PMEL Forum

K Sections => K3/4 - Waveform Analysis & RF Generation => Topic started by: Kyoungmin Kim on 07-28-2014 -- 07:39:10

Title: Attenuator calibration with network anlyzer
Post by: Kyoungmin Kim on 07-28-2014 -- 07:39:10
Anyone has experience attenuator calibration with network analyzer?
I tried to that instead of using measuring receiver. But I am not sure about accuracy of network analyzer and
TAR.
Anyone has experience in that plz let me know.
Title: Re: Attenuator calibration with network anlyzer
Post by: jimmyc on 07-28-2014 -- 13:08:52
what model network analyzer and attenuator?
Title: Re: Attenuator calibration with network anlyzer
Post by: silv3rstr3 on 07-29-2014 -- 06:33:44
I like the SWR function on the 8510C but I don't think it's as accurate as using SWR bridges and a measuring receiver.
Title: Re: Attenuator calibration with network anlyzer
Post by: Kyoungmin Kim on 07-29-2014 -- 08:45:14
Agilent N5230C analyzer and 8496B attenuator.
Title: Re: Attenuator calibration with network anlyzer
Post by: CalLabSolutions on 07-31-2014 -- 23:17:28
We have a software solution that will do that will calibrate that attenuator and many more.

It's called Attenuator SAM.  It is able to calibrate Step, Fixed and Kit attenuators with an VNA.  It will also collect the fill S-Parameters.

See http://www.callabsolutions.com/specialized-applications-for-metrologysam/
Title: Re: Attenuator calibration with network anlyzer
Post by: mysterymeat on 08-11-2014 -- 19:04:08
If Dynamic Accuracy is the right place to be looking... You can most likely get away with using it for some stick attenuators, 10 - 30 dB. But I would not certify an 8496B and claim it as the CAL standard. Depending upon the options your N5230C has, and frequency range, you start losing 4:1 tar around 70 dB, even with the sloppy specs of the 8496B.

http://literature.cdn.keysight.com/litweb/pdf/N5230-90016.pdf
http://literature.cdn.keysight.com/litweb/pdf/08494-90008.pdf
Title: Re: Attenuator calibration with network anlyzer
Post by: CalLabSolutions on 08-12-2014 -- 18:19:51
Accuracy and Cost of hardware is always a trade off.  But when you are using a VNA to calibrate attenuators you don't directly measure each step all the way down to 70 or 110 dB.  Each attenuator is measure separately on only to 40 dB.  All the values below 40dB are then calculated.   
http://www.callabmag.com/?p=1036
There are some limitations, for example you can't measure an HP 355 using this technique.  But compared to other methods its lighting fast.

Also, the accuracies can be improved if you account for the mismatch error between each of the attenuators in side the DUT.  And because of its superb linearity the overall accuracy can be greatly improved if you have the budget for a PNA-X.

Mike