Mass flow controller (Bronkhorst) data sampling rate

[b.erkal@utwente.nl]

Hello everyone, I have vi which controls my mass flow rate and measures it with a Bronkhorst mass flow controller. I need to control data sampling rate of the mass flow controller and I cannot acces the information for current data sampling rate. How can I solve this issue.

:
Mass flow controller: Bronkhorst D-6391-DF with flowbus
DAQ: cRIO-9057

Thank you in advance 

 

[GerdW]

Hi b.,

---

b.erkal@utwente.nl wrote:

I need to control data sampling rate of the mass flow controller and I cannot acces the information for current data sampling rate. How can I solve this issue.

---

Read the manual for your MFC to learn about its specifications. The sampling rate should be one item of the specs…

Which sample rate (aka loop iteration speed) do you reach now?

What is the desired sample rate?

On your image:

• All the error handling is missing…
• Why do you use MergeSignal to create a DDT wire before the WriteFile function? Why don't you create an array of samples by using BuildArray?

[b.erkal@utwente.nl]

Dear GerdW,

Thank you for your fast response. I am pretty new to Labview, so I might make little mistakes. Answers to your questions:

The current sampling rate is random. Sometimes it is 5 data points per second, sometimes 4 or 6. It does not have equally spaced time intervals. I write this according to the 'write to measurement file' block's excel output with real-time timestamps. Moreover, unfortunately, the datasheet of the mass flow controller does not have information about the sampling rate. (https://www.bronkhorst.com/getmedia/ea3f9d74-a738-49a2-8ad8-cfd07586398a/MASS-STREAM-D-6300-960067.p...)

The desired level of data sampling rate is at least 5kHz.

I would be happy to understand what you mean by missing error handling. What is the importance of it?

Merge signal to create DDT was a personal trial, I am sorry for the confusion.

Best Regards

BE

[GerdW]

Hi b.,

expecting a sample rate of 5kHz with a serial communication at just 38400 baud is completely unrealistic!

You can only transfer ~3800 bytes per second - including commands and term chars! And the MFC also needs time to process your commands before it can answer each request...

When you need such high sample rates then you should think about using a dedicated mf sensor!

[AnalogKid2DigitalMan]

The response and settling time of the sensor/controller is on the order of seconds (from the datasheet):

Response time sensor (63 %) D-631x / D-632x: approx. 0,3 seconds
other models: approx. 0,9 seconds

Settling time (controller) up to approx. 2 seconds
(depending on selected valve)

-AK2DM

[rolfk]

Gerd already answered you that a serial port will NEVER EVER allow you to get to 5kHz sampling rate, not even if you could get the baud rate much higher than 39k.

The real question is why you think to need such a sampling rate? What are you trying to measure? Flows are generally rather slow changing so measuring at high rates usually only achieves to require more expensive hardware and software!

If you measured the analog output of the sensor you could of course go up to 5kHz if you so dearly wanted (with an analog input module in the cRIO). But the sensor still only generates a few measurements per second since the measurement principle is integrating. The rest is simply noise that you will measure.

[b.erkal@utwente.nl]

Dear all,

Thanks for the helpful responses. I have an old custom made valve that oscillates flow with a given command. I wanted to validate its response whether i get the right mass flow rates or not. But, as you mentioned, the current mass flow controller isn't up to the task. I'm thinking of adding a mass flow meter to check my valve's performance. Otherwise, without that valve, I don't need a higher sampling rate.  

Appreciate your insights.

Best Regards