6128 with faulty FDD - unable to find working DSKTEST?

[jjne]

I have a faulty 3" FDD (one of the early EME-150A models) from a 6128 that has the apparently common fault of not being able to read discs (comes up with a "middling address mark" on my +3, have not yet tried on the 6128) but is able to recognise and format them.

A bit of history on this one - I acquired this faulty at a low price, and unfortunately did not check the 12v power supply kludge the previous owner put in - it was reversed.

The 6128 itself seems unaffected (which was a relief!) but the drive was dead. I replaced the belt, and noted that there were two ceramic caps (C205 and 206) that were split/exploded. Replaced these as well and the drive jumped back into life. Have been using the +3 so far for testing in case the 6128 had other issues related to the floppy controller.

I *think* the drive is basically OK from the above but it's possible that the 1066 servo IC linked to the blown caps is also damaged.

Anyway as I say the drive seems to be going through the motions but cannot read discs. So I dud a bit of go ogling and found reference to dsktest, as a common cause of the above symptoms is apparently speed drift over time. 

The only copy I've been able to find of this software though has been a DSK from a French site. But this doesn't appear to work at all through an emulator - launching resets the virtual CPC and loading either BIN through an emulator results in an out of memory message (I will confess I don't have any experience of disk access on the CPC, I've only ever had a 464 up to this point.

I've seen it mention that a CDT of this program exists somewhere but a search has brought up nothing. Can anyone help with this? 

Many thanks for any clues. 

[jjne]

Well, I am officially a numpty - RUN rather than LOAD to run these things lol. 

Will look into converting the BIN file to a cassette format as I don't fancy hot-swapping a gotek to the affected drive! 

[McArti0]

And the smartphone to DIN tape input?

On CPC command |tape.in
And run"

[jjne]

Thanks. Creating a suitable DIN lead shouldn't be an issue, I have the parts for that. However there does not appear to be any way of getting the dsktest program onto a tape-type file.

Am I right in thinking that connecting a Gotek to the rear FDD connector is simply a question of getting a floppy cable with 5.25 connectors and running it in reverse? Obviously subject to getting a separate 5v supply to it. I think that's going to be the only way of getting this to work without hot-swapping which I really don't want to do - alternatively I could connect the FDD via a 26/34 adaptor (I do have a few of these) in reverse direct to an older PC and run (probably better) tests from there.

[Bryce]

Thanks. Creating a suitable DIN lead shouldn't be an issue, I have the parts for that. However there does not appear to be any way of getting the dsktest program onto a tape-type file.

Am I right in thinking that connecting a Gotek to the rear FDD connector is simply a question of getting a floppy cable with 5.25 connectors and running it in reverse? Obviously subject to getting a separate 5v supply to it. I think that's going to be the only way of getting this to work without hot-swapping which I really don't want to do - alternatively I could connect the FDD via a 26/34 adaptor (I do have a few of these) in reverse direct to an older PC and run (probably better) tests from there.

There's a ROM version of DSKTest too, if you happen to own any sort of ROM Board.

Bryce.

[jjne]

Thanks. Creating a suitable DIN lead shouldn't be an issue, I have the parts for that. However there does not appear to be any way of getting the dsktest program onto a tape-type file.

Am I right in thinking that connecting a Gotek to the rear FDD connector is simply a question of getting a floppy cable with 5.25 connectors and running it in reverse? Obviously subject to getting a separate 5v supply to it. I think that's going to be the only way of getting this to work without hot-swapping which I really don't want to do - alternatively I could connect the FDD via a 26/34 adaptor (I do have a few of these) in reverse direct to an older PC and run (probably better) tests from there.

There's a ROM version of DSKTest too, if you happen to own any sort of ROM Board.

Bryce.

Aha!

That's excellent news, I have no issue burning an EPROM and it sorts all the issues out! Thanks!

Edit ah. No. You mean something like a Dandanator...

I have the *pieces* for one but it's not built yet. I guess that's my next job then.

I take it you can't just burn an eprom and install it in place of the boot ROM. In the cpc then?

[Bryce]

Thanks. Creating a suitable DIN lead shouldn't be an issue, I have the parts for that. However there does not appear to be any way of getting the dsktest program onto a tape-type file.

Am I right in thinking that connecting a Gotek to the rear FDD connector is simply a question of getting a floppy cable with 5.25 connectors and running it in reverse? Obviously subject to getting a separate 5v supply to it. I think that's going to be the only way of getting this to work without hot-swapping which I really don't want to do - alternatively I could connect the FDD via a 26/34 adaptor (I do have a few of these) in reverse direct to an older PC and run (probably better) tests from there.

There's a ROM version of DSKTest too, if you happen to own any sort of ROM Board.

Bryce.

Aha!

That's excellent news, I have no issue burning an EPROM and it sorts all the issues out! Thanks!

Edit ah. No. You mean something like a Dandanator...

I have the *pieces* for one but it's not built yet. I guess that's my next job then.

I take it you can't just burn an eprom and install it in place of the boot ROM. In the cpc then?

There are ROMs where that's possible, but the DSKTest ROM isn't compiled for that, it's a standard extension ROM. A simple "old school" ROMBoard, MegaFlash, or similar is all you need to launch it. There's a few here that you might be able to quickly put together on a Veroboard, if you have the right components available: https://www.cpcwiki.eu/index.php/DIY#ROM_.2F_RAM_Expansions
The ACU Board is quick to build with a single socket addressed to any ROM number between 2 to 6 is all you need. 

Bryce.

[jjne]

Hmmm, I don't know, putting something together on strip board to hack into a machine isn't something i want to risk. I'll look into available pre-built and proven gerbers and get one sent over from China.

Think I'll concentrate on getting an external gotek working, which seems a lot more straightforward.

[McArti0]

gotek is the basis

[jjne]

Having googled a bit more on this subject, I'm going to order some boards for a PicoROM - think this could prove to be a useful thing to have in the toolkit generally and is likely to be more reliable than hacking together a circuit out of strip board given my record with such projects 

Many thanks for the replies. 

[jjne]

It turns out that the Speccy +3 I have been using for testing (as I need a known-working system - there may be issues with the CPC FDD controller for all I know) actually has an RPM test built-in to its diagnostics.

And this is showing that the drive is running +55% fast (so about 460RPM). The built-in drive is only +1% fast so I don't think that's a spurious reading.

Need to study the circuit diagram now - not convinced it could have drifted that much naturally so I think there must be another component out. Either the speed control IC is shot or the speed sensor isn't reporting properly.

Unless these older drives aren't fully compatible with the later +3?

Looking again at the diagram this all comes back to the UPC1066 chip... Unfortunately if it is the chip then this becomes a parts unit, as the only place I can find selling them is in the States and with shipping it's nearly £30, for a part that may not fix the drive.

[jjne]

I think I've been overthinking this. Turning the pot down as far as it will go results in a speed 5% fast (so 315RPM). Out of spec, but slow enough that I was able to get it to read discs, save to them and have those files read back in a different drive.

So fundamentally it's working. The caps I replaced seem to be in the speed control circuit (it looks like it's some sort of 555-style oscillator governing the speed). So either the pot or resistor have gone badly out if spec or the new caps I put in (modern cheapo Chinese MLCCs) have parameters different enough to throw the circuit out. I'll have a bit play around to see if I can slow it down a bit further, then I'll have a good drive and, seemingly, a fully working 6128 for £25 😁