开云体育

Well that’s pretty easy.? Design for maximum flow first and check resolution.? A flow rate of 90 L/m is 1.5L/s… and if you do the math, using a 1 inch pipe (actual 1.029”) solving Bernoulli’s equations you find you must use an orifice (hole in a plastic insert/ washer) that is 0.0194m = 1.94 cm = 0.763” hole.? That is basically a 49/64” drill bit at 0.7656” true.? The resolution on that device to get to 2.5L/min = 0.004166 L/s, resolution required is 1.5 Pa = 0.0002175 psi.? That is tiny…. BUT the resolution is 0.30psi/ 52428 = 0.00000572 psi !!!.? That looks super good, but before you get really excited about this, note that the signal is gonna be noisy down in that range.? Turbulent air flow will cause the noise and without testing it we won’t know.

Dr. William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ

Owner - Operator

Big Signal Ranch – K9ZC

Staunton, Illinois

Owner – Operator

Villa Grand Piton – J68HZ

Soufriere, St. Lucia W.I.

Rent it:

Like us on Facebook!

Moderator – North American QRO Group at Groups.IO.

email:? bill@...

Show quoted text

From: [email protected] [mailto:[email protected]] On Behalf Of Gordon Gibby
Sent: Friday, March 27, 2020 7:59 PM
To: [email protected]
Subject: Re: [VentilatorDevelopers] #FLOWSENSOR

So if we have that sensor, and we need to be able to measure up to 90 L per minute, with a resolution of let’s say at least 2.5 L per minute, and we don’t want to waste a lot of pressure, what orifice do we need and what voltages will we get?

We don’t have to have perfect resolution. ? Getting it down to 1 L per minute would probably be astonishingly good. We’re just going to integrate the area under the curve to calculate the total volume. ?

On Mar 27, 2020, at 20:32, K9HZ <bill@...> wrote:

?
Background.
?
Calculations for wedge or orifice meter from Bernoulli’s equations reduce to:
?
? AIR =====èA1,P1àA2,P2à===== Flow, Q
?
Q = A2 * [ 2*(P1-P2)/(Rho * (1-A2/A1)^2)]^(1/2) …
?
Where,
?
A1 > A2, area in cm^2
?
P1 > P2 (because A1>A2), pressure in Pa... (1 kg/(m*s^2) or 1.450x10^-4 psi)
?
Q = Flow rate in cm/s
?
Rho is density of air = 1.2041 kg/m^3 or 1.2041x10^-3 g/cm^3.? Note that air density is a function of temperature via PV=nRT.? However, over the range of calculation, from 0C to 30C, density of air only varies from 1.29x10^-3 to 1.16x10^-3 g/cm^3. That is roughly 10%, and if calibrated at 68F (20C) with density 1.204x10^-3 g/cm^3, means less than 2% error with +/- 10degF.? Good enough for this application so that no temperature of the gas is needed (although some sensors have built-in temperature measurement as well).
?
Absolute Requirements:
?
1.???? Range of 0-90 L/min (= 1.5 L/s or 1500 cm^3/s).
2.???? Resolution of 0.03 psi (2.06 hPa)
?
Case 1:? BME280 Sensor, according to the adafruit website, ?is capable of resolution of +/- 1 hpa… or about 0.015 psi.? Range is 0-1100 hPa or 15.95 psi.?
?
Design A1, A2 such that delta P is 15.95 or full scale.? This is the largest pressure drop, but gives the highest degree of accuracy.? Note that we could drop down to half of this range or about 7.92 psi and still maintain the required resolution of 0.03 psi for full scale flow.? For now, use 15.95.
?
From the equations, ?solving for A2 in terms of P1, P2, Q, and Rho:
?
Q = A2 [ 2*(P1-P2)/(Rho * (1-A2/A1)^2)]^(1/2) or
?
Q^2 =A2^2 *2*(P1-P2)/(Rho*(1-A2/A1)^2) or
?
Q^2 =A2^2 *2*(P1-P2)/(Rho*A2^2/A1^2) or
?
Q^2*Rho*A2^2 – A2*2*[(P1-P2)+A1*Q^2*Rho] + Q^2*A1^2 = 0
?
Quadratic roots are:
?
For A1; d1=1” (is actually 1.029” ID sched 40 PVC pipe) = 2.61367cm, A1=5.365cm^2, P1-P2= 1100 hPa, P2=0, (so differential is 1100 hPa), Q=1.5L/s=0.0015 m^3/s
?
(using solver) A2=1.002 cm^2; d2= 0.787 cm = 0.309 inch. A “N” drill bit is 0.302 inches and will work perfect this.
?
Using half scale but maintaining the accuracy constraint, differential is 550 hpa, A2=.6054=0.00937m = 0.937cm = 0.368 inches.? A “U” drill bit is 0.368 inches and will work perfect.
?
MUST check Choke flow calculation (eg is this below supersonic flow?):? 42.7 SCFM = 1210 L/min… so all ok for this case.
?
Case 2: Adafruit LPS22HW
?
Won’t work…. Minimum floor is 260 hPa.
?
Case 3: First Sensor HUMB001DU7
?
<image002.jpg>
?
Range 0-61mv.? 0-15psi differential.? Solved for d2=0.0079m = 0.790cm = 0.311 inches; A2=0.4902cm^2 ; A 5/16” Drill bit works perfect for this at 0.312 Inches
?
Case 4:? Honeywell ABPDRRV015PDSA3 (Mouser)
?
<image006.jpg>
3.3V SPI/ I2C interface? Differential 0-15 psi.? Accuracy 1.5% of FS.? Solved for d2=0.0079m = 0.790cm = 0.311 inches; A2=0.4902cm^2 ; A 5/16” Drill bit works perfect for this at 0.312 Inches
?
Case 5:? NXP brand MPX53DB (Mouser $11 In Stock)
?
<image007.jpg>
?
50Kpa = 7.25 psi differential.? Bridge connection.? Vs = 3V, output 60 mv FS, 1.2 mv/mPa sensitivity.? Full scale on this for 1.5L/s is 7.25 psi differential.? Solving for d2= 0.00943m = 0.943cm = 0.371”.? A 3/8” drill bit is 0.375” and is adequate for this application.
?
?
Case 6:? NXP brand MPXV2010DP? (Mouser In stock $8.99)
?
<image012.png>
10 Kpa range = 1.45psi range (this is a good choice for maximum flow and least pressure resistance).? D2=0.0139m = 1.39cm = 0.547” hole. 10VDC Supply, 25mv full scale so problematic for output.? ?A 35/64” Drill is 0.5469” and a perfect fit.
?
?
Conclusion
?
Best choices are the NPX MXP53DB or (second) the BME280 IF they can be contained in a pressure-proof pipe coupler/ fitting (eg in a 1” tee before and after the “wedge” with just the wires sticking out).
?
?
?
?
Dr. William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ
?
Owner - Operator
Big Signal Ranch – K9ZC
Staunton, Illinois
?
Owner – Operator
Villa Grand Piton – J68HZ
Soufriere, St. Lucia W.I.
Rent it:
Like us on Facebook!
?
?
Moderator – North American QRO Group at Groups.IO.
?
email:? bill@...
?
?
Virus-free.

View All 85 Messages

#391

Join [email protected] to automatically receive all group messages.

开云体育

Re: #FLOWSENSOR #flowsensor

开云体育

Re: #FLOWSENSOR

#flowsensor