Amazon

Do I? :)

NT

Of course you do. You just don't notice it because it is so common.

I was recently being inrerviewed by a cute young thing about my mental state. She asked me if I was any different in the last six months. I replied, How would I know?

Haw! That's a good one!

No, seriously, few crazy people notice that they are crazy.

I'm in a research study about brain injury. Every six months I get an MRI and some silly tests and an interview. The hospital bill was $78,000, of which my insurance covered all but $740. They pay me $250 per visit to be MRI'd. So I'll come out ahead after my next exam.

The MRI is done in a full RFI screen room with a view, very impressive. There's a faint constant squish-squish background noise. That's the helium liquifaction machinery. MRIs used to vent the magnet boiloff, but now they reliquify it. "LN2 costs as much as beer, and liquid helium costs as much as whiskey."

They scan me at the VA hospital for some reason; I'm not a vet but the machine is available there. It's on a bluff overlooking the Golden Gate, amidst a cluster of old WWII gun batteries, near the Palace of the Legion of Honor museum. The setting is stunning, and it's fun to hike around in the ruins.

"Why yes I am. I've always been different."

..even under a low voltage standoff?

I've come across chocolate where re-voltage is an issue, but not too often. Similarly standoffishness is mostly not a problem. But one must be thorough and test.

NT

ever solve this problem?

No. We're sticking to rosin flux and solvent wash for now, which has always worked.

We're now doing the layout on a test fixture board that has about 270 relays. We'll use that as a test case for water washing these relays.

Building an east german "electronic" phone switch? I'm a fan of the clatter, noise and sparks from step relays and true relay racks for elevator control systems.

Do you give tours if somebody is in the area?

Yeah, the old Bell COs sounded like a WWI machine-gun battle. It's amazing the immense amount of logic that they did with relays.

Sure. Visitors are always welcome.

I'm using tiny Fujitsu surface-mount relays. They are barely audible when they switch.

I think we discussed the test board. There are 5 internal busses, A B C D and E. And there are 68 DUT-pin to bus mux's, each three DPDT relays. More relays connect test equipment onto the five busses.

This will replace lots of old test sets and test software. The new stuff will be coded in Python and generate HTML test reports.

I might post the schematic here one day. 25 pages, 255 relays, Ethernet control. Heck, maybe we should sell it.

I've recently started playing around with getting Python to control test sets- running on a Raspberry Pi. The old 34401A was a bad place to start - its SCPI serial interface is really fussy. Ethernet (and maybe GPIB via a USB adapter) should be much easier. It's nice having a bit of real hardware accessible (such as an SPI bus) as compared to using a PC.

--sp

My kids are really into the Python thing. We are takling to Keithley (serial) and lately Fluke (ethernet) DVMs, an Agilent SMU, an Agilent counter, and test board, and the DUT. It's very impressive and modern looking. All on PCs, generally Linux.

The B2901A SMU is a cool automated test tool. It will source/sink voltages up to 200, measures femtoamps.

I bought a Keithley SMU and sent the damned thing back. It was awful.

If handheld isn't a requirement, you can get an HP 70820A for beans.

Cheers

Phil Hobbs

Looks very nice- the dual-channel B2912A is not particularly cheap ($13.5K US). Maybe I'll make dedicated current sources with a few opamps/in-amps and a reference..

They Keysight guys were demo'ing some handheld boxes (look *very* similar to R&S products) that did SA, some kind of fudged TDR calculated from frequency domain data etc. The cost of the *options* (which were software keys mostly) multiplied the base cost by a huge factor.

SMUs are really just programmable power supplies. They shouldn't be as expensive as they are.

Being a boat-anchor fan, I just use a Kepco BOP and a LabJack for that sort of stuff. Razzpis and so on are cheap enough that it's probably worthwhile to retrofit some of those boxes. You can get BOPs for a few hundred bucks.

Cheers

Phil Hobbs

The Agilent is impressive. You can do a diode curve from fA to amps, microvolts to hundreds of volts. But there should be a somewhat less impressive, much cheaper version too.

Ours are going into production test stands, and it's nice to be able to send an SMU out to a cal lab and get it recertified annually, to keep all the paperwork legal.