Determine Cycle Time for a Vacuum Pump

Recommended Posts

I am using a version 1.21 U3 and I need to figure out how to determine cycle time for each vacuum pulse. I have the maximum pressure and vacuum showing using Min(Pressure[0,99]) but I can not determine how to capture the time at the max or minimum pulse. I would thing I could capture the time of the maximum pulses and then subtract to get the cycle time. Does this sound reasonable? I am sampling every 0.05 seconds.

I have read all the user manuals for DaqFactory Express and the Labjack and have been unable to determine how to do this.

My sensor is a differential pressure gage reading from +/- 1 bar. The output is analog voltage. I am getting a nice line graph of what is going on along with readings averaged for the last 100 samplings for the max and min pressures/vacuums.

If I could also get cycle time I would be a happy camper.

I tried this GetTime(Max(Pressure[0,20]))- GetTime(Max(Pressure[21,20])) on the Pressure channel but it is not working correctly.


Link to comment
Share on other sites

Some general thoughts, and then Azeotech will probably need to help with the specifics of implementing your algorithm in DAQFactory.

You need to choose a threshold that you use to define PumpOn. Looking at your plot, you might choose something like -0.4 inHg. You then use a state machine track your current state and detect changes in state. For instance, you would probably start your program in the PumpOff state, and then check each point in order to look for a point that is less than the threshold. When you find such a point, you change your state to PumpOn, start counting the time, and watch for a point that is greater than the threshold.

The other thought, is that depending your accuration and resolution requirements for time, you might need to use stream mode. You are currently using a time resolution 50 ms in command/response mode. If this resolution is sufficient, then you are probably in pretty good shape. I would expect Windows to maintain pretty good accuracy relative to that time interval (low jitter). Since the timing is dictated by software, for the lowest jitter you should avoid running other programs during your tests, and perhaps even bump up the priority of the program in Windows Task Manager. If you need finer resolution (down to 20 us), or can't tolerate the risk of jitter introduced by Windows, then you need to use stream mode which acquires data at precise hardware timed intervals.

Link to comment
Share on other sites


This topic is now archived and is closed to further replies.