I have been making new cal kit files for non-Keysight cal kits by editing existing files with the Cal Kit Editor. I've noticed that the Max Frequency in all the cal kits I looked at is 999000 MHz even though the specified maximum frequency of the cal kit is something less.
For example, the 85032B is a 6 GHz cal kit but the factory cal kit file shows the Max Frequency is 999000 MHz. Why is it not 6000 MHz?
Can someone explain this?
Thanks,
Whit
Yes, I have noticed that too. Where there are very good reasons for a reduced frequency range, they are reduced. For example, an X11644A WR90 waveguide cal kit is only specified to cover 8.2 to 12.4 GHz. Also, if cal kits (eg 85054 N kit) have sliding loads, the lowband loads have a maximum frequency of 2.001 GHz, and the sliding loads a minimum frequency of 1.999 GHz.
My company, Kirkby Microwave
Low cost Vector Network Analyzer (VNA) Calibration kits
sells low-cost VNA calibration kits. We always set the frequency range of the kits files properly, and don't use this 999 GHz stupidity. This does stop one using a kit above its frequency range, where accuracy would be undefined.
We have had lengthy internal discussions about this. Here is a summary. Quite a few mechanical calibration kits work beyond the specified frequency range and our customers use them beyond the range. For example our 3.5 mm cal kits are specified to 26.5 GHz but work well to 33 GHz. The initial 26.5 GHz limitation came from the VNAs themselves. We couldn't test them higher, so the cal kit specs became 26.5 GHz. This is the case for a few other connectors also.
So then in order to not limit customers initially, someone put 999 GHz in there. Now obviously 999 GHz is not correct, but if you put in 33 GHz, then users may think the cal kit is specified to 33 GHz, but it is only specified to 26.5 GHz. 999 GHz is an obvious number that says, it's just an arbitrary max. What we could do is add in a setting that users can check saying over-range allowed, or agree on a specific percentage of over-range. Both are possible improvements on our list of actions.
And lastly, HP/Agilent/Keysight cal kits are so widely used with a lot VNAs, some older, some newer, some in systems, etc, that we do not make changes lightly. We have to weigh the benefits versus the costs (of someone's software breaking).
Afsi Moaveni
Santa Rosa, California
Thanks again for your help, Afsi. I suspected there was a historical limitation somewhere along the line but rather than go with speculation I thought I'd get the real answer. Everything you say makes sense.
Whit
I sometimes use my 18 GHz 85054B cal kit on a 3 GHz 8753ES. To fully conform to the specifications of the 85054B, I should use the lowband load to 2 GHz, then use the sliding load between 2 & 3 GHz. In practice I don’t bother, as I know that the lowband load works fine up to 3 GHz, and it is too much effort to use a sliding load when a fixed load will do.
I have measured the properties of the 2 GHz lowband loads in the 85054B by using the sliding loads and a 20 GHz 8720D. I don’t recall the numbers, but the 2 GHz loads makes pretty decent load to at least 10 GHz. So when I use that kit on the 3 GHz 8753ES, I never bother with the sliding loads.
Dave.
We have had lengthy internal discussions about this. Here is a summary. Quite a few mechanical calibration kits work beyond the specified frequency range and our customers use them beyond the range. For example our 3.5 mm cal kits are specified to 26.5 GHz but work well to 33 GHz. The initial 26.5 GHz limitation came from the VNAs themselves. We couldn't test them higher, so the cal kit specs became 26.5 GHz. This is the case for a few other connectors also.
So then in order to not limit customers initially, someone put 999 GHz in there. Now obviously 999 GHz is not correct, but if you put in 33 GHz, then users may think the cal kit is specified to 33 GHz, but it is only specified to 26.5 GHz. 999 GHz is an obvious number that says, it's just an arbitrary max. What we could do is add in a setting that users can check saying over-range allowed, or agree on a specific percentage of over-range. Both are possible improvements on our list of actions.
And lastly, HP/Agilent/Keysight cal kits are so widely used with a lot VNAs, some older, some newer, some in systems, etc, that we do not make changes lightly. We have to weigh the benefits versus the costs (of someone's software breaking).
Afsi Moaveni
Santa Rosa, California