I have been using Labjack-2019-05-20 and found some performance issues (device Scan Backlog) when I sample at 3000Ksps using 28 channels 0-3, 48-71. so I changed to the Labjack-2018-08-30. The problem with it is the sampling has hiccups - suddenly the sampling data drops to zeros for all the channels for a very short time and then goes back to normal then eventually stops.
Has anyone experience this problem?
File Attachment:
.png%3Fitok=PHGBxf6M "EEV Blog Reviews LabJack T7-Pro & Software Tools")
.jpg%3Fitok=B1YcuDFK)
.png%3Fitok=Ybt-OlgU "LabJack + JEDI One Can Create a Powerful Distributed DAQ Ecosystem")
If deviceScanBacklog ever becomes large or continuously increases, the host computer running LJM is not keeping up with reading samples from the LabJack. LJM 1.2100 automatically increases the thread and/or process priority such that it's more likely to read samples fast enough. 1.2100 is currently in the beta installer:
https://labjack.com/support/software/installers/ljm
If you're still having issues after that:
What is the CPU usage?
For the sampling hiccups, what is the actual value of those hiccups? Is your graph able to display values less than 0? When LJM is unable to keep up with reading samples from the LabJack, it inserts -9999 values into the data stream:
http://labjack.com/support/software/api/ljm/function-reference/ljmestrea...
When stream stops, what is the error code?
The value is -9999 when this happens - again I am using the LJM from Lbjack-2018-08-30. I have more chance run into this problem with Labjack-2019-05-20.
Is the 2020-03-20-beta better?
This problem happens with LJM 1.20 very often and after I switched to LJM1.19, it works fine. But sometimes it happens and after I reboot the computer, it fixed the problem. But you do not know when it will come back again.
Let me know if anyone has this experience.
Yes, I expect the 2020-03-20 version is much better, which has LJM 1.2100.
Thanks - your support is really prompt :)