Nice. Hadn't heard of those. Looks like I'll be adding a few MCP1407's to my next parts order.

[without]

There will be the usual large inductive kick-back when the upper contact opens,

Inductive kickback from what exactly? When the capacitor is fully charged, there will be no current flowing in the coil. Opening the switch contact at that time will cause no arcing since there is no magnetic field being generated. You're likely thinking about a continuously held coil, which has current constantly flowing into it (and an associated magnetic field), but that's not the case here.

But as I said, I need a circuit with an on-off switch...

I presume you mean without an on-off switch. But you haven't fully answered my questions in reply #14 yet.

I think he means with a single pole push button switch.

I'm pretty sure there's a way to do it, but it will probably take multiple capacitors, a multi-pole double-throw dual-coil latching relay (perhaps several) and is going to become rather complex, rather quickly.

I can only tolerate power consumption when pulsing the solenoid.

I'll have to think on that.

Also, I don't think there will be much of an inductive kick in this type of circuit so a flyback diode should be unnecessary. Initially, when the button is pressed, the capacitor will begin discharging peak voltage and current through the coil. As the capacitor's charge depletes, the voltage and current will drop to zero along with the field generated in the coil. When the button is released, there will be an sudden inrush of current to the capacitor which will switch the coil. But as the capacitor charges, the current will drop to zero as will the field of the coil.

If you just want to play it safe, add a snubber circuit: put a 0.1 uF, 50V ceramic capacitor in series with a 47 ohm resistor across the coil to damp out the back-EMF or use a 'Transorb' or equivalent.

How quickly will you be switching from one state to the other? That could be a problem with this circuit if you change states before the capacitor is able to discharge (or charge) enough energy to switch the coil.

Use your switch to control a SPDT relay, which changes the connection of the capacitor.

That would have been one of the SAGE computer sites:

http://www.computermuseum.li/Testpage/IBM-SAGE-computer.htm

There was another in Minot. I saw some of the equipment hauled out of there when it was dismantled; it took a couple of hefty forklifts just to move the racks.

You may need to experiment to find a suitable capacitor value, but this should work:

I know I have a nuvistor somewhere in my collected piles of... *ahem* superbly handled integrated technology , I just don't seem to be able to lay hands on it at the moment. I hope it hasn't 'grown legs' and walked away.

However, I was able to find a few examples of early miniature electronics technology in my 'collection'. (see attached pic)

On the left, a couple of pencil tubes from the 50's. At top, a 2N158 germanium transistor from 1956 or `57. Note that it's bigger than a nuvistor. And on the right (just for fun), a few fetrons (look them up) from the 70's.

Now, to the original question, would my job or hobby change if transistors had never been invented?

Probably not. I was still repairing and maintaining tube-based equipment well into the 90's and currently sitting next to my workbench at home is a Harman Kardon Stereo Festival receiver from 1958 that's getting new caps and FM tubes, right after I finish refurbishing a Pioneer AM/FM/turntable/8-track console from the early 70's for an old flame um, friend. (Don't tell my wife.  )

But all this new-fangled stuff is fun, too.

Exactly. Nuvistors were used in microphones and a few miniature radio receivers that I know of. A former co-worker of mine (sadly deceased) had a portable radio that used nuvistors. It ran on a couple of C cells and a small 45V battery. I suspect the military used them as well in missiles and aircraft before transistors took over.

In 1959, the Nuvistor was approximately the size of point-contact transistors of that era:

http://en.wikipedia.org/wiki/Nuvistor

1959 was also the year that the planar method of transistor manufacturing was developed, which led to the development of modern ICs (and probably fueled the demise of microminiature tube development). Planar manufacturing made transistors cheap and much easier to manufacture in quantity.

It's worth noting that primitive point-contact ICs were first made in 1958 but were very expensive and time-consuming to produce. When the planar method was applied to ICs, well... here we are. 

It seems that they were designed to fail, they just failed a bit too soon to generate more revenue for the manufacturers.

That's the rule with consumer products: make it cheap, make it fast, make it just good enough that 90% last past the warranty period, then sell `em another one.

However in reality the back-light flickers annoyingly so that the numbers are almost impossible to read.

That's because the backlight is most likely a simple neon bulb.

(Wonder why they don't just use a row of LEDs? It is simple and intuitive, has no mechanical parts and can be read across the room!)

Because it's difficult to reliably power a row of LEDs directly from a line voltage that may vary +/- 5-10%. The additional power supply components would drive up the price.

And Lasers.

Gotta have lasers.

And be under $5.

[1]

That table does not describe the function  F(X)=XY’+YZ’D’+YXZD’

X   Y   Z   D   F(X) 
.
.
.
1   0   1   0   1    <--- Function is true for this condition
1   0   1   1   1    <--- Function is true for this condition
.
.
.

Either the function or the table is incorrect.

My Texas TI-50 calculator! Cost me about £50 back in the mid 1970's, now I can get more calculating power from something that costs £3!

SR-50, not TI-50.

Probable explanation: Each digit has 8 components, 7 segments and a 'dot', hence 8 segment leads and 4 digit leads.

As Crossroads said, you'll probably have to map them out by trial and error. Based on past experience, there is probably no rhyme or reason for the pin order.

If your choice of parts sources are limited, you'll have to use what you can find.

Unfortunately, since you don't know the specifics of the keyboard scanning circuitry, it's difficult to tell if any particular optoisolator will work. You will probably have to experiment and find out.

An optoisolator with an bipolar transistor output (NPN, PNP) will be polarity sensitive. An optoisolator with a FET or MOSFET output (NTE3085, H11F1) will not.