Is there a good book for learning about valves/tubes?

Sure, there's nothing about tubes that makes parallel architectures impossible.

You can do a lot of fun stuff that isn't practical. My old colleague Mark Galt, who helped me with my thesis project long ago, did an all-mechanical Nixie tube clock that is a thing of great beauty:

I just ran across it yesterday, in fact. Check out the 45-second video.

Cheers

Phil Hobbs

Some of those tube machines had hardware floating point.

There were giants in those days, for sure. I got to meet one or two of them when I was at IBM, and heard stories about many more.

It was much more practical in the old days, when (a) there weren't any good alternatives, and (b) you had your own tube design and manufacturing facility.

I don't think anybody makes 6CW4s anymore.

This from the inimitable Ken Shirriff:

Cheers

Phil Hobbs

The oldest working computer in the world uses dekatrons, valves and relays.

The Elliott 803 is a serial computer using germanium transistors and magnetic components for the logic.

Decca Navigator made a TTL serial computer for avionics applications.

They might be a starting point for an architecture.

Or you could be seriously radical and build a computer using neon bulbs. I doubt the reliability would be good :)

So you might have noticed there is a wide opinion on the matter in hand :-)

RCA's RC-** series of tube data books may also be useful. They include brief explanations of how tubes are constructed as well as several examples of practical circuits. One of the mid-1960s editions was among my first books on electronics.

Search for "RCA receiving tube manual"

Those things sure were ugly.

HP gear was made that way, lots of phenolic and ratty wires. The Tek ceramic terminal strip construction was maybe the first beautiful electronics.

I like to design beautiful electronics. It actually works better too.

Cheers

I used to own a couple of those modules (sans tubes) when I was a kid (circa 1972). I had no idea what they were, so I stripped the passives out of them and threw them away. :( Made perfect sense at the time.

I agree in general, but sometimes a man's gotta do what a man's gotta do, ya know?

Cheers

Phil Hobbs

The thing about PCB (schematic and layout) beauty is that the modest time spent tweaking cosmetics is actually another opportunity to think and review, to accidentally correct bugs or discover improvements. The time spent is usually repaid by getting it right first pass. It's like reading and beautifying your own code before you compile and run.

That board works first pass. So far. It makes adjustable delay and width and amplitude pulses from 0.5 to 45 volts peak, clean with 1 ns edges. I did that out of lockdown boredom. It's all analog, trimpots, because I didn't want to get involved with uP or FPGA code.

There must be a use for it somewhere.

I discovered these:

They make anything up to 75 volts into something usable, 12 in my case. My box can run from 24 or 48 volt warts.

For transmitting use: The Care and Feeding of Power Grid Tubes, by Eimac, if you can get hold of one. (I'll keep mine).

Has anyone mentioned the RCA receiving and transmitting tube manuals?

They had a good PMT manual too.

I'd like to see the datasheets on the EEV Klystrons that were used in a Comark UHF TV transmitter that I maintained. 65KW output, per tube. It's be e over 30 years, so I don't remember their type number. It was the last Com ark built with Klystrons. The next transmitter used Klystrodes, instead. Hi gher efficiency, less cooling required and lower electric bills.

ten so

dimension is

k up "1940

d free

it may be

tubes?

. In the

uestion

circuits

out

The thought just occurred that to me that perhaps he has no electronic experience at all and heard the word vacuum tube some where and wondered what the fuss was about. Twenty five years ago I had a young guy call me at the shop asking about repairing a jukebox that had no sound. I asked if it was solid state or vacuum tube and he had no idea. Being a ham from 1953 I was shocked at his reply. I thought ever ybody knew about tubes.

ten so

dimension is

k up "1940

d free

it may be

tubes?

. In the

uestion

circuits

out

about.

repairing

and he

ybody knew

Back in the '60s and early '70s, we had customers who didn't know whad bran d of TV they owned, even though they stared at their TV for hours every day . Others asked, 'Aren't all TVs RCA?'. Each service truck's inventory of sp are parts was tailored to a couple brands. If the wrong brand was given, th e tech had to haul their set to the shop, rather than make extra trips. It was amazing the high percentage of people who were too lazy to walk over to their TV to get at least a brand, if not the model number so the tech coul d take the service data with them. That was a big reason that I left consum er electronics.

I posted that somewhere up above in the thread. The RCA tube and transistor manuals were among my first books on electronics.

Oh, gigahertz makes just as much sense, with Nuvistor technology and some 3D printing of electrodes. Some of the old serial-CPU gizmos with delay lines were relatively capable (HP 9100A comes to mind) and a delay-line memory is wonderfully parts-count minimal.

I didn't know about Klystrodes--pretty cool devices actually.

Cheers

Phil Hobbs

If you want to minimise active devices, you could copy the old techniques of using magnetic components in logic gates.

I'm sure Peter Onion would be only too happy to discuss in detail. The full schematics are available at TNMoC

There are no delay lines in a 9100. Memory includes core, PCB inductive ROM, and diode ROM.