hi dear,
i have made a electromagnetic flowmeter,
i am getting linear output, i mean to say that when the flow rate increases the output voltage also increases, now i have to make it such a way that it will directly show the output " liters per hour" on a LCD.
final year project
please help
How to calibrate a electromagnetic flowmeter, please hel
-
- Posts: 12
- Joined: Sat Mar 25, 2006 1:01 am
- Location: mumbai
- Contact:
-
- Posts: 12
- Joined: Sat Mar 25, 2006 1:01 am
- Location: mumbai
- Contact:
Re: How to calibrate a electromagnetic flowmeter, please hel
thank JPKNHTP,
i am really glad.
Thats ok,
ill convernt the o/p digital and stuff, but how would i convert it to a "liters per hours" form the o/p is in mV.
i am really glad.
Thats ok,
ill convernt the o/p digital and stuff, but how would i convert it to a "liters per hours" form the o/p is in mV.
yours shiva
-
- Posts: 12
- Joined: Sat Mar 25, 2006 1:01 am
- Location: mumbai
- Contact:
Re: How to calibrate a electromagnetic flowmeter, please hel
again thanks
ya,
am going to just like that, flow from a standard instrument that my flowmeter,
ok suppose i am getting
13mV for 100LPH
16mV for 200LPH
18mv for 300LPH
than how would i convert those mV signals to LPH form
ya,
am going to just like that, flow from a standard instrument that my flowmeter,
ok suppose i am getting
13mV for 100LPH
16mV for 200LPH
18mv for 300LPH
than how would i convert those mV signals to LPH form
yours shiva
-
- Posts: 12
- Joined: Sat Mar 25, 2006 1:01 am
- Location: mumbai
- Contact:
Re: How to calibrate a electromagnetic flowmeter, please hel
sir, JPKNHTP
i am in second year of diploma in instrumentation, i am not that level of intellegence to calculate such things, but really i will start studying this ADC thing from right now and try to give a result.
i am in second year of diploma in instrumentation, i am not that level of intellegence to calculate such things, but really i will start studying this ADC thing from right now and try to give a result.
yours shiva
Re: How to calibrate a electromagnetic flowmeter, please hel
Hello shiva,
Note that your output isnt, strictly speaking,
"linear", in that you cant draw a straight line
through the points and the origin. Heck, you
cant even draw a straight line through the points alone!
This of course means your output is a curve,
and so to calculate the flow rate given the mv
output of your detector you need to do some
curve fitting. In curve fitting you'll fit
your data to some well-known curve such as a
second or third order polynomial. Once you
have the coefficients for your polynomial you
can program in the formula to convert from mv
to flow rate and show it on the display.
Have you done any curve fitting?
(Sometimes also called "linear regression")
Take care,
Al
Note that your output isnt, strictly speaking,
"linear", in that you cant draw a straight line
through the points and the origin. Heck, you
cant even draw a straight line through the points alone!
This of course means your output is a curve,
and so to calculate the flow rate given the mv
output of your detector you need to do some
curve fitting. In curve fitting you'll fit
your data to some well-known curve such as a
second or third order polynomial. Once you
have the coefficients for your polynomial you
can program in the formula to convert from mv
to flow rate and show it on the display.
Have you done any curve fitting?
(Sometimes also called "linear regression")
Take care,
Al
LEDs vs Bulbs, LEDs are winning.
-
- Posts: 12
- Joined: Sat Mar 25, 2006 1:01 am
- Location: mumbai
- Contact:
Re: How to calibrate a electromagnetic flowmeter, please hel
suppose i have done this liner regression thing then what would i do
yours shiva
Re: How to calibrate a electromagnetic flowmeter, please hel
Hi again,
Once you do the regression you'll have the
coefficients for the polynomial.
Lets say you choose a third order poly, then
you'll come out with four coeff's:
a3, a2, a1, a0, and use the following equation:
y=a3*x^3+a2*x^2+a1*x+a0
where
x is the value in mv, and
y is the flow rate.
Next you program in this formula into your
uP. Ideally, your uP would already have an
ADC built in so the output (mv) from your
flow detector would go right into the uP input
port, or perhaps after some amplification to
adjust the range of the signal to a value more
appropriate for the uP's built in ADC.
From there it's just a matter of displaying the
output on a display of some type.
I think your original question was how to
calibrate though?
To calibrate you'll put a working flow meter
in the same flow line as your new flow meter.
First you would record the reading with no
flow (zero flow). This will probably be zero
for both meters.
Then you'll set the flow to some small value
and record the meter readings from both meters.
Then you would increase the flow by some step
value and record both meter readings again.
Repeating this procedure, you'll come out with
a set of values to use for the curve fitting.
Does this help?
Take care,
Al
<small>[ March 25, 2006, 10:19 PM: Message edited by: MrAl ]</small>
Once you do the regression you'll have the
coefficients for the polynomial.
Lets say you choose a third order poly, then
you'll come out with four coeff's:
a3, a2, a1, a0, and use the following equation:
y=a3*x^3+a2*x^2+a1*x+a0
where
x is the value in mv, and
y is the flow rate.
Next you program in this formula into your
uP. Ideally, your uP would already have an
ADC built in so the output (mv) from your
flow detector would go right into the uP input
port, or perhaps after some amplification to
adjust the range of the signal to a value more
appropriate for the uP's built in ADC.
From there it's just a matter of displaying the
output on a display of some type.
I think your original question was how to
calibrate though?
To calibrate you'll put a working flow meter
in the same flow line as your new flow meter.
First you would record the reading with no
flow (zero flow). This will probably be zero
for both meters.
Then you'll set the flow to some small value
and record the meter readings from both meters.
Then you would increase the flow by some step
value and record both meter readings again.
Repeating this procedure, you'll come out with
a set of values to use for the curve fitting.
Does this help?
Take care,
Al
<small>[ March 25, 2006, 10:19 PM: Message edited by: MrAl ]</small>
LEDs vs Bulbs, LEDs are winning.
Who is online
Users browsing this forum: No registered users and 1 guest