Network socket timeout error message when trying to use Thermo EDS with Win 10

[emma_fisi]

Hi All,

I have recently (been forced to) upgrade my computers to Windows 10. I run PfE and the Thermo Pathfinder on different computers. Previously, they talked to each other just fine. Now, PfE will talk to Pathfinder to collect combined EDS and WDS data, but when I try to look at the data in the Analyze window, I get the following error message:  Error number 5 returned from Thermo EDS Interface (Socket timeout error) when calling function TEVB_GetNetIntensities".

Does anyone out there know what Error number 5 is...?

If I copy the MDB file to an offline computer that has both PfE and Pathfinder on it, I don't get the error message, and PfE will process the EDS intensities, but this is not always practical....

Thanks,
Em

[Probeman]

Hi Emma,
I moved this to the Thermo board as maybe Steve Seddio can respond. 

Error 5 is a Thermo NSS/Pathfinder socket timeout error.  I've seen this error when I had some spectra with high deadtimes. It might also be from a Win 10 network timing issue. 

But as far as net intensities go, just to speed things up I now utilize a recently added feature (Only Use EDS Stored EDS Net Intensities option in the Analysis Options dialog) by first putting PFE into off-line mode, then selecting all the samples from the Analyze! window, then clicking the Raw Data button. This extracts all the net intensities and these net intensities are now automatically saved to internal tables in the MDB file.

Then using this new Only Use EDS Stored EDS Net Intensities option, you can now only utilize these internally stored EDS net intensities for quantitative analysis, and everything runs *much* faster:

https://probesoftware.com/smf/index.php?topic=40.msg8637#msg8637

This way one can take their probe MDB file and not have to connect to the Thermo software for net intensity calculations. Makes everything much more portable. This feature works for Thermo, Bruker and even JEOL EDS spectra.

Of course, when you add new elements (or new samples) for quantification by EDS, you'll need to turn this flag off and re-connect to Thermo to extract the new net intensities.  Then you can turn on this "Only Use EDS Stored EDS Net Intensities" Option again.

So until you get this network timeout issue fixed, you can extract your net intensities on the other computer and bring it back over to the networked computer and then turn on the Only Use EDS Stored EDS Net Intensities option so it doesn't call the network interface.

[emma_fisi]

I've updated PfE, and I see the new option under Analysis Options - I like it! Unfortunately, I still get the same Timeout error whenever PfE is even trying to read the raw intensities from Pathfinder....

[Probeman]

Hi Emma,
Yeah.  I'm not trying to fix that, you'll have to talk to Thermo about getting that dealt with.  What I'm suggesting is that you do the net intensity extractions on the off-line computer which is working you said.  Then use this flag once that's done.

I'm just trying to get you going until you and Thermo figure out the Pathfinder/network time out issue. 
john

[emma_fisi]

Thanks! We'll definitely make use of the new flag to speed up our analyses on the offline computer

I'll pester Steve Seddio about the Thermo issue. I know how much he loves getting questions at the end of the day/right before the holidays 

[Probeman]

No worries. Just remember, once you get your network issues fixed you can also extract the net intensities on your on-line computer and use this new feature on that computer also.

[John Donovan]

So it seems that several users are running into Thermo Portal timeout issues when performing long runs and acquiring EDS spectra during the automation.  These timeout errors from the Thermo Portal interface seem to occur when obtaining the EDS net intensities during the automated acquisition and there are quantitative elements by EDS declared in the Elements/cations dialog.

Thermo is going to fix these timeout issues *real soon now*, but in the meantime there is a workaround that I posted here:

https://probesoftware.com/smf/index.php?topic=42.msg8432#msg8432

which basically involves responding to the question shown in the above linked post: "No". That is, do not calculate EDS net intensities for elements by EDS during the automated acquisitions. Note that this feature applies to all EDS interfaces, including Thermo, Bruker and JEOL.

The original reasoning for prompting the user with this question prior to launching the automation, was simply to speed up the automated acquisitions, since the net intensity calculations can take several seconds for each data point.  And there's no real point in displaying the EDS net intensities if there's no one there to see them! But if one is getting Thermo Portal timeout issues during long automated acquisitions, turning off the net intensity calculations during automation, will probably help.

On a related topic, to speed up the calculation of the EDS net intensities, once the automation has completed, please see this post:

https://probesoftware.com/smf/index.php?topic=40.msg8637#msg8637

Note that this saving and utilizing stored EDS net intensities feature also applies to all EDS interfaces, including Thermo, Bruker and JEOL.

[John Donovan]

I've updated PfE, and I see the new option under Analysis Options - I like it! Unfortunately, I still get the same Timeout error whenever PfE is even trying to read the raw intensities from Pathfinder....

The latest Probe for EPMA has modified error handling to better dead with socket timeout and socket write errors from PathFinder when the TEVB_EdsAcqRunning function is called (I wish Thermo would just increase the reliability of the Portal interface but that just takes too long).  Scott Boroughs has tested this new code and he says it seems to work great so far, for acquisition of EDS spectra through the Thgermo Portal interface.

So if you use Thermo Pathfinder for integrated WDS and EDS with Probe for EPMA, please update to the latest Probe for EPMA using the Help menu and if anyone still sees these socket errors please send me a screenshot and I will add additional error handling.