New subject: Yet another idea for recovering the data from Series80 tapes

30 Oct 2005

"Joe R." <rigdonj at cfl.rr.com> wrote:
...
  FWIW I tested about a dozen tapes a few years ago and
about 4 failed
 on the first pass, about that many failed in the next few days and
 only one lasted a week. 
The tapes stick unto themselves, so it is a good idea to unwind them slowly
rather than perform a retention (CTAPE) which is a violent and abusive
treatment for a tape that is ready to fall apart.

Whether you (think you) can afford two data extraction passes (vs
one pass) on a given tape *will* affect the strategy used. In the
one-pass case you have to salvage all you can from both tracks, while
in the second case you can rewind the tape (at the slow reading
speed, of course) and have a go at the second track. The HP-85 tapes
(and their 98xx cousins) are NOT serpentine: to read track 2 you
rewind to BOT and start reading using the second head.

Unfortunately the read/write IC (1820-2418) on the HP-85 can only deal with
one head at a time (only one DIO pin), so to employ the one-pass strategy,
you either have to wire another 1820-2418 strapped to read track two, or you
have do the analog (audio) recording I suggested initially.

...
  They DO use the same tape so they must be compatible.
I don't know about
 the track spacing but I would think they would be the same. 
Bad news. I checked the specs (TechData brochures from hpmuseum.net)
and came up with this table for read/write speeds
        HP 9825 22ips
        HP 9835 22ips
        HP 9845 22ips (didn't find any data, but I assume its 22 ips)
        HP 85   11ips
        HP 9815 10ips

So there goes any hope of cassette data interchange between the Series80
and the rest of the HP line. :-(

Apparently the HP-85 was a cheapo solution vs the 98[234]5 line. The speed
was slower and while this does not reflect on the total capacity of the
drive (around 240Kb), its access times are worse. This was a surprise to
me as I would have thought that HP would want to have some data exchange
capability between the newer low-end HP-85 and its high-end models (after
all they were expected to co-exist in the same environments).

...
  It should be
 easy enough to check with some MagnaSee. I THINK that's the name of the
 liquid that you put on mag tapes/cards to see the magnetic tracks. I have a
 can of it around here somewhere but no idea where it's at at the moment. 
I have been trying to get this thing for ages, and I couldn't. In the end
I found this
        www.sprague-magnetics.com/library/sprgmag.pdf
which works the same way. Initially I thought I could simply spray the stuff
on the tape and optically read the bits, but its too messy.

...
  Actually that sounds like a GOOD idea. Ratty looking
HP 85s are readily
 available. I'd take one of those and take out the tape drive ribbon cable
 and replace it with real wires then add some test points to it and maybe
 cut some sections out of the case to allow access to it's inards and make a
 test platform out of it. My advice is to go for 9915As I have seen them go for next
to nothing on eBay,
and even then remain unsold. Apparently people do not realize that these are
Series 80, or they are discouraged by the lack of screen and (most importantly)
keyboard!

I am also afraid that by creating the series80.org site, I caused an increase
in the prices of Series80 machines, but with the hpmuseum.net slowly making all
this info available anyway, I do not feel so guilty any more :-)

**vp

PS I found another metric screw up:
In the 9835TechData-5953-0982-12pages-Oct78.pdf file in hpmuseum.net
the length of the tape is given as 426.7 METERS, followed by the
more reasonable 140ft (the HP-85B spec sheet gives the length of tape
as 43m or 140ft).

BTW2 On the HP-85 the rewind time (end-to-end) is 29 sec. When I was checking
an early capstan repair and did a CTAPE (which does a fast forward to EOT
and then a fast rewind to BOT), it felt like that thing was taking forever
(probably it took longer because I was watching over it :-)