Really OLD logic levels

before

Don't forget the high threshold logic families (15 V) used in industrial applications.

Why would anyone use such high voltages for digital logic ? This would require quite high impedance levels and due to the stray capacitances, quite slow operation.

For instance the 5965 double triode was operated at about 100-150 V anode voltage supply. The early semiconductor diodes used to build diode/tube logic NAND/NOR would not have tolerated much more.

As strange as it might sound, the early semiconductor diodes were more unreliable than tubes.

I've never heard anything about vacuum state logic levels, if any companies even defined their own logic standards -- I would guess something in the neighborhood of, say, -20/0V (requires low current C- supply) or 50/100V (requires cathode bias = wastes power).

Tubes aren't so bad. I've done, I think it was -80/0V drive into a sweep tube, which is equivalent of driving a 6000V, 1A MOSFET to produce sub-150ns edges (comparable to... lazily driving an IRF540, I'd say). Such high voltage MOSFETs don't even exist [yet], so I guess toobs win that battle. :-p

Umm, read the link..

Tim

before

I worked with stuff negative voltage, even saw negative logic Collins logic, dec logic, zerox logic, general dynamics dynamic logic.

Greg

It was because the power supplies (highly redundant, a system would operate with a few hundred of 'em) weren't regulated, except by corona leakage to atmosphere.

Higher tech for regulators (gas discharge tubes, Zener diodes, bandgap references) was unavailable. :-)