Author Topic: HP 8720A VNA - Advice needed  (Read 2284 times)

0 Members and 1 Guest are viewing this topic.

Offline toastedcrumpetsTopic starter

  • Regular Contributor
  • *
  • Posts: 175
  • Country: gb
HP 8720A VNA - Advice needed
« on: November 11, 2019, 11:23:29 am »
Hi all,
I recently bid and won a HP 8720A in a B2B auction. I got carried away with the bidding and eventually ended up paying £1760 which was more than I intended. The specifications are excellent but I think I don't need most of the frequency range and will struggle to calibrate it up there too. The equipment arrived and appears to be in working order. The test ports are scuffed but the insides seem to be good. It passes the internal tests, but I'm waiting for some new connector savers/semi-flexible cables from Amphenol before I do any real testing.

First question, did I pay too much? This should be a fun question for everyone else to either laugh or cry at. If I've paid too much its too late now.

Second question, what is the underlying connector beneath the test ports? They seem to be female SMA/3.5mm connectors but is behind them a PSC-3.5mm connector? I ask as I'm tempted to change the front connectors to APC-7 as there's a Agilent cal kit I can access to play with at my work and some of the test fixtures I intend to use for materials work have these connectors. I can't find details on the front connectors, although I've not completely read the on-site service manual yet.

Third question, any advice about installing two connectors APC-7->SMA(M) and SMA(F)->APC7, assuming they're a symmetric through, thus calibrating it out, then breaking the "through" connect to my DUT? if i did change over to APC-7 for cal/cables/etc, most of my own hobby designs will probably use SMA hence the want to use adaptors. I realise the VNA probably doesn't have this ability itself, but (Edit: port extensions are available on both ports) One thing I am proficient at is mathematical modelling/programming, so I could develop my own cal model in python and fit it using the GPIB interface. I'm happy to process/plot the VNA data myself rather than use the front panel. Any comments/advice would be welcome.

Finally, is there anything you want a better photo of? I'll even break the magical cal sticker if you want internal shots, but I'm going to probably leave most of it connected until I'm more comfortable with the unit.

Thanks!

More pictures are here: https://photos.app.goo.gl/h9Eotp65HYpVhZxk6

870286-0
« Last Edit: November 11, 2019, 02:01:03 pm by toastedcrumpets »
 

Offline hendorog

  • Super Contributor
  • ***
  • Posts: 1635
  • Country: nz
Re: HP 8720A VNA - Advice needed
« Reply #1 on: November 12, 2019, 05:54:26 am »
Nice find, I'd love one of those. I found one (8720A) which went on eBay a few weeks ago by auction for $1700 USD - in the US, so you didn't do too bad :)

Maybe the connectors are serviceable - i.e. thats why they appear to be unscrewable? It will be in the service manual I expect. I'd leave them alone :)

Regarding the SMA. I think there are phase matched APC-7 to 3.5mm adapters for doing what you mentioned, that come with a cal kit.
Lacking those adapters there will be some error obviously if the adapters are not symmetrical, but it might not matter, or you might be able to correct for it, depending on what you are doing.

I don't think you can 'properly' mathematically remove the effect of the adapter unless you have both an APC-7 cal kit and an SMA/3.5mm cal kit, but there might be other approaches, and near enough might be good enough....


 
The following users thanked this post: toastedcrumpets

Offline TheSteve

  • Supporter
  • ****
  • Posts: 3785
  • Country: ca
  • Living the Dream
Re: HP 8720A VNA - Advice needed
« Reply #2 on: November 12, 2019, 06:59:21 am »
Price wasn't unreasonable.

I wouldn't mess with the factory 3.5 mm connectors - just buy cables to connect to them and adapters to what you need.

What frequency range do you intend to use? There are deals at times on decent calibration kits, you just need to keeps your eyes open and be ready jump on a good deal. If you intend to use SMA mainly then I think you'd want a 3.5mm or SMA calibration kit, skip APC7 altogether.
VE7FM
 
The following users thanked this post: toastedcrumpets

Offline toastedcrumpetsTopic starter

  • Regular Contributor
  • *
  • Posts: 175
  • Country: gb
Re: HP 8720A VNA - Advice needed
« Reply #3 on: November 12, 2019, 08:36:45 am »
Thanks for the replies so far.

I agree, I should leave the connectors alone, I will only mess with them if I absolutely need to. If anyone else is looking for information on the front panel connectors, the on-site service manual tells you how to remove them (it almost only says unscrew them, gold part first, then the actual test port) but gives no information on the actual interface. I think I've found an incorrectly listed ebay item (its not APC-7) that is close to my existing front panel test ports except at the VNA side (which is the bit I'm wondering about)
https://www.ebay.co.uk/itm/APC-7-DC-18-GHz-26-5-40-GHz-RF-COAXIAL-ADAPTOR-C-3-1-34/183610027276?hash=item2ac002ad0c:g:FbAAAOSwtutcJ02z
The actual details on the front panel kit are in the "HP 8720A System Accessories Manual" which is a binder which contains the "Connector Care Manual" which might have more information. If anyone can find this I'd appreciate it! Its not scanned on keysight.com.

Regarding frequency range, at the moment I only need up to about 6GHz (most of the materials testing on liquid crystals is below that) so I'm currently thinking about getting a Kirkby Microwave 85033 calibration kit. They look very well put together, properly characterised, and very cost effective. Almost 1/3rd of the price of the VNA, but quality costs money and its new.

I've got some ebay saved searches on the go for the 85033D, 85052B, 85052C, and 85052D calibration kits. Many are without the constants on disk and of unknown condition, so its a bit chicken and egg. If I had a known good cal kit like the 85033 I could evaluate somewhat the second hand stuff but I don't have one.
 

Offline TheSteve

  • Supporter
  • ****
  • Posts: 3785
  • Country: ca
  • Living the Dream
Re: HP 8720A VNA - Advice needed
« Reply #4 on: November 12, 2019, 05:30:36 pm »
If you do get a 85033D/E or 85052x they all use the same constants, so you should be able to transfer them to the VNA if it doesn't already have them stored internally(They all do, unless it predates all of those kits).
VE7FM
 
The following users thanked this post: toastedcrumpets

Offline hendorog

  • Super Contributor
  • ***
  • Posts: 1635
  • Country: nz
Re: HP 8720A VNA - Advice needed
« Reply #5 on: November 13, 2019, 01:57:07 am »
If you do get a Kirkby kit then its a good idea to request the s-param measurement files for all of the parts in it. I am not sure if that data is provided by default.

It's much easier to use that instead of the cal kit coeff's when doing the correction calcs offline on a PC.
 
The following users thanked this post: toastedcrumpets

Offline toastedcrumpetsTopic starter

  • Regular Contributor
  • *
  • Posts: 175
  • Country: gb
Re: HP 8720A VNA - Advice needed
« Reply #6 on: November 13, 2019, 08:29:43 am »
Thanks for the reminder hendorog! Funnily enough I just read your username while reading the changelog for the SOLT parameter fitting Octave file from in3otd. What a great resource from you guys (https://www.qsl.net/in3otd/electronics/VNA_calkit/calkit.html) and Tudor Dima (https://uk.mathworks.com/matlabcentral/fileexchange/6080-s-parameter-toolbox-z-y-h-g-abcd-t)

TheSteve, thanks for confirming the HP/Agilent/Keysight kits are made to a standard. The 8720A does have predefined cal kits as well as a user cal kit which can be based on any of the predefined, so no worries there.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf