Modelling a Tek S-6 sampler

Is that the traveling-wave thing? Post your schematic!

Traveling-wave samplers seem to have gone out of fashion; the really fast stuff is back to half-bridges. The SD-series (11801 scopes) were, I think, tw samplers, but I'm not sure about the new stuff.

Look up Agoston's patents for some theory.

I've got a pic of the guts of an SD-20 around here somewhere...

John

That will be some work. Do you have the classic Tek app notes of the era? Do you have spice models for custom made diodes? Do you need scans of some Tek documents?

Check this site for possible schematics for unit

=2E This seems to have a lot of manuals and circuits. Also TEK put out a manual pretty good on using TDR and all the theroy that they used.

Regards Cliff Schuring

It's the travelling-wave thing all right. I thought a half-bridge *was* a travelling wave sampler. One alternative I know of is the full bridge, as used in the HP54100 series.

But I got the model to work, after some additional diddling. I had put 300ps long transmission lines between the diodes, where in reality, they are only about 20ps long. That fixed it.

I have a photograph of the sampler with some simulation results on . The LTspice .asc file is there too.

Jeroen Belleman

Wow, that is awesome. How are you going to get the special SMAs back on? Do they need a specific torque or do you guess? Will the head perform the same after re-assembly? How did you determine the diode type? Just curious. I have a S-6 but never tested it, my lab is in disarray. Thank you for the work.

The 6-diode thing is the traveling-wave sampler, I think. It has internal delay lines.

The first samplers were mechanical contacts.

The first electronic sampler was a pentode tube, with the screen grid pulsed to sample the grid signal.

The first solid-state samplers used avalanche transistors to make the sampling pulse and a single point-contact diode as the sampler. Lumatron did this, then Tek in their type N plugin.

I'm not sure who did the first 4-diode bridge sampler, probably HP or Tek.

HP's 185 scope had a plugin that used a 2-diode half-bridge to get to

4 GHz, ca 1964. I have one.

LeCroy revived the single-diode sampler a few years ago.

Mark Kahrs has a really cool paper on the history of sampling.

This is my 2-diode sampler. It's good for about 70 ps, 5 GHz.

ftp://66.117.156.8/Sampler1.JPG

John

Yes, it's a pretty gadget and with respectable performance, despite having been designed in the 1960s.

There are a few lines about changing the SMAs in the S-6 manual but not a word about torquing them correctly. The centre conductor is spring-loaded and pushes against the metallized edge of the alumina substrate of the circuit. The substrate is very thin! An intermittent contact in one these was what pushed me to take it apart.

As for the simulation, I didn't determine the diode type. I guessed they must be low-Cjo Schottky diodes, so I selected the 2207 from Avago.

After re-assembly, it behaved just the same as before, but without the poor contact.

Cheers, Jeroen Belleman

I think the 60s were an awesome time for electronics, the pure hardware-driven variety.

I just took it apart because I felt like it. It's good to know it might still work after all!

Excellent.

The 1S2 sampler/TDR used a pair of point-contact GaAs diodes in melf-looking glass packages. They were pressed against notches in the pc board by a plastic holder gadget, and would SNAP! into place. The diodes were very shock sensitive, and the manual says that it "is possible" to replace then without ruining them.

The 5S14 and 7S14 samplers used soldered-in mercury batteries to back-bias the sampling bridges. They'd die after a few years, and it was an hour or so of hell to replace them.

John

Never understood that, did some overly academic engineer see a battery symbol in some analysis and thought he really needed batteries?

I notice you've got a few vias stitching ground planes together there...I thought you'd pooh-poohed that idea sufficiently?

Cheers,

Phil Hobbs

I did that when I was young and foolish.

John