"Oddball"(LGP-30)

[Dave McGuire]

On Jul 1, 2007, at 5:13 PM, dwight elvey wrote:
>>> I remember writing programs for on of these to solve a  
>>> statitical  problem
>>> for data on a mass spectrometer. I recall waiting on power up while
>>> it loaded the data from the metal tape( sometimes it needed to  
>>> be  rebooted ).
>>> I'd love to find one of these. I alway wanted to see how they   
>>> dealed with
>>> the different speeds of the metal tape and the memory. There was no
>>> exact speed control on the metal tape. It just ran from the  
>>> spool  speed
>>> with no capstan.
>>
>>   I'd imagine they just used a phase-locked loop for clock  
>> recovery  and clocked the data into the machine asynchronously.   
>> Is it more  complex than that?
>
> Hi
> I don't think that would work. First, the calculator had no RAM other
> than the delay lines. The metal tape had 2 rows of holes, one was
> clock and the other was the data. The motor that ran the tape was
> just a free running DC motor. The only thing I could think is that it
> may have buffered just a few bits and then burst them into the
> delay line as the slot for them curculated around. I suspect that the
> delay line was slower than a single bit of the tape but faster than
> 4 or so bits of the tape.
> Remember, the delay line has a fixed delay. The tape had a variable
> delay. It ran faster as the tape got near the end. There was
> no speed control on the motor. I was curious about this at the time
> and confirmed it with an oscilloscope.

   Ahhhhhhhh I see...so it's nowhere near as simple as I thought.   
Wow...how DID they get away with that??  What could they have used as  
a buffer, a few bits of core?  Flip-flops?

             -Dave

-- 
Dave McGuire
Port Charlotte, FL