how to design match standard in TRLM cal kit

[snafflekid]

As a followup to my previous thread on calibration standards:
https://www.edaboard.com/threads/274642/

I think that I can use the two traces on my board to use for THRU and REFLECT standards if I also create a MATCH standard.
What are the design requirements for the layout of a MATCH standard? There are some examples of LOAD standards on websites that show a microstrip with the end terminated in two parallel 100 ohm resistors.

My primary questions are: does the length of the MATCH microstrip matter? Would a screw-on 50 ohm terminator be adequate?

Picture of the board: View attachment 84584

this board is made of FR370HR

 

[drkirkby]

I don't know, but can make two suggestions.

1) Ask on the Agilent VNA forums. Many Agilent staff answer questions

2) https://www.amazon.com/Handbook-Microwave-Component-Measurements-Techniques/dp/1119979552 written by Dr. Joel Dunsmore, has a long chapter about making non-coaxial measurements on VNAs, shows PCB layouts, and descibes the standards used in quite some detail.

Someone on the Agilent forums with the user name "Dr_Joel" is very helpful, and answers questions very regularly. He is the author of that book.

BTW, I noticed on your other post you said you were going to hire a VNA which had the ability to bias the ports. I'm pretty sure my 10+ year old HP 8720D (50 MHz to 20 GHz) can do that, but I'm not 100% sure and can't be bothered to go down stairs and look on the back.

But given your VNA does not, would it not been cheaper / less hassle to buy a couple of bias-T's? If you calibrate with your standards on the common output with no DC voltage on the bias input, then apply DC, I can't see why anything would change. Obviously you are not going to be able to put a short across the output when there is DC applied. I think the smoke might come out of the bias-T.

Dave

 

[snafflekid]

Hello Dave,

Thanks for pointing me towards the Agilent forums. There is good advice on the VNA forum.

I did not realize that bias tees could be added on externally. But it makes sense that the tee can be a simple adapter screwed to the test port.
I have to buy a new VNA and I was adding tees as an option.
Maybe I should keep the tee external so that I have the choice to add or remove it!

Every answer creates a new question

 

[drkirkby]

Hello Dave,

Thanks for pointing me towards the Agilent forums. There is good advice on the VNA forum.

The people there are very helpful. The support from the forums that I have received has been excellent both on a new N9923A I purchased, and an obsolete 8720D I bought used. I'm quite often on there - the same user name as here. If buying another VNA, I would not personally consider any other than Agilent, since they make excellent VNAs and the support is excellent.

I did not realize that bias tees could be added on externally. But it makes sense that the tee can be a simple adapter screwed to the test port.

I can't see any reason why they could not be added externally. For what it is worth, I just checked the data sheet on my 8720D, and it does come with them built in. It is not an option. I think there are a couple of fuses on the back to protect the bias-T's.

I have to buy a new VNA and I was adding tees as an option.
Maybe I should keep the tee external so that I have the choice to add or remove it!

If I had a use for the bias T's, I'd probably just add the option for convenience, but I'm pretty sure you could add them externally. You might find the hardware is actually in there, but enabling them requires a software option. I know for the Agilent N9923A I bought, the options, which were all software ones, cost more than the basic VNA!

You might want to think about adding the TDR option - normally 010 on Agilent VNAs. IIRC, the section in Joel's book that deals with making non coaxial measurements, which is what you are doing, makes extensive use of the TDR option. There's a lot more to that option than just a simple inverse Fourier Transform of the frequency domain data. However, the data is derrived from an IFT - there is no pulse generator.

You might find the TRL calibration is an option, as that needs a 4 channel receiver. It is an option on mine (option 400), but my VNA does not have it fitted.

Buying options needs some serious consideration, especially if they can't be added later.

Dave