HP 85033E 3.5 mm calibration kit on 8720D VNA

I don't suppose this is the best place to ask, but for some reason I can't see to post to the Agilent forums. I have asked Agilent for clarification on this, but I'm hoping someone here might just know.

I have an HP 8720D VNA (20 GHz)
http://www.home.agilent.com/agilent/product.jspx?pn=8720D

and want to know if I've made an expensive mistake in buying an Agilent 85033E 3.5 mm calibration kit
http://www.home.agilent.com/agilent/product.jspx?nid=-536902695.536879921.00&lc=eng&cc=US
or not.

I was looking for a cal kit, looked in the calibration menus, and see the 85033E is listed as one of the calibration kits I can use, as you can see in this photo



(There is no such calibration kit as an 85033D/E, but a 85033D and an 85033E).

So I see an 85033E on eBay and bought it.

Today the cal kit arrived, and I happened to look on the Agilent web site, which indicates this cal kit is not supported with my VNA.

1) "Agilent 8720 Kits and Accessories" lists the acccessories for this VNA
http://na.tm.agilent.com/8720/kits.htm
but has no mention of the 85033E calibration kit.

2) At the page showing the calibration kit definitons for a variety of Agilent cal kits
http://na.tm.agilent.com/pna/caldefs/stddefs.html
it clearly shows on the 5th row down that the 872xD is not supported with the 85033E calibration kit, which would indicate I've just made an expensive mistake.

Does anyone have any idea what is going on here?

Dave

If you VNA has 3.5mm connectors, your calibration kit will obviously support your 8720D up to 9GHz.
Because 3.5mm connector is a industry standard and it can be used with any VNA brand.I think cal kit is not supported with recent models which have higher frequency of mesurements,that's why it's not been mentioned.

BigBoss said:

If you VNA has 3.5mm connectors, your calibration kit will obviously support your 8720D up to 9GHz.
Because 3.5mm connector is a industry standard and it can be used with any VNA brand.I think cal kit is not supported with recent models which have higher frequency of mesurements,that's why it's not been mentioned.

Click to expand...

I think you are over-simplying this.

Just because a cal kit has the right connectors and a suitable frequency range, does not mean the analyer will support it. The analyser needs to have data on the properties of the open and short. Unless HP publish the coefficients, or you measure them some way, you can't just use any cal kit.

Getting an Anritsu cal kit would not be much use, even if it had the right connectors.

On this analyser there is a choice of several N, 3.5 mm and APC-7 cal kits. (See the photo I posted). If you pick the wrong cal kit, the analyer will not calibrate properly. There will be small errors, since the distances to the open and short are different on different cal kits, even though they have the same connectors.

You can enter the data on the cal kit manually, or perhaps even download it vai GPIB. But the coefficients are not published on all cal kits.

Dave

You might find the comments on www.microwaves101.com about How to (not) trash a cal kit interesting. Hopefully they will not apply to the one you purchased.


Regards,

Azulykit

did you get the 4 port option or 3

drkirkby said:

Just because a cal kit has the right connectors and a suitable frequency range, does not mean the analyer will support it. The analyser needs to have data on the properties of the open and short. Unless HP publish the coefficients, or you measure them some way, you can't just use any cal kit.

Click to expand...

Dave, for the stuff explicitely sold as a cal kit, the electrical parameters should be available. That's the only reasons why people buy thess expensive cal kits, instead of individual open (trival) and short (trivial) and load.

If the cal kit is not listed as supported, this might be because it does not support the full frequency range or the cal standard definition data is not available on the required media. However, the data itself should be available in some format, so that you can enter it manually.

I managed to get to the Agilent forums in the end.

What I have found is there are huge discrepancies in the list of kits supported by my VNA and those which the firmware has support for. There are kits list as "unsupported" yet are in my firmware, and there are kits listed as "in F/W" but they are not in my firmware.

I need to work out the best way to get the data on some kits. If you look at some kits, the definitions in the files are different for different VNAs.

Take for example the 85032F 9 GHz N kit:
http://na.tm.agilent.com/pna/caldefs/stddefs.html

This is supposed to be unsupported in my VNA, yet it is in the firmware. But there are two files one can download for other VNAs. Those two files have different values for the same cal kit.

For the PNA series of VNAs using the 85032F 9 GHz N kit:
**broken link removed**

and for the 8753D (3 GHz VNA) using the same 85032F 9 GHz N kit:
http://na.tm.agilent.com/pna/caldefs/87xx/D53N5F.htm

so that is the same kit, having different values on different VNAs. For the 8753D, both opens, have the same C0 term of 89.939 fF. But for the PNA series, the value of C0 is 94.9227 fF for the male open, and 101.3493 fF for the female open.

From what I can gather a polynomial is fitted to data to give the best values of the capacitance parameters. But if the VNA has a more restricted frequency range (such as a 3 GHz VNA) on a cal kit that's usable to 9 GHz, then the optimal coefficients are different, since you can obviously fit a curve more closely using a polynomial if you restrict the range of the fit to 3 GHz rather than worry about 9 GHz.

I've asked Agient to clarify exactly what the status is with the 85033E 3.5 mm kit is in my VNA. But me and the Agilent guy think it should be ok since its in the firmware. But he is going to get it clarified for me.

I've bought 85032B 6 GHz N cal kit, which is supposed to be supported in firmware, but it is not. I'd like Agilent to clarify what the values I should be using are. I suspect I should take them from the modern PNA series data rather than from the older 3 GHz VNA.

I don't know if the only difference between the B (6 GHz) and F (9 GHz) suffixes of the 85032 are the loads. If so, I suspect using the 85032F data which is in my firmware will be ok. It might not be optimal, since the data for the 85032F in the firmware is fitted over 9 GHz, when the loads in the 85032B kit are only 6 GHz.

Perhaps, changing a couple of loads in the 6 GHz 85032B kit to 9 GHz loads might allow it to work to 9 GHz and me use the data in firmware for the 85032F 9 GHz kit. I'm hoping to get the differences between the 85032B and 85032F kits clarified, since if is only the loads, there's a cheap and easy method to get it to work to 9 GHz using firmware already in my VNA.

I believe I can only have one user calibration set in my VNA, which is annoying. So if you use a mixture of connectors, it would become rather tedious to load a different set of cal data every time one wants to use another connector.

To me at least, this is a rather confusing topic, though I think I am slowly getting my head around it!

Dave

- - - Updated - - -

Azulykit said:

You might find the comments on www.microwaves101.com about How to (not) trash a cal kit interesting. Hopefully they will not apply to the one you purchased.


Regards,

Azulykit

Click to expand...

Yes, I had seen that. As far as I can see, they are ok. But I am going to take some photos later today with real Nikon macro lens (not a normal lens with a macro mode) and investigate more carefully. But as far as I can tell, they are ok.

The cal kit has a recent calibration certificate, but I'm a bit weary of what values these cal certificates actually have.

I managed to pick up an 18 GHz N verification kit fairly cheaply (about 10% of the new price) and will send that to Agilent for calibration, as it is 20 years old, though it look unused.

I should be in a position very soon to have both N (6 GHz) and 3.5 mm (9 GHz) cal kits. Once I have those, I'll look around more for better (18 and 26.5 Ghz) cal kits, and sell off the ones I have. For now at least, 6 Ghz is all I need, so there is no hurry to go to the higher frequencies, but I will need them later, so I will try to acquire kits which go to the higher frequencies.

Dave