Real EE advice solicited

Projects General Guidance
1

I want to use the MPX2202 with an Arduino A to D and it needs to be amplified. I plan to use it to monitor water pressure and depth

The manufacturer has thoughtfully created an app note on how to do this. http://www.freescale.com/files/sensors/doc/app_note/AN1325.pd

On page 4 a schematic shows the recommended method using an MC33272 op amp. However, they show an 8v power input which I don't want. I want to use the 5v on the Arduino board. It appears that I will still have plenty to measure.

Can someone who knows what they are doing tell me what the Resistor values should be to use 5v, or tell me why I shouldn't.

Thanks in advance.

2

My first question is why would you select and think this sensor is rated and designed for liquid service. It's not, it's designed for gas (dry air at that) pressure only. From the datasheet:

The MPX2202/MPXV2202G series pressure sensor operating characteristics and internal reliability and qualification tests are based on use of dry air as the pressure media. Media other than dry air may have adverse effects on sensor performance and long term reliability. Contact the factory for
information regarding media compatibility in your application.

You have first focused on the electrical side of this sensor rather then first checking if it's compatible with the media you are trying to measure.

Lefty

3

It is commonly used for liquid and they describe silicone lubricants to prevent water from contacting the sensor element.

A flexible tube is used to connect with the nipple on the device.

4

growerdick:
It is commonly used for liquid and they describe silicone lubricants to prevent water from contacting the sensor element.

A flexible tube is used to connect with the nipple on the device.

Well if you say so. I worked in a oil refinery for decades and specified and worked on many types of pressure sensors and one thing in common to all pressure sensors rated for liquid service was the 'raw' sensor had a metal diaphragm that separated the sensing element from the liquid media to be sensed. The space between the active sensor element and the metal barrier was 'liquid packed' with some inert liquid material (silicon was common) that was compatible with the sensor element. Also proper external placement of the sensor relative to the liquid in the vessel or container being measured and remote sensing leads (your flexible tube) has to be 'liquid packed' as any mixture of air (gas) and the liquid would corrupt the reading as liquid is not compressible and will accurately 'transmit' the pressure but gas is compressable and will screw up the pressure the sensor 'sees' Vs the actual vessel or container pressure. So be sure of your application and that any liquid coming into contact with the sensor element is compatible with the sensor element material. As stated the manufactures datasheet states it's rated for dry air service only, so proceed at your own risk.

Lefty

5

They have an app note describing using a tube from water, which is what I need it for. The tube will be full of air and the device above the water level.

Is there any reason why using 5v will cause problems. I am not sure of the reason behind the 8v.

Thanks.

6

growerdick:
They have an app note describing using a tube from water, which is what I need it for. The tube will be full of air and the device above the water level.

Do you have a link for that? Pressure will not be accurately transmitted up to a sensor mounted higher then the full water level if there is air in the sensing lead between the liquid level and the sensor, it's physically impossible, the air will compress some but not 'transmit' the true pressure value. But maybe the app note is showing something different then what you think it is? Not trying to be difficult, but I have decades of experience in the reading of vessel pressure and vessel liquid level measurements based on liquid head pressure and the various methods needed to make such measurements.

Is there any reason why using 5v will cause problems. I am not sure of the reason behind the 8v.

Well the datasheet I read stated a typical sensor operating voltage of 10vdc with a maximum of 16vdc, but it failed to state a minimum operating voltage for the sensor, so I really can't tell you it would work as designed with 5vdc, but I would suspect it won't as I would think they would state that as 5vdc is a very popular operating voltage.
HTTP 301 This page has been moved
Lefty

Thanks.

7

You have information from a "Real" EE and yet you know enough to argue with someone who has direct information. His advice is good, based on my own experience with water control devices for the irrigation inudstry. If your own opinion is that his advice doesn't fit your model... Go build it and maybe learn... It would seem that is the only way you WILL LEARN... Some people are like that..

Bob

8

Thanks Lefty..

9

This need was researched several years ago and a solution is now needed. I had not looked at the information for over a year.

When I saw the schematic first referred to my question was why not 5v. That let me to posting the question. I am not trying to be difficult. I am trying to make something that will work.

Here is what I got from their tech support when this issue first came up:

Following to your inquiry, we would recommend to use Parker O-lube
silicone grease or DMS-T46 or T51. This type of grease is used by most
of our customer without problems. In fact the basic recommendation are
to use a silicone oil (or preferably grease) with high viscosity and
high molecular weight. In this case the size of the molecules are big
enough to limit the penetration of the grease inside our protective
silicone gel which is over the die. In terms of contaminants, the
silicon grease must be free of halogenures (Cl content < 50 ppm) to
reduce the risk of bond pad corrosion. On the other hand, don't forget
that whatever the material you will use, as soon as you put something on
our gel you have a high probability to see some offset drift. This is
coming from additional mechanical stress and/or gel swelling. The amount
of gel and global mechanical design are usually also part of the offset
drift.

Protecting pressure sensors with a silicon grease is definitely the best
way to use our products in harsh media. We don't have so much experience
in how to implement this solution 'industrially'. The grease fill is
done under vacuum to avoid air bubbles. Usually, an offset calibration
is necessary because the grease brings some mechanical stress that may
shift the offset value. This solution was tested and seems conclusive.

Usually in a water measurement application, a head tube is used. This is
simply a piece of tubing between the sensor and the container in which
the water level or pressure needs to be measured.
Air is trapped in the tube and this isolates the sensor from the liquid
water.

As said, you can use any silicon oil or preferably grease from other
suppliers with high viscosity and high molecular weight.

I would like refer to the following application note that would be worth
to be considered.

AN3728: Media Compatibility for IPS PRT Pressure Sensors
http://www.freescale.com/files/sensors/doc/app_note/AN3728.pdf

AN1516: Liquid Level Control Using a Pressure Sensor
http://www.freescale.com/files/sensors/doc/app_note/AN1516.pdf

AN1950: Water Level Monitoring
http://www.freescale.com/files/sensors/doc/app_note/AN1950.pdf

The last app note discusses using the sensor in a washing machine.

My question was a honest attempt to educate myself. I don't understand the seeming ridicule from Bob.

10

You were arguing with a person who has a great deal of time "In the Foxhole" both Lefty and I are aware that the sensor might be applied in the manner described, posting the "App Notes" is good education, however the situation isn't changed by your research. In the "Real World" app notes frrequently belong on the Loo.. better purpose for them there. Your solution would be inapropriate for 1. Potable Water (toxic contaminants), 2. Any media containing active halides

halogenures (Cl content < 50 ppm)

and this includes the florides like freon and both have high solution mobility. 3. Any petroleum based liquid (solvent). 4. There is a good reason that the Mfr suggests or specifies the device at some particular operating voltage (calibration) use at other lower voltages might be possible BUT are you ready to completely characterize the sensor at the new voltage? one side effect might well be compression or non linearity in the returned data. There is an old (when I was young)
Maxin that says... There is never enough time to do the job right and always time to do it over again" This is the reason why Lefty told you to buy the proper sensor rather than try to make a dry sensor work with a liquid. You stated thhat there would be a "Dry Tube" and Lefty mentioned that Air will compress but a liquid won't... so your returned data is anyones guess. BTW the link to the App Note is broken, I wonder why. Lefty has years of knowledge of what works and why. The advice is good and free as well so why not use it. 8 Volts is easy to do with a simple boost mode switcher and some light filtering 2 47uF caps and a 100 uHy choke should work well. My apparent anger isn't at all just my impatience with people who argue with answers that don't fit their own world model.
Good Luck in the Contest...

Bob

11

Docedison:
You stated thhat there would be a "Dry Tube" and Lefty mentioned that Air will compress but a liquid won't... so your returned data is anyones guess.

I would disagree with Lefty as well; using a dry tube as the OP describes will give a valid reading.

Not sure if I can come up with a good real world example but I'd start by pointing out the millions of homes in the world with private water wells that rely on a pressure tank half-filled with air and read the water pressure from the top of that column of air. Should the air bladder in that pressure tank burst and all air be lost from the tank it doesn't result in any change to the water pressure in the home. Well, the water pump starts cycling more frequently but that's another issue ...

12

Chagrin:

Docedison:
You stated thhat there would be a "Dry Tube" and Lefty mentioned that Air will compress but a liquid won't... so your returned data is anyones guess.

I would disagree with Lefty as well; using a dry tube as the OP describes will give a valid reading.

You are free to disagree, but that in itself doesn't make me or you wrong or right. I gave my reasons why it will not work and you just say it will regardless but state no technical reason or description of how it can be made to work, so there we are.

Not sure if I can come up with a good real world example but I'd start by pointing out the millions of homes in the world with private water wells that rely on a pressure tank half-filled with air and read the water pressure from the top of that column of air. Should the air bladder in that pressure tank burst and all air be lost from the tank it doesn't result in any change to the water pressure in the home. Well, the water pump starts cycling more frequently but that's another issue ...

The issue is he seems to want to be able to measure the water level of a vessel by sensing the head pressure of the water, not the vessel air pressure above the water level, and I assumed it would be a simple vessel only under atmospheric pressure. Those are two different measurements requiring different sensor protection and mounting methods. If it's a pressurized vessel that one wants to measure the liquid vessel then one needs to use a two port differential pressure sensor with the sensor mounted at or below the bottom of the vessel with the high pressure lead (which will be liquid packed) sensing the total vessel + liquid head pressure and the low pressure lead running to the top or above the vessel which will cancel out the vessel pressure resulting in just the pressure difference created by the head pressure of the liquid. This low side 'dry' lead is subject to condensation buildup resulting in liquid gradually filling up the dry leg and causing an error in the true level measurement. Measuring the vessel pressure above the water level can be done with a pressure sensor mounted higher then the vessel and of course the liquid level cannot reach the sensor, but this measurement in no way tells you the level of the liquid in the vessel.
There is a method to read liquid level above a vessels height using a 'dip tube' attached to a 'dry lead' mounted higher then the vessel. It's one of a very few methods of determining liquid level in underground buried vessels, but it requires an adjustable air pressure regulator to be able to force air bubbles into the 'dip tube' at the liquid levels highest possible level. What one is reading is the back pressure changes that result from the height of the liquid level in the vessel. It tends to not be a real accurate method but can be made to work. Texmate: Programmable Meter Controllers, Panel Meters, and Bargraphs
Lefty

13

Further searching turned up the device best suited for my need, MPXV5100GP for measuring water level. It runs on 5v, has integrated amplification and puts out .2 to 4.7, and Digikey has them for $11.81 q1

Here is data sheet: http://www.freescale.com/files/sensors/doc/data_sheet/MPX5100.pdf

Thanks everyone for responses.

14

There is ONE small thing to verify and that is whether the chosen device is characterised for "Potable" water (water for human consumption). There is a diffferrence as one can be used for most fluids including "waste"water (sewage, runoff water) BUT NOT water for people and the other is used only for drinking water. They can look identical from the outside but the internals are of different materials. I hope this helps.

Bob

15

growerdick:
Further searching turned up the device best suited for my need, MPXV5100GP for measuring water level. It runs on 5v, has integrated amplification and puts out .2 to 4.7, and Digikey has them for $11.81 q1

MPXV5100GP NXP USA Inc. | Sensors, Transducers | DigiKey

Here is data sheet: http://www.freescale.com/files/sensors/doc/data_sheet/MPX5100.pdf

Thanks everyone for responses.

Better but the datasheet still carries the advice:

The MPX5100 series pressure sensor operating
characteristics, and internal reliability and qualification tests
are based on use of dry air as the pressure media. Media,
other than dry air, may have adverse effects on sensor
performance and long-term reliability.

It's better in that they attempt to protect the sensing element with a silicon gel, but note that the metal diaphragm above the gel has a perforated sensing hole that allows the measured media to come in direct contact with the silicon gel. On industrial quality liquid service pressure sensors the metal barrier completely isolates the measured media from the sensor/gel filled measurement chamber, so not my first choice. But hey it may work for you, but keep in mind that water, even potable, is a pretty aggressive material and could very well raise problems with the silicon and sensor element over time. This would not fly in professional industrial applications where long term reliability and accuracy is worth the cost of proper chemical isolation from the measured media.

Good luck with your project;

Lefty

16

Hard Water??

Bob

17

retrolefty:

growerdick:
It is commonly used for liquid and they describe silicone lubricants to prevent water from contacting the sensor element.

A flexible tube is used to connect with the nipple on the device.

Well if you say so. I worked in a oil refinery for decades and specified and worked on many types of pressure sensors and one thing in common to all pressure sensors rated for liquid service was the 'raw' sensor had a metal diaphragm that separated the sensing element from the liquid media to be sensed. The space between the active sensor element and the metal barrier was 'liquid packed' with some inert liquid material (silicon was common) that was compatible with the sensor element. Also proper external placement of the sensor relative to the liquid in the vessel or container being measured and remote sensing leads (your flexible tube) has to be 'liquid packed' as any mixture of air (gas) and the liquid would corrupt the reading as liquid is not compressible and will accurately 'transmit' the pressure but gas is compressable and will screw up the pressure the sensor 'sees' Vs the actual vessel or container pressure. So be sure of your application and that any liquid coming into contact with the sensor element is compatible with the sensor element material. As stated the manufactures datasheet states it's rated for dry air service only, so proceed at your own risk.

Lefty

Hello Lefty,
As an independent consultant, I´m preparing various Test Runs for an Atmospheric Distillation Unit, in different Operating models (2 towers in parallel or in series) and in various production patters (several products alternatively produced: solvents). As this ADU has very few instruments, for these Test Runs I´m thinking of inserting additional sensors (temperature, pressure and fluids properties) into the ADU, connected to a CPU through Arduino Boards.
I defined the required sensors /transmitters, but I need an Arduino specialist and better if it has petroleum / refinery experiences. Please could you be interested in help? See the Flow Diagram
http://www.phmandar.com/onlinedoc/VASSA%20Esquema%20Proceso%20MFA_%20Test%20Run-%20Rev0c.pdf
PhMandar

18

Water has been described as the "universal solvent".

Why are some replying within the quoted text?