DAQ for rotational speed and flow rate measurements

[q123456789]

Hello Everyone,

I'm in a new project and want measure the rotational speed of a diesel engine and the diesel flowrate . I want use NI products, but couldn't decide which DAQ card is the most suitable for me. 

I want measure the instantaneous speed from a speed sensor which sends a pulses for each revolution. The speed I want measure is less than 3000rpm. The flowrate sensor will give me pulses as well. Also, I need it to have some AIO channels to measure the voltage and current signals. I would be very apprciative if you can provide me with some suggestions about the most suitable DAQ for me. 

Thanks

Mike

[Paul_C.]

Hi Mike,

The easiest way to measure the rotational speed and diesel flowrate would be to use a counter.  Doing a simple period measurement with a counter will calculate the time between two rising edges of a TTL signal.  Assuming the sensors output the proper TTL signal, a device with the proper number of counters should meet your needs.  Many of our M-series data acquisition cards come with 2-4 counters in addition to analog input / output and digital input / output lines.  Without knowing the form factor (PCI, PXI, USB, etc), number of channels, sampling rates, etc, I won't be able to make a specific recommendation.  I do recommend taking a look at our website here.  All you need to do is narrow down your DAQ card results based on channel count, form factor, measurement types, etc.  The only difficult thing you might have trouble finding is a card that will do the current measurements without you providing the external resistor.  This functionality isn't typically included in our multifunction DAQ line.

Regards,

Paul C. 

[q123456789]

Thanks Paul,

Based on the options available in the link you provided, I have decided to go with the NI PCI 6221 M Series DAQ card with RC68 -68 cable and the connection box. 

Also, thanks for your suggestion regarding how to calculate the speed, but I have one more question. Is it possible to get the "instantaneous" speed from counting the edges of the pulse train i get from the sensor?

Thanks

Mike

[Paul_C.]

Hi Mike,

I'm assuming that by "instantaneous speed" you mean you want to know the speed at a instant between pulses or at a specific point in a revolution.  To measure the speed we need two frames of reference.  For example, the way we calculate the period measurement is by counting the number of 80MHz pulses occur between the two rising edges of your sensor's signal.  So, the most we can know is the average speed over that 1 rotation in between rising edges.  If your sensor was able to output a rising edge more times per revolution, we could know the speed between each rising edge doing something called a buffered period measurement task.  This stores all the "speed" values calculated between all of the rising edges of your source signal.  In short, the closer the rising edges or ticks per revolution, the more "instantaneous" the measurement.

Regards,

Paul C.
 

[q123456789]

Thanks Paul once again,

The speed I'm measuring will vary in the renge of ~1500rpm. So if I get the speed information every ~0.04sec (i.e. each revolution) would be enough for me, and that is what I meant by "instantaneous speed". So did you mean that, the counter will save the speed values with their relavent time? if yes, can you give me some hints on how to do this in Labview?

Thank you very much

Mike

Message Edited by q123456789 on 09-23-2008 06:59 PM

[Paul_C.]

Hi Mike,

This is pretty easily done in LabVIEW with a buffered period measurement.  My recommendation would be to start with one of our example programs.  A good example to use is called Meas Dig Periods-Buffered-Continuous-High Freq 2 Ctr.vi.  This example can be found by going to Help » Find Examples in LabVIEW and navigating to Hardware Input and Output » DAQmx » Counter Measurements » Period or Pulse Width.  There are a few changes that need to be made to change this example to a low frequency period measurement using 1 counter.

1. Adjust the blue enum going into the DAQmx Create Channel.vi to "Low Frequency with 1 Counter".

2. Delete the "Measurement Time" control on the front panel and delete the broken wire on the block diagram after doing so.

3. Adjust the Samples to Read control to something like 10-100.  You can also adjust this to -1 if you want to just read all available samples.

It is important to note that these samples will just be the period of the gate signal and won't contain timestamp information about when they were taken.  However, you can use the order at which you receive the data to tell you information about when measurements were taken.  Another thing you can do is take a software timestamp prior to starting the acquisition.  This would give you a relative means to determine when measurements were taken or what the speed was at a specific time.

I hope this helps,

Paul C.

[q123456789]

Thank you so much Paul,

The time information is exactly what I need to know. Could you please explain more about how to use the "order at which data is received" or the "software timestamp" to get the time information?

I really do appreciate your help

Thanks

Mike

[Paul_C.]

Hi Mike,

"Software timestamps" can be achieved by using the get date/time in seconds vi on the block diagram.  When this vi is executed, it will check the system time from your computer clock and return the time.  This can be used prior to starting the DAQmx task to let you know aproximately (again this is not 100% accurate) when your counter acquisition starts.  

When I said, "order at which data is received", I was referring to the fact that each sample you receive corresponds to the value of the period between two rising edges.  In a buffered period measurement task, you will receive samples in order.  For example, the first sample will correspond to the period of the first revolution, the second sample corresponds to the second revolution, etc.  

I hope this helps,

Paul C. 

[q123456789]

Thanks Paul,

I deeply appreciate your help. 

Kindest regard, 

Mike

[Colin hughes]

Hi Mike

We use Freq to Voltage converters for flow measurments.

Colin