Academia.eduAcademia.edu

Figure 5 – uploaded by Octavian Postolache

See full PDFdownloadDownload figure
Figure 5. Graphical representations: (a) experimental (0) and interpolated (-) measurements (Ey y=21,2% and (b) linearization error after ADC-PWM A/D conversion (E,w=0.96% for p>5 p.s.1. and ky, =1,42)

Figure 5 Graphical representations: (a) experimental (0) and interpolated (-) measurements (Ey y=21,2% and (b) linearization error after ADC-PWM A/D conversion (E,w=0.96% for p>5 p.s.1. and ky, =1,42)

Related Figures (4)

Figure 1. Pressure drop (dP) between upstream (HP) and downstream (LP) sides of an orifice sensing element
Figure 1. Pressure drop (dP) between upstream (HP) and downstream (LP) sides of an orifice sensing element
Figure 2. P&I diagram used for flow meters’ experimental tests Figure 2 represents the piping and instrumentation (P&I) diagram we used for flow meters’ experimental tests. It includes the following set of equipments: an electro-pump (CPT45 Pentax) controlled by a three phase compact general purpose inverter (VS mini J7 Omron), a tank (14 gallon capacity), a manual valve, a pressure indicator (PI), an electromagnetic flow meter (Rosemount 8712C), a Coriolis flow meter (RFT9712) and an orifice flow meter under test (in-house made). Figure 3 represents the basic hardware of each flow transmitter (F1, F2 and F3) measuring channel, which includes the current output of each flow meter (4-20 mA), the current to voltage converter (RCV420 integrated circuit), the conditioning circuit (CC) and the PWM-ADC. Obviously, the difference circuit, implemented with an operational amplifier (OA), can be replaced by an instrumentation amplifier (e.g. AD620) but the proposed solution is cheaper and works properly as long as the output current delivered by RCV420 integrated circuit is lower than 49 mA.
Figure 2. P&I diagram used for flow meters’ experimental tests Figure 2 represents the piping and instrumentation (P&I) diagram we used for flow meters’ experimental tests. It includes the following set of equipments: an electro-pump (CPT45 Pentax) controlled by a three phase compact general purpose inverter (VS mini J7 Omron), a tank (14 gallon capacity), a manual valve, a pressure indicator (PI), an electromagnetic flow meter (Rosemount 8712C), a Coriolis flow meter (RFT9712) and an orifice flow meter under test (in-house made). Figure 3 represents the basic hardware of each flow transmitter (F1, F2 and F3) measuring channel, which includes the current output of each flow meter (4-20 mA), the current to voltage converter (RCV420 integrated circuit), the conditioning circuit (CC) and the PWM-ADC. Obviously, the difference circuit, implemented with an operational amplifier (OA), can be replaced by an instrumentation amplifier (e.g. AD620) but the proposed solution is cheaper and works properly as long as the output current delivered by RCV420 integrated circuit is lower than 49 mA.
Figure 3. Basic hardware of each flow meter measuring channel: flow transmitter (FT), transmission support (TS), current to voltage converter (RC420), operational amplifier (OA); analog-to-digital converter based on pulse width modulation (PWM-ADC), comparator (Comp.), digital input (DI), digital output (DO), digital input-output ports (DIO), analog switch (S), microcontroller (uC)
Figure 3. Basic hardware of each flow meter measuring channel: flow transmitter (FT), transmission support (TS), current to voltage converter (RC420), operational amplifier (OA); analog-to-digital converter based on pulse width modulation (PWM-ADC), comparator (Comp.), digital input (DI), digital output (DO), digital input-output ports (DIO), analog switch (S), microcontroller (uC)
MATLAB routines were developed to evaluate the best value of the time constan (t) of the low pass filter included in the PWM-ADC. A correct dimensioning of this parameter is essential in terms of A/D accuracy and resolution. A very low value of t generates an excessive quantization error. On the other hand, a very high value of t lowers the quantization error amplitude but makes the ADC very sensitive to noise. Thus, it is crucial to obtain the appropriate t value in order to improve ADC performance. Figure 4 represents the amplitude variation of the average voltage modulus, across capacitor’s terminals, as a function of the pulse width (Tp), during a conversion cycle (2" pulses generated by the microcontroller). Input voltage range is normalized relatively to full-scale amplitude and the amplitude of the voltage variation (vertical axis) is normalized to ADC step amplitude (Q). The simulation parameters were: Vy=-V,=0.5, n=5 bits, and the ratio between pas a oe ee aaa Tee eeeen cai RR gti Tees (4 (OND jail 4 j_nhly pulse width (Tp) and RC
MATLAB routines were developed to evaluate the best value of the time constan (t) of the low pass filter included in the PWM-ADC. A correct dimensioning of this parameter is essential in terms of A/D accuracy and resolution. A very low value of t generates an excessive quantization error. On the other hand, a very high value of t lowers the quantization error amplitude but makes the ADC very sensitive to noise. Thus, it is crucial to obtain the appropriate t value in order to improve ADC performance. Figure 4 represents the amplitude variation of the average voltage modulus, across capacitor’s terminals, as a function of the pulse width (Tp), during a conversion cycle (2" pulses generated by the microcontroller). Input voltage range is normalized relatively to full-scale amplitude and the amplitude of the voltage variation (vertical axis) is normalized to ADC step amplitude (Q). The simulation parameters were: Vy=-V,=0.5, n=5 bits, and the ratio between pas a oe ee aaa Tee eeeen cai RR gti Tees (4 (OND jail 4 j_nhly pulse width (Tp) and RC
Related topics:
EngineeringFlow measurementUltrasonic flow meter
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved