The Ampex 1163 Auto Reverse Hi-Fi Stereo Tape Deck
This auto-reverse Ampex audio tape recorder of the mid 1960's is a rather interesting piece of work.

There was no other audio tape deck quite like it before or since, and while many owners were forced in the
end to give up on it (as its drive belts gradually decayed into mush), to my mind it remains a fine example of
inventive mechanical and industrial design. No Hi-Fi tape recorder had a better pedigree, and one (radio)
observatory even recorded the first signals detected from a Pulsar on one of these...
From the start these tape recorders were designed play in both directions, so that one didn't need to
physically change the reels over at the end of the tape. They were also remarkably easy to lace up as well;
just put the reel of tape on the exposed left hand hub and draw the tape across the heads into a slot in the
right hand reel cover, then just push the start key up. The 'pronged' tape-up spool grabs the tape and starts
woking. On coming to the end of the reel the machine automatically goes into reverse, and plays the other
'side' of the tape, if required it could be set to repeat the tape continuously. The pre-recorded tapes were
of good quality and the whole 'system' was simple to use. And as American Hi-Fi enthusiasts preferred
their automatic record changers to the 'Englishman's' single playing 'transcription turntable', this new long
playing tape system with no pops and crackles seemed a pretty good bet. But there again a reasonable
turntable was a lot cheaper than one of these complex machines, and reel to reel tape was never as quick
as disk in accessing specific tracks, so the end result was perhaps fairly obvious.

Taken on their own though, these Ampex machines were a rather esoteric alternative to many a 1960's Hi-
Fi tape deck. Unfortunately though in the UK, because of the exchange rate they were very expensive and
for the 150 price of one of these you could buy the best stereo Ferrograph together with a load of tapes to
record on it. In the end I imagine most Ampex 800 and 1100s ended up being sold off at a large discount
(which indeed mine was).

But in these days, when it takes at least a laser beam and several million transistors in very large
integrated circuits to record a bit of music (and then often not particularly well), this Jurassic tape
recorder is an object lesson in what one does not actually need. This relic from the age of steam (well,
perhaps the 60's mini skirt) records audio at least as well as a MiniDisk, and miles better than MP3 or
most home computers. To do this it employs just one old fashioned mains induction motor, two flywheels,
twenty one transistors (each of which are mounted in little sockets), and an elegant arrangement of belts
and tyres. Infact it only uses five transistors per channel to do the actual audio recording and playback, the
rest are used in the auto-reverse and loudspeaker power amps.

The basic structure of these machines consists of a cast alloy 'picture frame', on which various modular
sub-assemblies and items of electronics are bolted. This picture frame concept was probably a hang over
from those tidier days when Hi-Fi was usually built into furniture, and just the control surface left
accessable. Some recorders were supplied in simple perforated screeened boxes for just this sort of
application, though my own recorder is 'built' into a portable carrying case and is complete with power
amplifiers. This makes a fine self contained 'Hi-Fi' when used with the optional lid mounted speakers.

The transport itself is quite something, and is perhaps the definitive application of two large rubber bands!
Well, not actually rubber bands perhaps, but large 'O' rings made from Neoprene or was it Nitrile (I
forget). One belt drove the pair of large capstan flywheels and the other the reel hubs. The capstans were
mounted at the entrance and exit to the curved head block casting. The tape tension across the heads is
controlled by a lightly sprung 'finger' bearing on the back of the tape and the differential rotation of the
capstans. This is caused by the slight stretching of the main drive belt as it travelled between the two
flywheels. This was quite subtle in action, but provided enough tape back-tension for the recorder not to
need pressure pads (so reducing scrape flutter). Having two balanced quite large flywheels would also
help with speed stability generally and did provide some of the lowest wow and flutter figures of the time.
I am reminded here of the Nagra T with its two servo capstans tightly controlling tape spead and head
contact, though of course using vastly more electronics (and expense). Unfortunately, the drive belts in this
Ampex would after a while stretch and slip. This in the end would kill off these 'prosumer' audio (and the
remarkably similar video) machines well before their time. The regular adjustment of clutches and
changing of drive belts every few years, was in those days a normal operation. But the big problem was
that somebody changed what the belts were actually made of at some point (Neoprene to or from Nitrile or
Polyurathane or liquorice - something like that). With the result that they just didn't last long enough. And
just months after a major service fast wind would become something of a joke, and that was the premature
end of many an Ampex audio and industrial video machine. Having replaced the belts in my own recorder
several times over the years, I decided to obtain some 'O' rings as used in the oil industry (of the correct
size and similar shore hardness) from James Walker of Cockermouth in Scotland. Some ten years after
fitting these alternatives, the transport is still alive working fine and it just remains to be seen how long the
tyres last...

For those few who may be interested; the James Walker part numbers are:

JW Drg. No: 72-1995-30 and 72-1595-30.
Material: PB80 DTD 458A - B1
And their address is: James Walker & Co Ltd
Cumbria CA13 0NH
Great Britian

Meanwhile, the original Ampex belts have just melted away in their mini grip bags...

Why use horrible drive belts in the first place? Ah, a good question! Cheapness is the main answer I
imagine, because (even in the Nagra portables) you can get away with one motor in the transport instead of
three. Ironically though, belt drive is a rather good way of decoupling noisy vibrating motors from tape
capstans. And if you don't need the blistering wind speeds of a professional machine, one motor can be
arranged to give quite an acceptable wind performance. Belt drive can also have other advantages, in that
if you couple your motor to a pair of balanced flywheels of high inertia, you have the makings of a stable
and low noise tape transport. Indeed before the arrival of servo driven capstans, these Ampex machines
had state-of-the-art speed stability figures. (Actually, when the Revox A77 was introduced - the first servo
capstan tape recorder - some sound engineers said to its detrement that they could actually hear the
machine's servo working.) The motor, main transport casting and outer picture frame casting are all
decoupled from each other, and the reel hubs are driven by a second belt and a series of tyres and
clutches. Most of the metalwork in these Ampex recorders are nicely machined die-castings, and 'bent tin'
is only used for the electronics casing. Again, while Nagra would use machined-from-the-solid alloy
components, these old Ampex machines used alloy castings in place of bent tin. However, most 'domestic'
manufacturers of tape recorders at the time used bent tin and cardboard...

This ability to be able to play a tape in forward and reverse directions relied on a clever rocking
arrangement of pulleys that employed drive belt drag to correctly engage forward and reverse motion of
the reel hubs. The single, centrally mounted, bi-directional, mains synchronous motor drives the belts by
two pulleys, one of which is stepped to allow for speed changes (these were three speed machines). This
large motor has two sets of windings and is fitted with a cast and balanced cooling fan which also acts a a
flywheel. Finger operated sliding levers move the various internal components into and out of engagement.
The three tape heads (one for reverse play, erase, and forward record/play) are mounted upside down, in
the best fashion on a machined alloy bridge-like casting. This assembley also contains the forward and
reverse audio track changeover slide switch.

So how does this simple, non logic controlled tape machine automatically change the direction of tape
movement with the use of only five transistors? (One might ask...) Well, one of the drive belt idler pulleys
has a permanent magnet built into it which is in close proximity to a sensing coil. When the recorder is
under power the motor and all the belts are running, and the idler is therefore rotating about twent times a
second. This induces a 20Hz signal into the coil. Now, if in record mode, the user notices that the machine
is getting near to the end of the tape, he just pushes down a small lever next to the main on/off switch. The
lever connects the coil's output directly to the left recording amplifier's input, which now puts this 20Hz
signal directly onto the tape. These machines don't actually record in reverse, so while recording at the
end of the reel, the user has to swap the tape over in the 'time honoured' tradition. However, once that the
20Hz reversing signal is on the tape, the machine will (if instructed) change direction at that point. This is
done simply by always feeding the left channel off-tape signal through a 20Hz low pass filter and an
ampifier into a solenoid. In play mode, the previously recorded 20Hz reverse signal is picked up by the
tape head, low pass filtered, amplified and the solenoid actuates. The solenoid 'toggles' a pair of
switches; one to reverse the drive motor windings, the other to switch the playback heads over. With the
result that the motor slows down, stops and reverses, the heads switch to the other tracks, and we are now
happlily travelling back the way we came, simple eh?

The audio electronics are also about as simple as one can get, with the play responses cut at about 30 Hz
(to prevent any speaker damage from the 20Hz reverse signal). Just 5 (new for the 1960's) silicon
transistors per channel in the whole line level signal path. This particular recorder has the optional 6 Watt
RMS transformer coupled two transistor (!) power amplifiers.

I am rather ashamed to admit that I 'audiophiled' recorder up a bit over the years and it now has 'special
wiring and 'audio' caps. Thankfully not visible from the outside though, and the various 'improvements' did
seeem to bring extra 'Hi-Finess' (well possibly, might have done). Actually this old beast is still quite a
good audio recorder with the right tape etc, and can reasonably keep up with my other machines. It does
though suffer sometimes from a bit of track wandering, but that is probably because the pinch rollers are
over years old.

Oh, and I must of course mention Warmack Distrubting in Illinois, who over the years have provided the
most remarkable spares service for these machines. They are still able to supply many spares, and newly
manufactured belts and tyres to Ampex specifications are available. Many thanks Mr. John Warmack.

Update - well the James Walker belts gave up mid 2003 and I ordered some very good replacements
together with new tyres from Mr Warmack. The machine is now working well again some 38 years after it
was first manufactured...
1960's American High End.

The Ampex is shown here with a Dynaco FM3 tuner and a pair of Koss ESP6 electrostatic headphones.

I have owned both the Ampex and the Dynaco since new. Actually, the Dynaco was only recently replaced
by a 'modern' Yamaha CT7000. The phones were a lucky find a few years ago for 5, and after some
restoration seem to work pretty well. The Koss ESP6 were of course the World's first electrostatic
It had not occurred to me before I sat down to write this piece, but it is probable that these machines were
actually part of what once might have been thought of as the next step beyond the stereo gramophone
record. In those happy days before the nasty compact cassette, a natural improvement in fidelity over the
good old vinyl disk was thought to be the domestication of the 'wild and unruly' reel-to-reel tape
recorder. Even before the microgroove record, the first high fidelity stereo recordings were made
available to the public on magnetic tape during the 1950's. Expensive, large and short-lived, this
pioneering domestic stereo system was long admired for its fidelity. Time passed and the stereo LP
became the de facto standard, but in the mid 1960's Ampex and others started marketing high quality pre-
recorded stereo reel to reel tapes. Recorded at 3 and 3/4 inches per second on 7 inch reels, the two
stereo tracks were recorded in opposite directions. Copied in real-time (?) and with careful attention to
quality, these new tapes could quite hold their own against most LPs, particularly in America where the
fashion at the time was to use 'autochanger' type record decks with rudimentary crystal or ceramic
cartridges Under these circumstances tape would promise increased longevity, and of course had no
snap, crackle and pop (or indeed 'rumble'). The tapes just needed playback machines that were of high
quality and above all simple to use. Ampex were perhaps in the ideal position to build this hardware, and
one wonders (for them) which came first; the machines or the tapes. One might recall that in those days
Ampex were not only in the professional tape machine business, but were also making 'prosumer'
audiovisual equipment as well as 'custom' (the old term for High End) domestic audio equipment.
(Click for images of the machine's interior.)