#653

Episode 653
I got a free analyzer. What to do with it?
Be a Patron:
/ imsaiguy

[@amirb715]

"R" is for reference (not return loss) and is a "portion" of the source signal that goes into your device. "A" and "B" receivers receive the signals that you want to measure against the reference (A/R and B/R are measured) so B may be the output of the device so B/R is S21 and A may come from return loss bridge so A/R is your S11. That's how these instruments were used. We had one at school back when I was a student...

[@NimbleJack3]

+1. This is how the channels are intended to function. There are actually dedicated bridge products designed to plug into an input channel and provide both directional coupling and RF power measurements with high precision.

[@jafinch78]

Neat to see the innards of the HP8757A. The older designs are so easy to work on if needed... pending parts availability or DIY knowledge enough to customize suitable. Looking forward to firing up the PMI/Wavetek 1038's this season since I don't have sensors yet for the Gigatronics and HP power meters. Finally scored a set of service and operation manuals for the NS20 and NS206/207... not the NS201's I have three of... however I think enough to read through to boost my confidence enough so I don't damage anything and can better reverse engineer if I wanted to use as signal generator. ESD fear! Guessing like the N10, I'll have a bunch of buttons to replace.

[@herbertsusmann986]

Amazing how all that old test equipment that used to cost $20K+ (in those dollars) is now either $50 or free today. I have a lab full of that kind of stuff, still works. They were built to last from HP and Tek (except HP scopes....). Quality parts through and through. Before the bean counters took over at HP.

[@NimbleJack3]

Only the unusable trash costs $50. There's still many working examples of antique HP tech out there in labs and on the market for hefty sums. This 8757 is relatively worthless due to the unknown performance and self-test fails - it's more money than it's worth for anyone to try and restore it to working order and/or calibrate it, unless you are an enthusiast who is willing to potentially perform component-level repair. Hence why it never sold on craigslist.

[@NimbleJack3]

I advise against ramming little bits of wire into the data inputs. Those things have voltage rails on them, and shorting them to ground or the data lines is a great way to break what's left of the analyser.

[@IMSAIGuy]

Don't worry "I'm a professional'

[@AECRADIO1]

A, B, R, WITH 'R' BEING THE COMBINED INPUT/OUTPUT, POSSIBLY.

[@NimbleJack3]

A and B are the test channels, while R is the reference. The idea is that you can use R to measure the power before it enters the device under test, then measure the power after and display it as an A/R or B/R ratio. This setup could be used to characterise 3-port linear devices such as a power splitter, or to measure both the S11 and S12 of a two-port device at the same time (by using a coupler or bridge).

[@KEVINPAGEkf4znl]

What is a "DUT" and where can I get one for my for my HP 8752A ? Thank You de KF4ZNL

[@IMSAIGuy]

DUT stands for device under test

[@KEVINPAGEkf4znl]

@@IMSAIGuy OH OK Thank you

[@davidr4210]

what is the difference between the 8757 and 8756? i heard the 8756 are a pain to cal and adjust, but are they harder to work with too?

[@IMSAIGuy]

I have no idea, sorry

[@nickcaruso]

Maybe I missed it but why is it scalar? Does it only measure the real part, or one part at a time?

[@NimbleJack3]

It's scalar because the inputs are only for power level data from external detectors (the predecessor to power sensors). They have no way of measuring the phase of a signal, just the instantaneous RF power level moment to moment.

[@nickcaruso]

@@NimbleJack3 thanks!

[@RobB_VK6ES]

Can you use the TG built into your HP SA?

[@IMSAIGuy]

Yes I could probably get it working but it would require sensors and cables (way too expensive)

[@NimbleJack3]

EDIT: I am mostly wrong, see below. No. To my knowledge, the 8757x family is only designed to work with dedicated signal generators that have a compatible "source interface" DB connector (such as the 836x0 sweeper, or its' successor the PSG).

[@IMSAIGuy]

@@NimbleJack3 nope, this one is HPIB

[@NimbleJack3]

@@IMSAIGuy Oops, you are right - total brain fart. Yes, it's a GPIB control, but there's also three auxiliary BNC connectors for sweep, stop sweep, and z-blanking required. I don't think a spec an will be able to work.

[@IMSAIGuy]

@@NimbleJack3 It also required a 27.777khz modulation