Microdrive Unit Design

Nagging hardware related question? Post here!
User avatar
Cristian
Aurora
Posts: 960
Joined: Mon Feb 16, 2015 1:40 pm
Location: Veneto

Re: Microdrive Unit Design

Post by Cristian »

RWAP wrote: I have a feeling that these were used in the early QLs but then maybe for reliability, they slowed down the motor and used the keyed system to identify the newer version?
So is there an ultimate answer to this question? Indeed today I realized that some of my mdv units have straight edges! They seem to read old tapes better than other keyed edge units :o
Are there some serials or codes to identify the various issues?


User avatar
Dave
SandySuperQDave
Posts: 2765
Joined: Sat Jan 22, 2011 6:52 am
Location: Austin, TX
Contact:

Re: Microdrive Unit Design

Post by Dave »

Ralf R. wrote:
RWAP wrote:And....making the motor a bit faster (I think there was a potentiometer on the PCB) gives more sectors. So I think, one of your mdv is running a bit faster than the other. I was able to get 235 sectors, but had not used this speed.
This is the opposite of my experience.

During format, he tape loops and sectors are written, counting down from 255 to 0*. At some point, the starting point passes the head and sectors 255, 254, 253... are overwritten by sectors 4, 3, 2, 1, 0. The faster the motor, the more sectors are overwritten. This sets the number of sectors laid out on the cartridge. Then, during a verify process, the highest number sector that hasn't been overwritten is determined, and any detectably bad sectors are omitted.

You would think that arranging it so all 255 sectors are on the microdrive with no overlap, but in my tests I found that the head gap and tape speed sets a limit on data density that is already quite close to the theoretical limit for the media. For basic data integrity, the maximum sensible sector count is about 230-235 sectors. I'm personally much happier with a cartridge that formats to 220-225 sectors.

*practically, not logically


User avatar
tofro
Font of All Knowledge
Posts: 2685
Joined: Sun Feb 13, 2011 10:53 pm
Location: SW Germany

Re: Microdrive Unit Design

Post by tofro »

Actually, the tape would need to run slower to get a higher density (less amount of tape runs by the head in the same time interval) and, thus more sectors. I'm not sure how much the length of tape varied between different cartridges, but, as Dave said, were already pretty much at the theoretical limit. There's also always a length of tape that cannot be used (the spot where the loop is glued together). Depending on where exactly this is and how long it is, we will also lose between 0 (unlikely) and 2 sectors.

Tobias


ʎɐqǝ ɯoɹɟ ǝq oʇ ƃuᴉoƃ ʇou sᴉ pɹɐoqʎǝʞ ʇxǝu ʎɯ 'ɹɐǝp ɥO
User avatar
Dave
SandySuperQDave
Posts: 2765
Joined: Sat Jan 22, 2011 6:52 am
Location: Austin, TX
Contact:

Re: Microdrive Unit Design

Post by Dave »

Tobias,

It's near the lower limit. It's not the charge embedded into the tape that matters, but the crispness of the change of state. On playback, the sharpness of that transition is what is read. On microdrives, the transition is nearer the lower boundary.

I hope this is clearer.


User avatar
Dave
SandySuperQDave
Posts: 2765
Joined: Sat Jan 22, 2011 6:52 am
Location: Austin, TX
Contact:

Re: Microdrive Unit Design

Post by Dave »

Hehe... That didn't really help. Let me try again another way.

The tape signal during read is AC coupled, as read by the head. Only state changes are readable. Steady state signals quickly revert to zero. The more tape a transition occurs over, the more ambiguous the tipping point is and the more vulnerable reading is to things like temperature changes altering the switching points of components. These timing changes are quite critical in microdrives.

Faster motors give fewer sectors for the looping/overlap reason above, and slower microdrive motors push the reading process into quite marginal territory with regards to AC coupling as even at Sinclair spec settings the signals were quite near the boundary of reliable reads. So we have a peak speed where sectors can be written and read reliably which is different from the maximum number of sectors you can get on a cartridge.

Add in that the guard time (the gap between the sectors) is actually quite small, and speed differences between drive units could theoretically allow a microdrive cartridge to be formatted on one unit with say 215 sectors, but then written to on another unit with a faster motor speed such that a sector could just start to impinge on the next sector, overwriting it. It's such a close thing that with the erase head being in advance of the read/write head, when this happened the erase could impinge on the first few bytes of the subsequent sector, but data write would stop before that point reached the head!

Microdrives worked as much by luck/experimentation as anything else.


User avatar
Dave
SandySuperQDave
Posts: 2765
Joined: Sat Jan 22, 2011 6:52 am
Location: Austin, TX
Contact:

Re: Microdrive Unit Design

Post by Dave »

I designed a microdrive replacement PCB that resolved/ameliorated some of these issues on the analogue side, but it didn't provide much benefit - though it did make a lot of unreadable cartridges recoverable. I'm sure by just mentioning it, Tetroid will release a much improved version within a few weeks ;)


Martin_Head
Aurora
Posts: 847
Joined: Tue Dec 17, 2013 1:17 pm

Re: Microdrive Unit Design

Post by Martin_Head »

Think of the old reel to reel audio tape recorders. The faster the speed of the tape, the better the quality.

If you are doing digital signals, you need the 'quality'

Also I could see a potential problem in slowing the tape speed to get 256 sectors on the tape. You are libel to end up with a gap between sectors 0 and 255 which could contain sectors, or partial sectors left over from the previous format. Which could confuse the mdv driver no end.


User avatar
Outsoft
Super Gold Card
Posts: 695
Joined: Sat Apr 19, 2014 1:30 pm
Location: Italy
Contact:

Re: Microdrive Unit Design

Post by Outsoft »

RWAP wrote:I had a bit of a sort through my microdrive units yesterday and have just been getting another QL up and running.

I decided to use some of the non-keyed microdrive units (ie. they have straight edges, not the jigsaw pieces which seem to appear on most units).

Having gone through several drives to get two working units, I was pleasantly surprised.

FORMAT mdv1_

produces up to 231/233 sectors

Compare that to the other microdrive units, which tend to return 218/222 sectors as a maximum.

mdv2_ gives slightly lower results, but it does appear that there is a difference between the keyed and non-keyed units.

I wonder which were the earlier ones?
Very interesting thing ;)

I don't know what kind of MDV unit are installed on my 2 QL's (one ITA and one GER) but are all 4 working very well (more than 2000/3000 md dumped) ;)


User avatar
Cristian
Aurora
Posts: 960
Joined: Mon Feb 16, 2015 1:40 pm
Location: Veneto

Re: Microdrive Unit Design

Post by Cristian »

RWAP wrote: (ie. they have straight edges, not the jigsaw pieces which seem to appear on most units).
produces up to 231/233 sectors
Dave wrote: the maximum sensible sector count is about 230-235 sectors.
Yesterday I created a felt pad and put in an old and very used cartridge:
230/246 sectors, then 236/246 :o
On the keyed-edge unit I obtained 218/221.
Also I realized that the straight-edged units appear relatively recent, with almost no wear and less time signs.
Also their PCBs are very different: they look simpler and newer with very thin paths.
Have we maybe found the last (and improved) mdv models?


Derek_Stewart
Font of All Knowledge
Posts: 3932
Joined: Mon Dec 20, 2010 11:40 am
Location: Sunny Runcorn, Cheshire, UK

Re: Microdrive Unit Design

Post by Derek_Stewart »

Hi,

I have carried this one stage further, replaced the Microdrive unit, with a QLSD.

I have never needed a Microdrive since 1990s.


Regards,

Derek
Post Reply