FORMAT ram1_mdv2

[RWAP]

I have an oddity with a JS ROM QL (Issue 6 motherboard).

The QL on its own works like a dream - the microdrive units are in really good condition.

I have tried two Super Gold Cards and both give the same results.

FORMAT ram1_mdv1
works fine to make a full image of the microdrive in mdv1_

FORMAT ram1_mdv2
basically looks as though it is accessing the directory drivers, but never runs the motor on mdv2_ - it eventually reports format failed.

I have fully tested mdv2_ from SuperBASIC, and all the usual DIR, SAVE, LOAD and FORMAT commands work without any problems.

Any suggestions as to what might cause this? Maybe I need to pull the 68008 chip ?

[RWAP]

I have now tried the same QL with a Trump Card and FORMAT ram1_mdv2 works fine.

I have also tried the Super Gold Cards on another JS ROM machine and again the command works as expected.

Just for some reason it does not work with SGC+ this QL .....

There must be something wrong somewhere... but where do I start?

[twellys]

In the QL service manual, Sinclair recommends a couple of mods for the microdrives for issue 6:-
* Adding two resistors on top of IC23 (pins 19 & 21) to -12V line (RAW1 & RAW2)
* Cutting and joining tracks

For starters, you could see if the mods are done?


As the SuperBasic Microdrive commands work, but the SGC+'s doesn't, I would suspect the timing of the signals to the Microdrive ULAs - so check mods first, then 'scope or logic analyser.

Tim
(Newbie to QLs, good-ish at debugging digital designs)

[RWAP]

The mods are in place.

Alas anything beyond that is too much for me

[twellys]

Off topic, if you're pretty sure which chips and their pins you are targetting, a logic probe might be the way forwards.

You put the logic probe on a pin/track/etc. and it tells you if it's logic 0, logic 1, alternating, etc.

I acquired a Thandar TP1 Logic Probe a lot time ago. It only does 5V logic levels, but it's good for legacy computers like the QL.

Just a thought.

Tim

[dilwyn]

I have now tried the same QL with a Trump Card and FORMAT ram1_mdv2 works fine.
I have also tried the Super Gold Cards on another JS ROM machine and again the command works as expected.
Just for some reason it does not work with SGC+ this QL .....
There must be something wrong somewhere... but where do I start?

Might be worth pulling out the 68008 just to try. It cured many SGC problems, usually due to current consumption and/or noise affecting signals. I don't know if mdv1 and mdv2 differ in tolerances to such things.
As an easy test, if you have a ZX Spectrum microdrive, connect that as mdv3_ and see what happens.

As Twellys has mentioned the Service Manuals mention some mods for issue 6 where a track is cut and two wire links added on the solder side (page C4 of the Service Manual) and devotes the whole of Appendix D to microdrive fault finding.

Dilwyn

[RWAP]

I tried it without the 68008 - made no difference. I also swapped the 8302 and 8301 ULAs amd even swapped the 8049 for a Hermes, still no difference.

Unfortunately, I don't have a ZX Microdrive to test the unit with (and the QL is now all boxed up to be shipped to a customer).

I am not too certain that anything in the service manual would help, as neither microdrive unit exhibits any problems when used within a standard QL setting, or even with a Trump Card inserted.

It is just this one command, when either of 2 Super Gold Cards are plugged in.... It even looks as though the SGC is detecting the microdrive (judging by the time taken before it reports format failed). I have even swapped the ROM chip in the Super Gold Card, in case it was that.

Very peculiar indeed - probably one of these little QL mysteries that will remain unsolved.

[vanpeebles]

Sometimes the QL is like an old Austin Allegro that never started right on a damp misty autumn morning.

[twellys]

Or any morning - we had to roll our Allegro down the hill to get it started. Fortunately we lived on the top of the hill at that time.

The simile is quite apt though, as my work is interfacing to a 680x0 at the moment.

[Dave]

If you swap mdv1_ and mdv2_ drive assemblies, does the problem follow the drive, or continue to be at mdv2_?

This will at least isolate if it's logic on the main board or on the microdrive itself that is causing the problem.