all good!

No way I'm resoldering that drive cable though, so now I need to modify the case for the centronics port 😛 So glad to have this worked out, and now I have a silly amount of spare stock for disk-related ICs and sockets!

Animalgril987

Maybe you could use your spare chips to make DDIs for owners of 464s?

Bryce

Hi CWiis,
      there's normally a black wire connecting the ground near the power socket to a point on the floppy circuitry. This was probably done by Amstrad to resolve a ground issue on the floppy circuit. I don't see it in your pictures, is it under the board? If it's missing it might be the cause for your mod not working.

Bryce.

Cwiiis

Quote from: Bryce on 10:54, 10 August 22Hi CWiis,
      there's normally a black wire connecting the ground near the power socket to a point on the floppy circuitry. This was probably done by Amstrad to resolve a ground issue on the floppy circuit. I don't see it in your pictures, is it under the board? If it's missing it might be the cause for your mod not working.

Bryce.
This was the problem indeed, it was connected on the underside, but to the wrong pin :picard:  - I learnt a lot and improved my desoldering skills debugging it at least!

GUNHED

http://futureos.de --> Get the revolutionary FutureOS (Update: 2023.11.30)
http://futureos.cpc-live.com/files/LambdaSpeak_RSX_by_TFM.zip --> Get the RSX-ROM for LambdaSpeak :-) (Updated: 2021.12.26)

Cwiiis

In case the last couple of posts weren't clear, this is now working perfectly 🙂

Powered by SMFPacks Menu Editor Mod