Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
OT: HP70843B 12 GHz BERT info needed
|
I posted this in the hpagilent group a while back, but no one seems to have info. That group seems pretty quiet lately, and I don't see indications that it's moved to groups.io yet. Since it's mostly the same folks in both groups, I thought I'd put it here for a bit where there's more traffic, and hope for some more visibility. Sorry that it's not Tek stuff, but it is highly related to fast pulse generators, which are near and dear to our hearts.
Anyway, here goes - it may be fun and useful to talk (some more) about fast pulsing technology, and GaAs devices, etc. I'd appreciate any info on this unit or its guts. "I just picked up this beast for cheap, figuring there would be some fun microwave goodies inside. It seems to be a fairly obscure unit, with no info on the guts, and barely any about the unit itself. I found an operating and programming manual, and that's it. I've already parted it out and found lots of interesting stuff, including the four main source output modules, rated for tr and tf of about 30 pSec. These should be good for fast pulse generator work, but not unless I can figure out how to power them up and activate the control lines. The outputs apparently can run up to 12 GHz (or Gb/sec) data rate, up to a couple volts swing on 50 ohms. They appear to be some type of GaAs amplifier, so could be complicated making the biasing all work out right without any info or some reverse engineering of the support circuitry. Does anyone know of any technical info on the innards? This is from the era of "No User Serviceable Parts Inside," and no manuals about the insides. My plan is to put some of the sections back together enough to trace out the supply connections, and power up the pieces and see if I can get some of the items operational. I realize it would have been cool to use it in its complete form with lots of capability, but it would have taken lots of effort and more stuff to cable it up to a 70004 display, and maybe some SW, just to see if it worked. I'll be satisfied if I can get the output modules figured out and working. There's also a bunch of other microwave parts, and lots of EL-series "modern" high-speed ECL parts all over the place. The SMPS is big - over 400W out I'd say, and I plan to use the whole carcass as a big multi-output supply after I figure out the modules and such. It has heavy -5.2V and -2 V, handy for ECL work, besides the usual +5V and +/-15V. There's room to add other supplies where the high speed pulse guts were, and plenty of cooling - six 3" fans pull air from one side, through the guts, and out the other side. Two are for the PS, and the others for the working guts up front." That's what I wrote before. I've been studying it more and have some interesting observations that I'll report on later. Ed |
|
Hi, Ed.
toggle quoted message
Show quoted text
I'm forwarding your message to another Ed I know who may know something about this beast. I envy you for finding it! Jim ------ Original Message ------
From: "Ed Breya via Groups.Io" <edbreya@...> To: [email protected] Sent: 7/21/2018 10:53:01 AM Subject: [TekScopes] OT: HP70843B 12 GHz BERT info needed I posted this in the hpagilent group a while back, but no one seems to --- This email has been checked for viruses by AVG. |
|
Can you post some photo of the board somewhere on the internet? It is
toggle quoted message
Show quoted text
always interesting see what is inside mysterious instruments Il giorno sab 21 lug 2018 alle ore 20:20 Jim Ford <james.ford@...> ha scritto: Hi, Ed. |
|
I'll try to post some pictures in the hpagilent group later. There are quite a few pieces in this complicated system, but I've gotten it somewhat figured out, structurally.
The 70843B is normally hooked to a 70004A or equivalent display MF via two local control cables (I forget the name - something HP specific control interface), and the 70004A needs a synthesizer (I forget the model) module installed, which supplies the clock (100 MHz - 12 GHz). The source unit in the 70843B amplifies and squares up its clock input with a GaAs amplifier, and sends it to the front panel in true and inverted forms via two of the output amplifier modules. The clock internally also goes to a band-switched variable phase shifter system for skewing over a wide range. Since the system knows and controls the clock rate, the bands and amount of phase shift can probably be worked out precisely to desired time delay or whatever else is needed. The clock from the input amplifier goes through an HP8768K coaxial relay which routes it to a selected phase shifter module - a 6-12 GHz variable (with three varicap sections), a 3-6 GHz (similar), or a <3 GHz block with apparently PIN-switched delay lines of 250 pSec, 500 pSec, and 1 nSec, plus one varicap variable section. There is also a straight through mode that bypasses all the shifters. A second HP8768K selects the appropriate output and sends it to another GaAs amplifier, then that becomes the internal clock for data pattern generation. The RX unit has the exact same kind of phase shifter assembly, but hooked up a little differently. The input amplifier clock outs are not used, just terminated, and its version of its delayed clock is used for decoding the data stream. The TX and RX units each have a block full of small board sub-modules (about a dozen) that do the sub-rate multiplexing (TX) or de-multiplexing (RX). I have been able to ID some of the parts, which are mostly GaAs IC source-coupled logic drivers or switches, made by NTT, circa 1990s. I'll list them all here for reference, once I figure out all the types and quantities. These run at the top speeds involved, and apparently the rest of the stuff runs at one-quarter or one-eighth the clock rate. I'll have more info later, and maybe some pictures too. Ed |
|
Here's my other post on hpagilent from before:
"I've been doing some searching on the part number of the output modules 5086-7956. Got some hits including apparently used in some version of the 8590 series SAs. I couldn't find anything like it so far in documents on these SAs though. Maybe it's just a wide-band (100 MHz-12 GHz) amplifier, used in other gear too." I've been studying the output module more, trying to trace out the small board attached to it. I opened up the module too. The circuit is very tiny stuff on sapphire substrates. Looks like three identical GaAs amplifier or logic chips cascaded. I need to set up a microscope to see more detail. Live voltage measurements on the interface from the big board to the small board/module were confusing because in operation, I think the whole amplifier is biased up and down to get the right DC out, while its input is AC coupled. Without knowing what conditions the instrument defaults to, it's hard to tell what it's telling the module. Opening it up will help to tell at least what's what in there and the basic signal path and structure. Will post pictures and rough schematics if I figure out enough to make some sense. Ed |
to navigate to use esc to dismiss