Test and Measurement Forum

DrDAQ & Sensor make precision barograph (GENERAL INFO)

Post all DrDAQ questions here

DrDAQ & Sensor make precision barograph (GENERAL INFO)

Postby Glovisol » Sun Apr 12, 2015 4:03 pm

An inexpensive but highly accurate barograph can be built with a type MPX5100DP Freescale (Motorola) Semiconductor piezo sensor and the DrDAQ + Picolog. Due to the high sensitivity and precision of the DrDAQ/Picolog combination, it is possible to resolve barometric pressure variations down to 0.1 mmHg. For simplicity this topic will be treated as follows.

Post 1: GENERAL INFO
Post 2: SENSOR DESCRIPTION
Post 3: TECHNICAL INFO

The MXP5100DP is an analog differential sensor with a voltage output which can be directly connected to the DrDAQ input through a simple resistor/capacitor network and an Op Amp or alternatively the already described DrDAQ Universal Interface. This arrangement does away with the system complexity (hardware + software) of a digital sensor like the Bosch BMP085/BMP180. The MXP5100DP is priced below £10 and is readily available. The enclosed screens illustrate actual barograph performance.

As will be later explained in detail, the sensor needs to be pre-loaded with a NEGATIVE PRESSURE on the Negative (P2) input port and then barometric pressure will be measured on the Positive (P1) port. However all attepts to stably preload with vacuum the P2 port failed, because it is impossible to eliminate the long term microleaks, due to the low density of air. For this reason the successful, highly stable preload solution has been to apply preload pressure to the P1 (Positive) port with a column of Vaseline oil 1600 mm high (REVIEW NOTE: preliminary experiment only, later colum height has been increased to 2200 mm) and to measure barometric pressure at the P2 port. The top of the oil column is closed to barometric pressure by means of a suitable plug, so only oil weight is loading the port. With this arrangement the sign of pressure variations is inverted, but this problem is easily solved with Picolog mathematics.(REVIEW NOTE: later work showed that system is improved with P2 port plugged and pressure measured by P1 port, also eliminating the inversion in the sign of pressure variations).

System performance is indeed notable, as the DrDAQ/Picolog combination allows the measurement of minute barometric pressure variations and their stable recording. Extremely good system temperature stability has also been demonstrated. Another novel system characteristic is that any barometric pressure can be easily simulated with a syringe (this being impossible with single ended sensors) as will be shown in a future post. Finally Picolog allows barometric pressure to be displayed with any Physical Unit (mmHg/HPa/InchHg/etc.) and/or reduced to absolute or to sea level value, by simple mathematical manipulation.
Attachments
Barotrue 100e.jpg
Barograph view with single parameter display
Experimental setup.JPG
Components of lab experiment
Experimental preload.JPG
1600 mm oil column
Barotrue 100a.jpg
Barograph display near switch-on
Barotrue 100d (R).jpg
Barograph display after 7 hours
Last edited by Glovisol on Thu Apr 30, 2015 4:27 pm, edited 2 times in total.
Glovisol
Advanced User
Advanced User
 
Posts: 114
Joined: Sat May 11, 2013 10:03 am
Location: San Marzano Oliveto, Asti, Italy

Re: DrDAQ & Sensor make precision barograph (GENERAL INFO)

Postby Glovisol » Tue Apr 21, 2015 6:21 pm

SORRY, FORUM SEEMS TO BE IN TECH TROUBLE & IT IS IMPOSSIBLE TO CARRY ON WITH THE UPLOADING OF THIS PROJECT.

glovisol
Glovisol
Advanced User
Advanced User
 
Posts: 114
Joined: Sat May 11, 2013 10:03 am
Location: San Marzano Oliveto, Asti, Italy

Re: DrDAQ & Sensor make precision barograph (GENERAL INFO)

Postby Martyn » Wed Apr 22, 2015 6:09 am

Permissions should be sorted, your description updated, I still haven't managed to synchronise post counts yet.
Martyn
Technical Specialist
Martyn
Site Admin
Site Admin
 
Posts: 2339
Joined: Fri Jun 10, 2011 8:15 am
Location: St. Neots

Re: DrDAQ & Sensor make precision barograph (PERFORMANCE)

Postby Glovisol » Thu Apr 23, 2015 8:18 am

PERFORMANCE TESTS & RESULTS

Since last post, system has been modified by removing the plug from the preload oil column and placing it at the differential (P2) input. In this way pressure variations influence the cell with the sum of barometric pressure + weight of oil column. The ideal system would use a preload mercury column (a height of 20 cm would suffice and no density variations would be noticeable) but unfortunately sale of mercury metal is forbidden by law in all E.U. The oil column (with a height of 2.2 metres) is as effective, but unfortunately oil density changes with temperature and the cell output voltage must be corrected in consequence. This proves the extreme versatility of the DrDAQ/Picolog combination: an LM35 temp. sensor has been inserted into the column and senses oil temperature variations. It is then a simple matter to use a math channel to correct the cell voltage.
PC screen 400a shows nearly perfect compensation over a 12 hour run with a sensor/column temp variation from 18.3 to 16.5 °C. Small differences (less that 0.5 HPa) of barometric pressure are easily detected.

The # 500 run is detailed by 4 PC screens and shows excellent performance and stability. Minute barometric variations impossible to detect on a "normal" barometer/barograph are easily seen here and provide a powerful insight on weather changes. The sistem is so sensitive it must be kept away from doors, as it detects variations in air shifts when they are moved.

Future post will provide detailed technical info.
Attachments
Barotrue 400a.jpg
400a - temperature compensation of cell voltage
Barotrue 500a.jpg
500a - detailed 6 hour run
Barotrue 500b.jpg
500b
Barotrue 500c.jpg
500c
Barotrue 500d.jpg
500d
Last edited by Glovisol on Thu Apr 30, 2015 4:29 pm, edited 2 times in total.
Glovisol
Advanced User
Advanced User
 
Posts: 114
Joined: Sat May 11, 2013 10:03 am
Location: San Marzano Oliveto, Asti, Italy

Re: DrDAQ & Sensor make precision barograph (TECH DATA 1)

Postby Glovisol » Wed Apr 29, 2015 4:23 pm

6. Baro tech.JPG
Barograph schematic diagram and technical data
TECHNICAL DESCRIPTION 1

The barograph is successfully completing a final 96 hour test run & PC screens and data will be published in a future post.

The DrDAQ Barograph is based on the MXP5100DP differential pressure sensor. This sensor is ratiometric, e.g. the supply voltage directly influences the output voltage on a 1:1 basis. To ensure accuracy, a precision calibrated supply is required, with a long term stability of a few PPM. Looking at the schematic diagram, a cheap ready made packaged miniature, A.C. mains to 12V supply feeds an LT1021-5 reference IC. This device guarantees that the +5V fed to the pressure and the temperature sensors will have a drift of less than 100 uV (microvolts) in the long term (years).
Looking at the MXP5100DP specification, this device has an excellent linearity in the 15 – 115 KPa (150 – 1,150 HPa) pressure range. The Barograph pressure range is 600 to 790 mmHg (millimeters of mercury) which translates into 800 to 1,053 HPA. The sensor is differential, with input P1 accepting positive pressure and input P2 accepting negative pressure. Unfortunately the excellent linearity of the device does not go to zero volts, but it has a typical pressure offset of 200 mV: e.g. with equal pressure on both ports, output voltage is 200 mV and then linearly increases with pressure differential with a typical sensitivity (in air!!) of 4.5 mV/HPa (45 mV/KPa). Therefore for our pressure range of 800 - 1,053 HPa, our sensor would output a voltage of:
800 * 4.5 = 3,600 mV ------------TO---------- 1,053 * 4.5 = 4,740 mV
In theory for the Barograph application, the sensor should be pre-loaded at half range (933 HPa – 4,200 mV) by creating a vacuum at the P2 port. Variations in barometric pressure sensed at the open P1 port would then swing the output within the range shown above.
In practice it is possible to create a suitable vacuum at port P2 (a syringe is suitable) but it is impossible to maintain such a vacuum in the long term, because of the pressure differential with the external world. To obtain long term sensor stability it is necessary to keep all ports at near barometric pressure, then no pressure leaks are possible.
The most stable configuration is obtained by plugging the P2 port (which then sits at the barometric pressure present at the moment in which the plug was installed) and pre-loading the P1 port with a suitable liquid medium. Barometric pressure will then impinge on the liquid column of port P1 and pressure variations will be read by the sensor. In theory the best liquid for pre-load should be Mercury, which is unfortunately no longer available commercially: with Mercury a pre-load column height H of 20 cm (8”) would suffice, although it would be necessary to investigate Mercury side effects to the inner of the pressure cell (corrosion?).
The pre-load has been obtained with a column with H = 2.2 m (86,6”) filled with Vaseline oil. The sensor sensitivity is influenced by the medium and it was found that it doubles with Vaseline oil. Hence the system operating point has been fixed with pre-load at 1,000 mV and Picolog calculates pressure with a two part expression, one part swinging with output voltage variation, while the other part is fixed. More on this point in future Technical Descriptions. (REVIEW NOTE: The pressure calculation expression has been modified to three parts, as explained in the following Technical Descriptions)
Even though the sensor and the precision power supply have excellent temperature stability, oil density changes with temperature, so that temperature variations would affect cell output voltage, altering reading accuracy. For this reason a temperature sensor reads the oil temperature and corrects cell output voltage accordingly. This correction is done in the math channels of Picolog calculating the Barograph pressure in mmHg. Looking at the schematic, the LM35 temp. sensor output is loaded with R51 and noise is suppressed with C51. Temp. sensor output, through the DrDAQ Universal Interface, reaches EXT.3 Input. Noise of pressure sensor output (pin 1) is filtered by R52/C52. Sensor output is read by the DrDAQ at EXT.1. The EXT.2 input monitors the high precision +5V supply voltage and is useful in detecting the (unwanted) occurrence of long term shifts, either in the supply, or in the DrDAQ. The two resistors R53/R54 halve the value of the supply voltage to bring it within the range of the DrDAQ (2.5V).
Next posts will be as follows:
TECHNICAL DESCRIPTION 2 will provide practical and mechanical information.
TECHNICAL DESCRIPTION 3 will show how the Picolog channels must be programmed.
Attachments
Preload-1.JPG
1. General view of Barograph under construction
Preload-2.JPG
2. Pre-loading with oil
Preload-3.jpg
3. Pre-loading with oil
Last edited by Glovisol on Sun May 03, 2015 4:45 pm, edited 4 times in total.
Glovisol
Advanced User
Advanced User
 
Posts: 114
Joined: Sat May 11, 2013 10:03 am
Location: San Marzano Oliveto, Asti, Italy

Re: DrDAQ & Sensor make precision barograph (TECH DATA 2)

Postby Glovisol » Thu Apr 30, 2015 9:25 am

TECHNICAL DESCRIPTION 2

The MXP5100DP Pressure Sensor is mounted on the Barograph Sensor Assembly. This unit is assembled on a thick wood plank to achieve rigidity and stability of mechanical dimensions. Assembly drawing with dimensions and other details is enclosed. NOTE: it must be understood that pipe diameter dimensions are NOT CRITICAL. The only important dimension being column height H. Therefore pipes of any diameter can be used, as long as a tight fitting is obtained to ensure absence of oil leaks.

The Barograph must be protected by sudden air flows and temperature variations and the best solution is to place it inside a big diameter (120 mm - 5") plastic pipe, as those used for sewerage or for large electrical installations. Bottom and top of this pipe are closed with a plug having a small (3 mm dia.) hole. In this way the Barograph assembly is protected from sudden external environmental changes. For this reason dimension "P" is left open, as it depends on the diameter of large pipe being available.

Once the Barograph Sensor Assembly is assembled and cabled as per schematic, it must be filled with vaseline oil. The assembly must be put in an horizontal position with top slightly raised and temp. sensor LM35 not inserted (pipe B must still be disconnected from the pressure sensor and kept in a higher position, as shown in photos of previous post). Then pipes A & B are filled with oil, pipe B is connected to the pressure sensor. Place also about 20 drops of oil into pipe C, making sure that oil drops into Port 2 of the sensor, then plug pipe C.

Secure pipes B & C to the pressure sensor with electrician's bands pulled tightly and place the entire assembly in a vertical position. With the LM35 sensor inserted, oil level at top should be approx. 50 mm (2") from top.

IMPORTANT: when plugging pipe C some extra positive pressure is exerted on Input P2, which affects initial system operation. To minimise this problem, keep a large part of the pipe squeezed flat while inserting the plug.

Before testing, it must be understood that the assembly must rest in a firm vertical position for at least one week/ten days to allow for pressure at P2 to stabilise and to eliminate all minute and invisible air bubbles in oil. My experience is that stable readings are obtained only after such a period of time.

With the dimensions shown the REFERENCE VOLTAGE OUTPUT of the pressure sensor is:

OUTPUT VOLTAGE = 1,000 mV @ 18°C and @ 730 mmHG (973.1 HPa)
Attachments
7. Baro mech.JPG
Barograph Sensor Assembly and Technical data
Glovisol
Advanced User
Advanced User
 
Posts: 114
Joined: Sat May 11, 2013 10:03 am
Location: San Marzano Oliveto, Asti, Italy

Re: DrDAQ & Sensor make precision barograph (108 HOUR RECORD

Postby Glovisol » Fri May 01, 2015 11:49 am

RECORDED TEST RESULTS - 108 HOURS

A recording run of over 108 hours is attached. Note that for the first time the sampling interval is 10'. In normal operation the best sampling interval should be 20 to 30'. Previously published results were for much shorter intervals, because short term data was needed to optimise Barograph parameters. The original .PLW file is also uploaded, from which all operating parameters & data can be inspected by means of Picolog Player. Using the Player facility it is possible to enlarge/change the displays at will, as well as transfer the Spreadsheet recorded data to Excel or other databases.

A comparison with other baromeric pressure measurement sources is also attached which demonstrates excellent Barograph performance. Barograph data is compared to Official Data of Turin's Observatory, published on site and to a classic, in house, Torricelli type Mercury barometer. Here data was taken at 12 Hour intervals for the last 60 Hours with a cal. factor of 4.

Next post will show all Picolog equations/calculation data.
Attachments
Barometer 900a.PLW
108 Hour Run downloadable .PLW file
(28.07 KiB) Downloaded 128 times
Barotrue 900e Scripted (Medium).jpg
PC Screen of 108 Hour run @ 10' sampling interval
8. 60 Hr tests coomp..JPG
60 Hour comparison between Barograph and other measurement sources
Glovisol
Advanced User
Advanced User
 
Posts: 114
Joined: Sat May 11, 2013 10:03 am
Location: San Marzano Oliveto, Asti, Italy

Re: DrDAQ & Sensor make precision barograph (TECH DATA 3)

Postby Glovisol » Sat May 02, 2015 9:40 am

TECHNICAL DESCRIPTION 3

During testing a way to better and to simplify at the same time Barometric Pressure compensation vs. Temperature was found. Compensation to cell output voltage has been eliminated and compensation is directly applied to the Barograph mmHG math channel, as shown below.


3.1 BAROGRAPHY MATH CONVERSION EXPRESSIONS
Barometric pressure, after the invention of Torricelli’s Mercury Barometer, is expressed in mmHg (height of Mercury column in mm). To convert mmHg to HPa (Hectopascals) mostly used today:

HPa = mmHg * 1,333 [1]

Barometric pressure (B.P.) is influenced by height above sea level (S.L.). For this reason Barometers and Barographs can be calibrated in two ways: either to read B.P. referred to S.L. regardless of height (Relative Pressure Pr) or to read B.P. at the height above S.L. (Absolute Pressure Pa). All measurements shown in our posts are done at Absolute Pressure at the height of my location, which is 220 m. above S.L. Pressure variation with height is not linear and the conversion expressions are:

Pr = Relative Pressure (Pressure at sea level) - HPa.
Pa = Absolute Pressure (Pressure at height Ha) - HPa.
Ha = height above sea level in metres.

Pa = Pr * (EXP(-Ha/7000)) = Pr * e(-Ha/7000) NOTE: expression following e is the exponent [2]
Pr = Pa / (EXP(-Ha/7000)) = Pa / e(-Ha/7000) NOTE: expression following e is the exponent [3]

Expressions [1], [2] and [3] are easily calculated by an Excel spreadsheet. For example let Ha=140 m; Pr=1011,9 HPa,
then:

Pa = 1011,9 * e(-140/7000) = 1011,9 * e(-0,02) = 1011,9 *0,980199 = 991,9 HPa


3.2 PICOLOG CHANNEL SETUP
Setup information and parameters for all Picolog channels are shown in Table 1 (uploaded).
Channel descriptions are given in logical sequence and not according to Table 1. (REVIEW NOTE: Table 1 has been amended twice to eliminate errors. With this note I am posting VER. 3 which should be O.K.)


CHANNEL 1, Sensor Temperature, is EXT. 1. Scale is 10 – 30°C and the calibration equation (set in “Options”) has the form: X * T, where T is the calibration factor. Before placing the LM35 sensor into the oil column, check reading against a mercury thermometer and find calibration factor. I found: T=99.

CHANNEL 5, Cell Supply Voltage, is EXT. 2. As already explained, this voltage must be accurate and have a long term stability of 100 uV or better. The ideal value is V = 4,990 mV, which can be obtained by trimming the adjust pot on circuit of the LT1021 Voltage Reference IC with the help of an accurate DVM. This voltage is reduced by a factor of 2 by the DrDAQ resistor input network to the value of 2,495 mV: hence the channel equation X*2. Channel scale is 4,900 to 5,000 mV. The absolute value of the supply voltage is not so important, as long as it does not drift with environmental changes and time.

CHANNEL 7, Cell Voltage, is EXT. 3. As written in Post TECHNICAL DESCRIPTION 2, the reference point of Cell Voltage, with dimensions shown, should be 1,000 mV @ 18°C and 730 mmHG or 973.1 HPa. With other dimensions and/or different height and/or different conditions, one could find a pressure different from 730 mmHg, but system must be set up in any case, by finding a given cell voltage at known conditions: it does not matter if the reference voltage is different, as long as conditions (Pressure, Temperature and Height) are known. This operation must be attempted after having allowed the system to settle vertically for a minimum of seven days. Pressure data can be obtained from any local weather forecasting internet site and/or from an accurate mercury barometer. In general sites give updated Pr and Pa values every few minutes. Knowing Ha, Pr data from the internet site can then be easily converted to the local Pa value using [2].

To set up channel 7, first calibrate voltage value shown by Picolog by comparison with an accurate DVM. In my case the calibration factor is 1,007. By using the expression: X*1007 in Picolog, in my system the cell voltage readings in Picolog and on the DVM are the same. Then all said conditions must be noted.

CHANNEL 4, Ambient Temperature, is the DrDAQ own temperature channel. This channel also must be calibrated against an accurate Mercury thermometer: in my case the correction equation is: X*0.8, as the DrDAQ temp. reading is 20% higher than true. Bear in mind that, in operation, Cell Temperature will always be higher than Ambient Temperature, because the LM35 sensor normally sits 1.5 metres higher than the DrDAQ.

More setup information on next post.
Attachments
9. Table 1 - Picolog Setup.JPG
Table 1, amended VER. 3 - Setup expressions & Data
Last edited by Glovisol on Mon May 04, 2015 10:35 am, edited 3 times in total.
Glovisol
Advanced User
Advanced User
 
Posts: 114
Joined: Sat May 11, 2013 10:03 am
Location: San Marzano Oliveto, Asti, Italy

Re: DrDAQ & Sensor make precision barograph (TECH. DATA 4)

Postby Glovisol » Sun May 03, 2015 3:16 pm

Barometer 1500a.PLW
This .PLW file shows the complete working of temp. compensation
(28.07 KiB) Downloaded 135 times


TECHNICAL DESCRIPTION 4

3.2 PICOLOG CHANNEL SETUP (continued)

UNDERSTANDING THE PHYSICS OF THE PRESSURE CELL

CHANNEL 2, Barograph, is a Calculated Parameter Channel. This channel calculates the actual, temperature compensated, Barometric Pressure in mmHg, with the following formula:

Pa = (A * ((200/1000)) + (λ * 0.5) + (B * μ) [4]

With reference to Table 1 of previous post, the actual factors and constants are as follows.

A = Vc - Cell Voltage from Channel 7
B = To - Sensor Temperature from Channel 1
Pressure Calibration Factor: λ = 1,020
Temperature Compensation Factor: μ = 1.05

Here below is the detailed explanation for expression [4].

The basic Barometric expression relating cell output Vc in mV to Barometric pressure P’a in mmHg is:

P’a = 0,2 * Vc or Vc = Pa/0.2 or Vc = Pa * 5

To bring the pressure value to the correct level, we use factor: λ * 0.5, or 1,020 * 0.5 = 510. (I prefer to always use values conceptually near the thousands of the theoretical mV value of the pressure cell output).

Temperature sensitivity, due to density variations of the Cell Assembly filled with Vaseline Oil, has been measured to be:

-5mV/°C (negative temperature coefficient).

This means that, at constant Barometric Pressure, cell output decreases by 5 mV for any 1°C of oil temperature rise and consequently the value of Baromeric Pressure shown by Picolog would change by:

0,2 * (-)5 = -1 mmHg (pressure error due to oil temperature variation).

It is now possible to develop a simple expression to eliminate this error. Luckily the Barograph works indoors and must withstand a small temperature variation of To = ± 2°C in the range: 18 – 22 °C, the mean value being 20°C. If we now add the oil temperature value, multiplied by a convenient factor μ, to the basic expression, Picolog will take the temperature variation into account as follows:

Pa = (Vc*(200/1000)) + (1020*0.5) + (To* 1.05)
Vc = 1000 mV
To = 20°C
μ = 1.05
Pa = 200 + 510 + (20 * 1.05) = 200 + 510 + 21 = 731 mmHg

If now temperature To increases to 21°C:

Pa = (200 – 1) + 510 + (21 * 1.05) = 199 + 510 + 22.05 = 731.05 mmHg (error: 0.05 mmHg)

And if To decreases to 18°C:

Pa = (200 + 2) + 510 + (18 * 1.05) = 202 + 510 + 18.9 = 730.9 mmHg (error 0.9 mmHg)

This is an excellent result, which can be further improved by small empirical variations of factor μ. Furthermore this compensation will hold, with minor errors, even throughout a larger temperature range.

But we must also consider the delay between Vaseline oil temperature variation and change in the density of the oil. The uploaded file shows how these parameters vary with temperature changes. To obtain this file, the sampling interval was reduced to 30” (from 10’) and oil temperature (by turning a stove on in the room) was increased from 20.79°C to 21.89 °C in 1,530” or 25.5’. It took 2,250” or 37.5’ for the oil density to reach a good, if not complete, equilibrium. Therefore the delay between oil temperature rise and pressure cell voltage change is in the order of 720”, or 12’. To avoid fluctuations in readings and in graphs due to ambient temperature variations, the sampling interval should be at least twice this value, e.g. 24’. In the graph these fluctuations can be easily seen, due to the very short sampling interval.

In reality oil density carries on changing, even though at a much smaller rate, up to 2,850”, far after the time the heat has been stopped and temperature is no longer increasing, therefore the maximum delay is 1,320” or 22’: accordingly the optimum sampling interval should be set at 30 to 40’, for maximum accuracy.

All above described data can be seen in 30" intervals on the spreadsheet of the enclosed .PLW file, which can be opened with the Picolog Player facility.

The next post will describe the remaining channels/parameters.
Attachments
Barotrue 1000a - Undercomp. example.jpg
Example of undercompensated Barograph Equation
10. Barotrue 1500 density delay.jpg
Showing heat propagation delay in the oil column
Last edited by Glovisol on Mon May 04, 2015 10:51 am, edited 1 time in total.
Glovisol
Advanced User
Advanced User
 
Posts: 114
Joined: Sat May 11, 2013 10:03 am
Location: San Marzano Oliveto, Asti, Italy

Re: DrDAQ & Sensor make precision barograph (TECH. DATA 5)

Postby Glovisol » Mon May 04, 2015 8:22 am

TECHNICAL DESCRIPTION 5

3.2 PICOLOG CHANNEL SETUP (CONTINUED).....

CHANNEL 3, Barograph HPa, is a Calculated Parameter Channel. It uses the simple conversion formula:

Barograph HPa = (Pa mmHg) * 1,333


CHANNEL 6, Differential HPa, is a Calculated Parameter Channel. To set it up, find the mean absolute pressure Pa at your altitude. Since the mean pressure at sea level is 760 mmHg, or 1,010 HPa, you can use [2] to find constant k which will calculate difference between mean and actual pressure.

EXAMPLE:

Altitude above S.L. Ha = 260 m.
Constant k = mean Pa @ given Ha = Pr * (EXP(-Ha/7000)) = 1,010 * (EXP(-260/7000) = 1,010 * 0,963538 = 973 HPa. If now actual Barometric Pressure Pa is 976 HPa (CHANNEL 3) Picolog calculates differential HPa as:

Differential HPa = 976 - 973 = 3 HPa



3.3 SETTING SYSTEM UP

Here below the setup procedure. The advantage of reproducing the Barograph with the same physical dimensions as given, is that cell voltage and calibration factors will be equal to or very near the values found in my work.

Once the Sensor Assembly has stabilized in a suitable vertical position (7 to 10 days) the precision power supply has been set up at V = 4,990 mV and all interconnections are made, DrDAQ Channel 4 should give an Ambient Temperature within the range 18 – 24 °C. The ideal being to have an ambient at 18°C. Oil temperature shoud be approx. 0.4 to 0.6 °C higher, due to the fact it is reading warmer air sitting 1.5 metres higher in the room. With these conditions, Channel 7, EXT. 3, should read a voltage in the range: 940 to 1060 mV. Using a good quality DVM, check the DrDAQ reading and alter calibration constant of Channel 7 to make the two readings coincide: in my case the calibration constant (which should be 1,000 with a calibrated DrDAQ spot on) was 1,007.

3.3.1 CHARACTERISTIC OF SENSOR ASSEMBLY

At this point local absolute pressure Pa must be known. Then we can take note of the characteristic of the Sensor Assembly.

Taking the previous EXAMPLE:

Our altitude is Ha = 260 m. and the Observatory on Internet gives us a relative (S.L.) pressure Pr = 1,010.0 HPa
Then our local pressure Pa = 973 HPa which translates into Pa = 973/1,333 = 730 mmHg. Giving EXAMPLE data, we can record the following set of references:

Ambient Temperature Ta = 18.5° C.
Oil Temperature To = 19.0 °C.
Local Pressure Pa = 730 mmHg.
Supply Voltage = 4,990 mV.
Cell Voltage = 960 mV (as read on the DVM and on Picolog, Channel 7, EXT.3)

3.3.2 SETTING UP CHANNEL 2, BAROGRAPH in mmHg

Following the previous EXAMPLE, we must now find the Pressure Calibration Factor λ in Channel 2 equation: this is very simple, because we must choose λ to obtain the reading of 730 mmHg on this channel, bearing in mind that A, in this equation is cell voltage: let us start with
λ = 1,000

Pa = (960 * (200/1000)) + (1,000 * 0.5) + (19 * 1.05) = 192 + 500 + 19.95 = 711,95, or 712 mmHg: this value is too low.

Therefore 730 – 712 = 18 and the correct λ = (500 + 18) * 2 = 1,036 with this value:

Pa = 192 + (1036 * 0.5) + 19.95 = 192 + 518 + 19,95 = 729.95, or 730 mmHg.

This channel is now calibrated and temperature compensated and Barometric Pressure recording can start.


Conclusions in the next post.
Glovisol
Advanced User
Advanced User
 
Posts: 114
Joined: Sat May 11, 2013 10:03 am
Location: San Marzano Oliveto, Asti, Italy

Re: DrDAQ & Sensor make precision barograph (CONCLUSION)

Postby Glovisol » Tue May 05, 2015 7:59 am

12. 136 hour test comparison.JPG
CONCLUSION

I do not know if other users will consider this project practical enough to be reproduced, considering the present general availability of cheap Barometric devices. In my opinion what this project has proven is the incredible power and versatility of the DrDAQ / Picolog combination, available at a very affordable price.

The multi-parameter, multi-channel measurement and storage capability of this combination makes for an extremely powerful tool in many design applications and effortlessly allows the solution of many circuit development problems. Stable and accurate measurement and storage of many different parameters in time allows the designer to devote himself to other activities, while the DrDAQ does its work..

Just one example: after the theory of temperature compensation was devised and the constant μ = 1.05 was found by calculation, the DrDAQ allowed an extremely fast check: three Calculated Parameter math channels were set up: one with μ = 0.85, another with μ = 1.05 and another with μ = 1.25. Then the results of the different compensation constants were immediately available at the same time on the same PC screen!!

Further info in the uploads.
Attachments
3. Assy with cover (Large).jpg
Sensor assy. protection detail
2. Bottom closeup 2 (Large).jpg
Sensor arrangements
1. Complete assy (Large).jpg
Complete Sensor assembly
Barotrue 1700d.jpg
Barograph 36 Hour run
Barotrue 1700c.jpg
Barograph 36 Hour run
Glovisol
Advanced User
Advanced User
 
Posts: 114
Joined: Sat May 11, 2013 10:03 am
Location: San Marzano Oliveto, Asti, Italy

Re: DrDAQ & Sensor make precision barograph (GENERAL INFO)

Postby Glovisol » Wed May 20, 2015 9:08 am

IMPORTANT MODIFICATION TO THE DrDAQ BAROGRAPH

Additional work on the DrDAQ Barograph showed that the unwanted pressure cell output voltage variation did not agree with the calculated oil density variatiuon with temperature: in fact output voltage change was notably greater than what caused by density variation.

Further research found that the additional temp. variation was caused by pipe C which pugs Port 2. An increase in ambient temperature causes the residual air inside the pipe to expand, parasitically reducing cell output voltage, because of increased pressure at differential Port 2 and vice-versa. This mechanism has nothing to do with oil density variation.

The simple modification is to remove pipe C and plug Port 2 directly with conglomerating tape. This modification is shown in the uploaded drawing. Once the modification is done, system must be allowed to settle for a minimum of one week. Future posts will show non only the new temp. compensation expression, but a very important simplification and improvement of the barograph.
Attachments
7.1. Baro mech mod..bmp
IMPORTANT MODIFICATION
Glovisol
Advanced User
Advanced User
 
Posts: 114
Joined: Sat May 11, 2013 10:03 am
Location: San Marzano Oliveto, Asti, Italy


Return to USB DrDAQ

Who is online

Users browsing this forum: No registered users and 0 guests