On Wed, 21 Feb 2018 00:06:05 +0000, Dave Plowman (News) wrote:

In article ,
Johnny B Good wrote:

====snip====


Whilst extremely low inertia motors can be made by making the rotor
'ironless' using a rotor made entirely from just the copper wire formed
into a moving coil shape[1] by the use of an epoxy glue, this doesn't
seem to be a good idea when you want to eliminate flutter from even the
most residual of commutation/cogging effects on the motor's angular
speed.

I have seen one in bits, but can't really remember the motor design -
except it was pancake shaped.

It was certainly ahead of its time in terms of a very stable transport.
You could lock a colour VTR to it and edit pictures exactly where needed
(within the limits of the PAL system) Which was very handy for doing
dance etc numbers to pre-recorded music with multiple passes. It was
some time before the likes of Studer caught up with the more
conventional single capstan design.

Its other trick was you could use any size of spool on either side with
zero problems.

Probably done by threading the tape between a pair of guide pins on a
small 'turntable' either side of the capstans. These rotating turntables
were spring loaded and drove either a potentiometer or some optical
position sensor to generate a control signal to the reel table motors so
as to regulate the back and the take up tensions generated by a constant
current source drive to these motors.

The system used by the GX630DB (and possibly also the GX747) wasn't
quite so sophisticated. They both used direct drive reel table motors
with a constant current drive to provide the necessary tape tensions but
relied on a reel size selector switch to provide a fixed level of torque
which was a compromise for 7 and 10 inch reels which was far from ideal
with 3 and 4 inch reels, especially when loaded with triple play tape[1]!

This, of course is just one example of a servo controlled tape
tensioning system. The other, near identical version you'd likely see
differed only in that the tension sensing arms would simply be a single
guide pin on the end of a spring tensioned swinging arm or a sprung
loaded slider in a guide slot which drove a pot or optical position
sensor.

Ultimately, seen only in high speed open reel reel data storage systems,
you had the tape loaded into pneumatic silos[2] either side of the head/
capstan assembly (capstan in this case not always required) with some
form of sensing to maintain the amount of tape in the loop of tape being
held by air pressure/vacuum in the silos by bi-directional control of the
reel table motors (the supply/takeup reel tables not only provided back
tension but were actively accelerated in the opposite sense to eliminate
any excess back tension from attempts to accelerate the tape pack from
tape tension alone. This truly did isolate the critical section of tape
passing the read/write heads from the effects of inertia in the supply
and take up spool drive system.

[1] When I was creating my own Absolute Azimuth White Noise Calibration
test tapes some thirty odd years ago, I took advantage of the extremely
thin triple play tape which allowed it to be flipped over and played in
the opposite direction so as to play back the stereo in phase from a mono
noise source recording via the back side of the tape. With a 4 track
stereo tape deck, it was most vital to have the left and right head gaps
in perfect alignment if you wished to retain mono mixdown compatibility
with other, equally correctly lined up tape decks. The effect of "Head
Scatter" was of minor importance in this case.

It took only just three or four iterations of such testing to arrive at
an absolutely correct Azimuth setting of the record and playback heads
from which I was then able to create several copies on 7 and 10 inch
reels of LP tape. Since the only suitable triple play tape I possessed
was loaded onto 3 inch reels, this presented a huge tape tensioning
issue. I got round this by winding the tape over a 7 inch reel full of
tape to act as a large diameter 'hub' where I could select the 10 inch
reel tensioning option to approximate a back tension better suited to
such thin tape.

Loading a 3 inch reel's worth of tape over the top of a full 7 inch reel
wasn't a problem since there was still adequate space on the full reel to
add the additional tape without going beyond the limits of the reel. This
also made it easier to deal with the business of re-loading the tape 'the
wrong side up' to complete the auditioning phase of the adjustment
process where the mono mixdown, though muffled, would allow me to
determine whether I'd reached the optimal point of absolute azimuth that
would generate the highest tone of 'hiss' without still having to
temporarily nudge the replay head gently either side of its current
alignment to prove which way, if any, the optimum setting lay.

If a nudge to the left was required to find this point, I'd adjust the
head azimuth by half this required amount before repeating the test
recording process with the tape reloaded the right way around and then
having another listen to the result. I think the initial iterations were
done by simply swapping the reels over and twisting the tape over between
the reels and the first guides. Literal reloading the tape the 'wrong
side out' onto the reels was left for the final stages to verify the
result without imposing a possibly detrimental effect arising out of
using the twisted tape path method.

I think I created one or two dedicated test tapes but I also recall
making good use of the space after the end of several existing recordings
on some of my 7 and 10 inch reels of LP tapes where the recordings had
left a good 5 minutes or more of blank tape at the end where it would be
well out of the way when auditioning the recordings.

[2] These pneumatic silo tensioners were as close to an 'inertialess'
spring tensioning system as you could get since it got rid of the mass of
a typical tensioning arm and its spring, leaving only a few milligrams'
worth of tape mass and micrograms' worth of 'air spring' to upset the
dynamic tape tensioning forces in the critical section of tape being
scanned by the read/write heads. This really was the ultimate in tape
handling taken to the extreme!

Of course, such extremes weren't justified even with 24 track studio
decks but were par for the course in mainframe computer rooms where time
very much equalled money and extremely reliable high speed data tape
processing was the order of the day when, despite even such extremes of
performance, it still represented a bottleneck in the whole data
processing chain.

--
Johnny B Good