VNA calibration
How to do a sma 3.5 calibration with that? My DUT is 3.5
Just cal using the 2.4mm and use a small low loss adapter that will allow you to connect to your device under test.
You will need an adapter that will take you from 2.4mm to 3.5mm. Make sure you get the correct gender. (m) or (f)
How to make sure that the adapter is good enough? What is the effect of using adapter with the result of measurement?
This will depend on the frequency of your measurement. (higher electrical length) Assuming your adapter has very low loss (0.02dB) it will only present a slight phase offset to your measurements. So if you need to measure accurate impedance values you would need to compensate for it. One easy way is to export the s2p file to a simulator and place a loss-less coaxial line at both the input and output ports of the data-file and enter the length of your adapter as a negative value on both ends (assuming the measurement is a full 2-port, S11 only would only require one correction). Some VNAs have an option to adjust electrical delay in the display menu. Do this with the adapters in place on the smith chart or polar format and adjust the phase until your unconnected and calibrated cable is fully on the right side of the chart with no positive or negative reactive impedance. Also some newer VNAs have a function called Port-extensions that will do this for you.
If the adapter is really short you could ignore the error and still get usable results, assuming you are not measuring in the high GHz range.
EDIT: When using electrical delay offset or port extensions, double check to make sure that any exported data-files (touchstone(s2P) or citifiles) have the correction in them. They often are not included so the phase corrections you apply are usually just for display purposes. These methods only work if your adapter is very low loss.
If the adapter is already include in calibration, should we need to compensate it?
That would depend on how you use the adapters and I don't really know your entire situation.
This would depend on if you are using 2.4mm cables. If you are, then you would calibrate to the e-cal module without adapters. Then to measure your device you would then use the adapters to interface to your device. In this case you would have to do what I outlined earlier.
If you are using 3.5mm cables and then use the 3.5mm to 2.4mm adapter to mate to the e-cal module for calibration. Then you will need to remove the adapter for measuring your device. In this case you will need to add electrical length to the input and output ports in order to maintain your calibration reference plane.
The concept of reference plane is the key to understand how to deal with calibration corrections. Always think about what your analyzer sees as your calibrated point then any additional connectors added or removed after calibration has occurred will require correction.
What I mean is this:
I want to ask you about this: "Then you will need to remove the adapter for measuring your device." What does it means? Why it must remove the adapter?
You also said: "you will need to add electrical length to the input and output ports in order to maintain your calibration reference plane." How to do that?
The adapter is needed for the sole purpose of mating your cabling to either the e-cal module or the device you are testing. The adapter can only help you do one of these things. I'm assuming you have either 2.4mm cables or 3.5mm cables. Which do you have?
How to add or remove electrical length?:
http://www.vnahelp.com/tip9.html
What VNA model are you using? with this I can help you step by step.
My VNA is E5062A from Agilent.
User manual can be obtained here:http://cp.literature.agilent.com/lit...5061-90060.pdf
go to page 209 and it lists the following: (Ive bolded the 3 options that should help you)
"Improving the Accuracy of Phase Measurements
This section describes the following functions that can be used to improve phase
measurement accuracy.
• “Electrical Delay” on page 209
• “Port Extension” on page 210
• “Phase Offset” on page 211
Electrical Delay
Electrical Delay is a function that adds or removes a pseudo-lossless transmission line with
a variable length against the receiver input. Using this function enables you to improve the
resolution in phase measurement and thereby measure deviation from the linear phase. You
can specify the electrical delay trace by trace."
Edit: I hope this helps, I've got to run for now. Good luck!
also check out this Agilent discussion forum:
http://forums.tm.agilent.com/communi...2003a53d60c4d4