Contact an Expert

Thread: Multiple calibrations


Permlink Replies: 8 - Pages: 1 - Last Post: Jun 13, 2012 12:29 PM Last Post By: daras Threads: [ Previous | Next ]
nsm

Posts: 40
Registered: 02/20/12
Multiple calibrations
Posted: Jun 12, 2012 10:39 AM
Click to report abuse...   Click to reply to this thread Reply
Hi,

I am using a pulsed PNA-C with the Z5623A H81. I need to do return loss measurements of both the ports of the amplifier that I am measuring. I am doing the measurement through a switch matrix. All the measurements are done in an automated manner.

1. I first do an enhanced response cal and then save the cal as a .csa file.
2. Then I switch the matrix so that during measurement, I reverse the device. Then I do a 1 port cal.Save this as a .csa file.(A different file name than the previous one)
3. I find that when I do this, the enhanced response cal goes awry. What I mean by this is that, when I do the measurements, I do not get the same response that I get when I do only the enhanced response cal. I am wondering if the enhanced cal gets overwritten by the 1 port cal.

rgds
daras


Posts: 905
Registered: 12/01/05
Re: Multiple calibrations
Posted: Jun 12, 2012 10:49 AM   in response to: nsm in response to: nsm
Click to report abuse...   Click to reply to this thread Reply
An Enhanced Response cal and 1-port cal share the same error terms for input port:

Directivity
SourceMatch
ReflectionTracking

Because of this, if you have an Enhanced Response correction in a calset and then perform a 1-port cal, the Enhanced response cal will be over-written. The opposite is also true. Therefore these two different correction types can not share the same calset.
nsm

Posts: 40
Registered: 02/20/12
Re: Multiple calibrations
Posted: Jun 12, 2012 10:53 AM   in response to: daras in response to: daras
Click to report abuse...   Click to reply to this thread Reply
Hi Daras,

Thank you for the response. After I finish the enhanced response cal, I save the .csa file. Then I preset the PNA and then do a new 1 port cal and save the file. The preset is also not helping. How do I overcome this problem?

rgds
daras


Posts: 905
Registered: 12/01/05
Re: Multiple calibrations
Posted: Jun 12, 2012 11:46 AM   in response to: nsm in response to: nsm
Click to report abuse...   Click to reply to this thread Reply
Ok - maybe I am not understanding the exact sequence of events here. if you do an enhanced response cal and then save a .csa file, preset the box, do a 1-port cal and then save it into a different .csa file, then when you make the measurement, the correction that will be applied will be based on the last state that was recalled. so if you want to make a measurement and apply the enhanced response correction to it, then recall the first .csa file that you save, connect the DUT and take a sweep. now if you want to make a measurement and apply the 1-port calibration, recall the 2nd .csa file you saved, connect the DUT and take a sweep.

I am betting that you are using the default save option at the end of the Calibration wizard, which saves the calibration in the channel cal register. Also I am betting that for both the enhanced response cal setup and the 1-port cal setup you are using the same channel (probably channel 1?) so in both cases the calibrations are saved into Channel 1's cal register and even though you save two different .csa file, each time one is recalled, it overwrites the contents of the channel 1 cal register. so if you want to make an enhanced response measurement and a 1-port measurement in the same state (without recalling a different state in between sweeping the measurements), then you need to setup 2 different channels for each of the measurements, so that both corrections can coexist in the same state.
nsm

Posts: 40
Registered: 02/20/12
Re: Multiple calibrations
Posted: Jun 12, 2012 5:52 PM   in response to: daras in response to: daras
Click to report abuse...   Click to reply to this thread Reply
Hi Daras,

The sequence of calibration is as follows.As you have pointed out we are using Channel 1 for both the measurements.

1. Set the PNA for the frequency of operation of the DUT.
2. Do an enhanced response cal.
3. Save the state and cal file as enhanced_resp.csa
4. Preset the PNA.
5.Set the PNA for the frequency of operation of the DUT.
6.Do a 1 port calibration.
7. Save the state and cal file as 1port_resp.csa.

Now for the measurements.

1. I recall the enhanced_resp.csa for the measurement. I find that this cal is not the same as the enhanced_resp.csa. It is being overwritten by the 1port cal. The preset on the PNA does not help to save the cal and state as a different file?

DO we have to create a separate channel for bot the cals to co-esist?

Rgds
daras


Posts: 905
Registered: 12/01/05
Re: Multiple calibrations
Posted: Jun 12, 2012 6:12 PM   in response to: nsm in response to: nsm
Click to report abuse...   Click to reply to this thread Reply
nsm wrote:
1. I recall the enhanced_resp.csa for the measurement. I find that this cal is not the same as the enhanced_resp.csa. It is being overwritten by the 1port cal. The preset on the PNA does not help to save the cal and state as a different file?

this seems a bit odd. if you are recalling the .csa file it should restore the channel cal register to whatever was in the channel 1 cal register when you saved the .csa file. it should not matter what other calibrations you have performed in that channel since you saved the .csa file. it should over-write anything that is already there. can you post both your enhanced_resp.csa and 1Port_resp.csa files and also note what firmware revision you are using?
nsm

Posts: 40
Registered: 02/20/12
Re: Multiple calibrations
Posted: Jun 12, 2012 10:57 PM   in response to: daras in response to: daras
Click to report abuse...   Click to reply to this thread Reply
Attachment Rx_June6_F1P_T1.csa (1,000.4 KB)
Attachment Rx_June6_F2P_T1.csa (1,000.3 KB)
Hi Daras,

I am using the PNA E8362C firmware revision A.09.33.09. Please find the files attached. The F1P is aone port call and the F2P is the enhanced response cal.

Rgds
nsm
Dr_joel


Posts: 2,706
Registered: 12/01/05
Re: Multiple calibrations
Posted: Jun 13, 2012 12:22 PM   in response to: nsm in response to: nsm
Click to report abuse...   Click to reply to this thread Reply
While daras sorts out the recall stuff, might I recommend that you save each as a named user calset, such as nsm_oneport nad nsm_enhanced; in that way they are explicitly saved independetly. I almost never use the default cal register as its is easily corrupted by other calibrations and it's hard to know when it gets corrupted; by corrupted I mean that you can do a 1 port cal, then a response cal, and then a 1 port cal on the other port and it's all ok, then if you do an enhaced response, the 1 port and respone cal are overwritten, but the old 1-port cal on port 2 is left around (it's intentional, but it can also get confusing; it's intentional so you can build up a lot of different calibratoins such as response and 1 ports, on each port, one at a time, but using the guied power cal eliminates the need for this now).

So, save to a user cal set and see if that fixes everything. Also, I think you can use up to 9.42 on the C versions, and I know of lot of things are fixed in that version; and that will be the final major branch for C versions as we just introduced a new version of the PNA-L:

http://www.home.agilent.com/sec/product.jspx?nid=-536902643.1017780.00&pselect=SR.Looking&pageMode=OV&lc=eng&cc=US

oops, I just saw that it is a PNA you are using; anyway, you can use 9.42 on that as well.

Edited by: Dr_joel on Jun 13, 2012 11:24 AM
daras


Posts: 905
Registered: 12/01/05
Re: Multiple calibrations
Posted: Jun 13, 2012 12:29 PM   in response to: nsm in response to: nsm
Click to report abuse...   Click to reply to this thread Reply
Hi nsm,

I recalled these files on an E8364C PNA running A.09.33.09 version of firmware and it behaved as expected. when I recall the F2P file, the S21 measurement is corrected with an Enhanced Response correction (you can confirm by looking at the status bar - it should say "C Enh Resp"). And when I recall the F1P file, the S11 measurement is corrected with a 1-port correction (Status Bar says "C 1-Port"). So basically, I am not observing this issue of an improperly overridden calset. Since we can't recreate the issue (and part of the problem might be miscommunication), I would recommend that you call the Agilent support hotline or contact your local Agilent rep, so that we can provide better hands-on, face to face support.

Point your RSS reader here for a feed of the latest messages in all forums