IBM 2260 acoustic delay line

[Jon Elson]

On 12/12/2015 09:13 PM, Eric Christopherson wrote:
> I'm reading about those terminals and find it just 
> fascinating how they used acoustic delay line memory to 
> remember the pixels. But I have lots of questions: 1. Did 
> the cables connecting the 2260s to the display controller 
> actually contain the delay lines themselves, over the 
> whole length; or were the delay lines just inside the 
> controller and then some electronic signal was sent out to 
> the terminals?
The delay lines were little coils of steel wire in a 
housing. Probably some 30 feet or so wound up in a spool 
about 9" diameter. See these for some examples.  The 2nd one 
might actually be a 2260 delay line.

https://upload.wikimedia.org/wikipedia/commons/e/ef/Torsion_wire_delay_line.jpg
http://www.glennsmuseum.com/ibm/med/med_ibm_wire_memory.jpg

I actually have a bunch of 2260 cables here, a couple coaxes 
for sync and video, and about 15 wires.
> 2. I would think that the wave travelling along the delay 
> line would weaken over time. How was it refreshed?
You read the output through a read amplifier, squared it up 
to a digital signal, and re-launched it onto the wire.  if 
you wanted to change the info, you switched a multiplexer 
and inserted the new data instead of recirculating the old 
character.
> 3. What kind of speed could be acheived, and did this 
> depend on the number of connected terminals? 
It depended on the 360 CPU, which were not all that fast.  
But, WAY faster than just about anything at that time.  I'm 
not sure what you really mean by "speed".  You could alter 
the characters on the page you were viewing, and then 
transmit that to the computer.  This all happened at IBM 360 
channel speed, so quite fast.  (Actually, it was almost 
certainly running on the multiplexer channel, so each 
character transferred caused a channel request to put it in 
a buffer, then when 4 were in the buffer, the channel would 
cause a memory transfer.  (4 bytes was the memory word width 
for an IBM 360/50.)  So, that all happened in milliseconds.
Still, this was lightning-fast for that time (1972 or so).  
Since the OTHER way to do programming was editing decks of 
punch cards with a keypunch, then submitting it for batch 
processing and getting your error messages back 4 - 8 hours 
later, it was REALLY a step forward.  I wasn't in any 
courses where I was actually allowed to use the 2260, though.

Jon