I have a turbine flowmeter giving a frequency output (-5 : 5 V) with a wave shape as in the figure attached. This signal is fed to a Digital-to-analog converter to give a 4-20 mA signal to feed to my DAQ.
Now, I need to set the range for the converter (to know which frequency gives which current), to do this I should use a function generator, and I won't be able to produce the same shape as the output from the flow meter, so the question is:
Is the output signal (4-20 mA) dependent on the shape of the input wave or just the frequency?
Hi Ibrahim O. Elhagali ,
the device you need, I wouldn't call a DAC but a Frequency-to-Current Converter. Maybe it's only a matter of designation; in order to decide: Could you please provide a datasheet or a manual for your device?
If it is indeed an FCC, the general answer to your question is: No, the shape of the signal doesn't matter as long as the duration of the halfwaves doesn't fall below a certain value. This value is either given directly in the datasheet or you can derive it from other data like: fmax = 1 kHz, duty cycle 0.2 .. 0.8. In this example, the minimal period would by 1 / fmax = 1 ms, and a duty cycle of 0.2 would correspond to a duration of 0.2 ms.
A typical FCC, suitable for an analog input signal like yours, feeds the input signal into a comparator, which results in a digital signal, and only the rising (or falling) edges of this digital signal are used to determine the frequency.
To make things a bit more complicated: With a shape as in your figure, and depending on the input circuit (probably capacitive coupling?), when the frequency is very low it might be that the rising edges from C to zero and from zero to A are recognized as two independent rising edges. This problem would result in a sudden increase of the output current when the flow decreases monotonously. Rectifying the signal by a diode might be a solution; in the resulting signal, the dashed portions of the signal in your figure would be replaced by 0 V sections.
Another important parameter is the input voltage range, of course.
Hello Ibrahim O. Elhagali,
The things would be more clear if we study the datasheets of each part. Generally, the flow measurement arrangement produces digital (square) output pulses with 50 percent duty cycle assuming the flow is constant. Now, the output is converted into 4- 20 mA by a suitable frequency to analog industrial grade output. My answer for waveform shape is no. It should not matter. You can check the shape from the sensor output directly and also confirm the duty cycle using oscilloscope. Obviously you can calibrate the F to current module by using a suitable function generator by keeping desired amplitude of square wave. Actually the F to current module datasheet must include this relationship with the accuracy of the same.
Good luck.
I Am sorry in previous post i misunderstand what is actualy question .Based on known data and assumption there is my opinion . So if You have standar flowmeter with pulse output which gives pulse between every cycle of rotor wheel (mechanical ) then 4-20 mA is highly dependent on freq . You need to meassure time between 2 pulse ( wheel full cycle) and that time ( indirect freq.) is directly related to flow ( fluid velocity) Mass and volumetric flow could be easily calculated with known denisty . Shape of wave input does not have matter . ( of course distorsion on wave ,which represent shape ,is directly related to quality and accuracy of transmision signal )
For flow meter simulation, even a square wave with 50 % duty cycle should do. Even unipolar, although you might have to fiddle with the amplitude and offset to match the receiver's characteristics.
What you currently see in the flow meter's output looks very much like the induction from a magnet into some coil. So the amplitude might vary over flow (higher flow -> higher amplitude), at some point distorted by the receiver's input protection.
A "usual" receiver will just count pulses over time (e.g. 1 s), updating the output signal once every measurement period.
That's it - no rocket science.
Please share the data sheet of the digital to analog converter you are using . That would help in deciding what would be an acceptable waveform for the simulated input
Dear Imbrahim,
Please provide the datasheet of both the flowmeter and the 4-20 mA converter for a more accurate advice.
- Badges
- Science topic
- Similar topics
- Psychometrics
- Measurement