Missed measurement samples in SignalExpress

[sskulpone]

I am trying to use NI Labview SignalExpress - Tektronix Edition v2.5.1 to log Pk-to-Pk V measurements on a Tektronix TPS 2024B scope via RS-232C serial interface (USB-Serial adapter, FTDI chip). I have installed a Tektronix supplied "SE_Step_1K2K" driver (it isn't clear if the built-in "Tek TDS1000/2000" step supports this scope), as well as National Instruments VISA 5.41 full (only to check the VISA address and make sure the scope responds). I am finding that trying to record the desired measurements results in missed samples (e.g., if I expect to see 10 samples I only get 3 captured).

I added the "Tek TDS / 1K 2K" step to a new project, then set the correct VISA resource in the VISA Resource tab. This syncs the channel settings from the scope into the step. Within Step Setup I then toggle (checkmark) the Peak to Peak measurement on the Measurments tab for Channels 1 and 2. On the Recording Options tab, I checkmark the Record column for two measurements. "Record While Running" is set on. View > Update Signal Views While Running is set to "Update None" and in Logging options, "Prepare log data for viewing" is set to "After logging completes" (as I was led to believe this can save resources for faster acquisition rates).

When I hit Run, it takes a while (10 seconds?) for the Log entry to appear in the log viewer section in lower left corner after the "Recording" status appears. I start the 1 Hz trigger cadence on Channel 1 after this appears (assuming that the scope is not ready to acquire before then), and let it go for 10 or 15 triggers (there is an electric crack each time, so I know when each trigger should be fired). I hit Stop, then after the recording finally stops and I check the data via Export to Microsoft Excel, I only get, say 4 to 7 samples instead of 10 or 15.

Is there something obvious I am doing wrong?  Speaking with Tektronix tech support, they conclude SignalExpress must not be fast enough to keep up, and suggest I either go LabView or program a loop in VISAcmd.  Admittedly, their support of the SignalExpress product doesn't go much beyond getting it installed and speaking with the scope.

Background: independant of SignalExpress, the scope has a saved setup that is recalled to perform sonic output testing. Trigger signal is on Channel 1 and signal to be measured is on Channel 2 (hydrophone/microphone). Trigger occurs at 1 Hz, then the Pk-to-Pk value is measured on-screen for Channel 2. However, this scope can't save these measurements (not that we are aware of) so the operator has to write down the measurement as it happens, at 1 second interval for 10 samples in sequence (any longer and they tend to miss the samples). Hope was to use SignalExpress to capture this instead, thereby relieving operator of need to record manually.

Thanks for any tips.

[Queen_Nadine]

Hello sskulpone,

I have a few questions for you:

1. With respect to the samples that are actually acquired, are you receiving expected values?
2. Have you tested acquiring readings using Measurement and Automation Explorer (MAX) and VISA Test panels?
3. Is there consistency with respect to the data being acquired? Are you missing exactly one out of two points or is there a random frequency of missed data points? This may help us pinpoint what exactly might be occurring.

Regards,

[sskulpone]

Nadine,

Thanks for your reply.  In answer:

1. Of the samples acquired, they are the expected values (matching a subset of measured values on-screen)

2. No, I have only verfied the VISA resource name and basic connectivity (*IDN?) with MAX before jumping into SignalExpress.

3. I am not certain but believe it is random because I do not always receive x% of the expected samples.  For instance, if I run the test 5 times with 10 triggers each time, the number of captured measurements is not consistent for each run.  As mentioned previously, after clicking Run in ***** it takes a 10s of seconds for the dated Log entry to appear and I do not start triggering until after then (assuming that otherwise I might be missing recording samples prior to then).

Sakorn

[Queen_Nadine]

Hello Sakorn,

This seems quite unusual. I'm not sure why this could be occurring, so any extra information or screenshots that you might be able to gather would be useful to determine which part of the system is producing unexpected results.

Could you attach some screenshots of the data that has been acquired on a few different runs?  Also, what happens if you disable logging and simply look at the data coming in on both channels? Is there any change in the amount of data that is acquired? Could you also attach screenshots of those signals?

Regards,