开云体育

Have a HP 8485A sensor I bought used, turns out it had been damaged due to overload, seller refunded, now have a physically nice but electrically dead HP 8485A sensor.

Have a HP 8481A sensor surplus to my needs.

I'm interested in the possibility of converting my 8481A to an 8485A (3.5mm male connector), albeit not calibrated, in lieu of the cost of buying a working 8485A.

Found a Keysight publication that seems to be quite comprehensive coverage of this era sensor:



This is titled "Keysight 8480 series coaxial power sensors operating and service manual." I wasn't aware the manual existed; it is current publication from Keysight.

The manual has pretty extensive "how to disassemble and reassemble" and other maintenance activities.

The underlying thermocouple sensor element appears to be the same part number for 8481A and 8485A, so this brings me to my question - could I create a thermocouple sensor with a 3.5mm HP male connector that could subsequently be recalibrated and used as an 8485A?

I've searched and haven't found any reports of the practicality or result of actually trying this procedure.

Thanks for any comments, observations, or requests for clarification.

Thank you for detail understanding!

"F1EKU via groups.io" <rfconsulting.fr@...> writes:

From experience, the 8481A series does not perform up well beyond 18
GHz. The response drops very quickly towards 19/20
GHz, and at 26.5 GHz, the response difference is more than 3 to 4
dB. The N connector is not responsible for this poor
performance, as 85025A detectors easily perform up to 40 GHz with only
a few glitches due to VSWR beyond 18.6 GHz.
Conversely, an 8485A often performs up to 30 or even 33 GHz.

I can confirm that. Last week i tried to do flatness calibration on my
83752A sweeper. As i only have a 8485D that accepts a max input level
of -20dBm, i tried the 8481A instead. But exactly as you said, the
levels where low by at least 3dB above 19GHz.

After doing some testing yesterday, I discovered that the 8481A that was to be the recipient of the dead 8485A's connector etc. assembly was itself dead (8481A won't zero and responds about 20 dB down incident signal, so dynamic range is something like 20 dB, not 50 dB.) I don't immediately recall the 8481A's history, however, I thought I had stored it away in working condition (and I'm the only person that would have access to it.) Given that unexpected result, and the current-era prices for used 8481A, my project idea won't be able proceed. So all that said, I'm going to keep eye out for 8485A (basically as a reference against which to calibrate 8673D synthesized signal generator.)

Thanks again for the contributions to the question, it is very much helpful.