Power outage in San Fran

Anywhere above 4 mA, you're within 1 dB of the shot noise in 50 ohms, so there's not too much point doing anything else unless the dynamic range is too large (i.e. you'll be below 1 mA sometimes).

A quiet 50-ohm amp can get you down to ~70K noise temperature if you don't actually connect a 50-ohm resistor to its input.

Cheers

Phil Hobbs

One of our products has a dumb textbook photodiode+TIA and does 160 MHz. We added a cascode transistor and a faster amp in a test board and got to around 500. The opamp feedback resistor is over 50 ohms but not hugely so, so the s/n is better than using a 50 ohm load + voltage amp, but not by a big factor.

Phil uses large-area detectors and mourns the loss of every photon. We do fiber, and 2 milliwatts of optical power in a 9 micron fiber is about 40,000 times the power density of full sunlight. The active area of our detectors needs a microscope to see, and the capacitances are very low.

ADA4817? or some other opamp?

Right I'd like something a bit bigger. ~laser beam size.. A few mm on a side.

George H.

I usually use a current-mode amp, more bandwidth and higher slew rate. I was playing with using MMICS as the amp, but it's not worth the hassle. I need DC coupling and flat frequency response.

There are some cheap, low noise telecom ROSA parts, but they are usually AC coupled, with AGC, so aren't suited to pulse use.

The odd physicist can. Back in the day Pippard was working on Fermi surfaces he had a 2MW electromagnet that was only run at night (when it was cheaper) and had to give notice to CEGB prior to switch on or bad things happened. The regulated PSU supplied 27kA at 75V to make a steady

The paragraph starting "By 1959" just before Josephson junctions gives a bit more detail.

NMR and HEP storage rings these days are a lot less energy hungry thanks to modern superconducting magnets.

A very nicely written obituary. I remember him well.

John

Marcus Oliphant was another of that kind of physicist

In 1955, Oliphant initiated the design and construction of a 500 megajoule homopolar generator (HPG), the world's largest. The HPG was intended to be the power source for a synchrotron, but this was not built. Instead, it was used to power the LT-4 Tokamak and a large-scale railgun that was used as a scientific instrument for experiments with plasma physics.

What isn't mentioned in the Wikipedia article was that the HPG started off with liquid metal brushes - a mixture of potassium and sodium - which event ually blew up, very nearly killing one of the technicans involved. They wer e replaced with more conventional copper-graphite brushes which worked fine , and didn't endanger anybody.

yes, that sounds about right.

Fairchild bought a Cambridge Instrument EBMF 10.5 electron beam microfabric ator which they used somewhere in Californian to make the masks for their 1

I had to do a bit of work on the scanning amplifier that drove the electron beam around to get it quiet enough to give clean patterns with - then - ne w electron beam resist that they were using, but it's not the kind of work that could have taken out a substation if it had gone wrong.

This dates back to the mid-1980's, and one might imagine that any machine t hat old would have been junked by now, but the two that were sold to Austra lia at the same time still seem to be in use.