Fast-assed signal switch

Hamamatsu makes a sampling scope that takes an optical input into a photocathode and then sweeps the electrons, streak-tube-style, across an aperature into a photomultiplier. They demoed it for me at the factory in Hamamatsu [1] but, in typical sampling-scope style, couldn't get it to work.

John

[1] Worth the trip, if ever you get the chance. It's an immaculately clean, nicely laid out factory, with everybody wearing slippers, with blue gas flames everywhere. Sort of a cross between Silicon Valley and Hell.

Right, that's the streak camera back end. Perfect for looking at optical signals between about 300 and 900 nm, but much less useful otherwise, due to the O/E conversion problem.

I've been to Japan once, about 10 years ago, but although I gave it a first-class try for a whole week, I couldn't learn to like raw fish. I'm also 6'4" and 260 pounds, so 14 hours in an economy class seat is a good approximation to Purgatory. (Back then, IBM had a sensible policy--you were allowed to go business class if the flight was over 8 hours. My ticket from JFK to Narita was $5k.)

Cheers,

Phil Hobbs

Actual flames? What for?

robert

Glasswork would be my guess. They make PMTs and other types of vacuum (and gas) tubes.

Best regards, Spehro Pefhany

--
"it\'s the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

Yup, they're melting glass everywhere. They even have lathes that use flames instead of cutting tools.

The photocathode activation is cute. A PMT has its seal-off tube still in place, and it's in a light box, and powered up into a microammeter. Very home-made, with fluorescent tubes for the calibrated (?) light source and old needle-type meters, 6 or so tubes at once. Inside the sealoff tubes are little boats full of caesium compounds or whatever, and there's a flame under part of each tube. The operator uses a magnet to slide the boats in and out of the hot section of the tube, evaporating the photocathode stuff onto the inner surface of the tube, and maybe the dynodes too. Lots of skillful juggling gets six tubes tuned up in one session.

They make one pmt that looks like a basketball. They're the ones that exploded (thousands of them) at a water-filled neutrino detector.

John

The earliest commercial samplers, the Lumatron junk and the Tek N, used an avalanche pulser and a single diode sampler, which was very nonlinear and kicked a huge pulse out the input. HP's first sampler was an avalanche pulser too, at 700 MHz. They used it to accidentally discover the SRD, and soon they used srd's and got to 1 GHz, then soon to 4 GHz with a dual-diode feedback sampler.

Tek's 7S14 dual-trace sampling plugin, much later, used an avalanche pulser with no srd. It was sold as a 1 GHz gadget but was typically twice that. They used mercury batteries to bias off the sampling diodes (a full 4-diode bridge, I seem to recall) and they are a nightmare to replace, so few functional units still exist. Too bad, they were very nice, with signal delay lines and excellent internal triggering.

John

Don't you mean "imploded"?

Don't be a PITA. Nobody likes a PITA.

John