"Tyrotek" analog computer
hilpert at cs.ubc.ca
Mon Jan 12 02:06:26 CST 2009
dwight elvey wrote:
> > From: hilpert at cs.ubc.ca
> > I'm currently working on refurbishing a small analog computer from the
> > mid-60's. It has the name "Tyrotek" on the front panel, in the middle of a sine
> > wave curve. (My) web searches turn up zippo about this name (other than some
> > guy using the word as a moniker).
> > The only other clue to a manufacturer is "Non Linear Mfg." in the foil of the
> > printed circuit boards. No model number, no identification plate.
> > While it is smaller for an analog computer and appears to be intended as an
> > educational or training unit, it is built to industrial standards, with a nice
> > sizeable plugboard using standard IBM plugboard wires, fiberglass PCBs,
> > discrete FET-input op-amps, etc., so it's not just some gimmicky thing.
> > Eventually I'll have a web page for it, but right now I'm curious as to it's
> > manufacturer origins. Just a guess, but one line of speculation I'm making is
> > that "Non Linear Mfg." may have been the manufacturing arm of Non-Linear
> > Systems of digital voltmeter fame, and Tyrotek was a subsidiary of, or
> > otherwise associated, with NLS.
> Sounds intresting. I do question the date. It is more likely in the
> 70's. You might look at some of the date codes from the parts.
> How many amps does it have? Can they all be configured as integrators?
> Any special operators like multipliers? Does it have a two speed
> integrator function like the Comdynas?
Oh, it really is mid-60's, date codes are one of the first things I look for
on any piece of quipment. Latest date code is 6734 on a 'lytic cap,
66xx on some other components.
If you were questioning the date because of my reference to op-amps,
("discrete FET-input op-amps"), to clarify, I mean the entire op-amp
is discrete, not discrete FETs feeding an op-amp IC. There are no ICs
in the unit.
There are a total of 10 amps. 4 can be configured as integrators or summers, 4
are just summers. 2 have associated circuitry to make them into multipliers
or dividers, or can be used as just summers.
I'm not sure exactly what you're referring to as the two-speed function, but
there was an option (going by the front panel labeling and some wiring) (option
not installed) to change the "Time Scale" for the integrators. I believe it
changed the integrating caps for all the amps, but it's a 'global' change, not
individually selectable per amp.
The unit had been languishing in my basement for a few years, on the verge of
being thrown out. Now that I've gotten into it, it has been fun figuring out
something 'completely different', never having worked on or with one previously.
Of course, I don't have the manual for the thing. I've done the schematic,
which has made it possible to figure out a lot of it.
There's not a lot on the web about analog computers, but there are a couple of
scans of manufacturer brochures and an article or two that are cluing me in to
the standard practices of the time.
It's currently in a million pieces across 3 workbenches, almost ready for
reassembly. (Most of the nuts and screws and a few other metal parts had grown
that white powder that develops under certain conditions on some metal
surfaces. I've seen it on other equipment, I presume it's an oxide of the metal
plating, although it also has the appearance of a mildew or fungus. I'm not sure
what element the plating is. Fortunately, it's only a 'mechanical' issue, it isn't
affecting the circuitry or contacts.)
If you have experience with these things I may be back asking some questions in
a while, once it's all together and functioning and I can put up some pictures.
What can be done with it when it's functioning will be another question.
The only built-in 'output' device is a solitary analog meter. I have my eye on
an HP analog chart recorder buried at the radio museum, to use as an output
recorder, but I expect it has ink issues in the way of getting it going.
(At least the plugboard looks cool.)
More information about the cctalk