Shop OBEX P1 Docs P2 Docs Learn Events
Pressure sensor 4-20mA — Parallax Forums

Pressure sensor 4-20mA

ArchiverArchiver Posts: 46,084
edited 2004-07-20 23:32 in General Discussion
Hi guys & girls,

I am working on a HVAC project. I have to read in an 4-20mA pressure
sensor (between -0,5 and 7 BAR brand DIXELL). I've searched a lot on
internet and didn't find a proper solution to connect this sensor to
my BS2P-24. I've already made an easy circuit wich contains an 12BIT
ADC and a resistor (150ohms) so I can get max 3V over my adc inputs.
When I read the ADC in, I get VERY different values. My power supply
is extremly stable (checked with an oscilloscope). Is there
somebody who has created a simular project and could give me a
better way to connect the pressure sensor with my stamp?

Kindly regards
Tim
Belgium
Student ICT

Comments

  • ArchiverArchiver Posts: 46,084
    edited 2004-07-08 11:50
    I'm unsure of hte coding sequence of these guys unless they're an analog
    measurement.

    But, since the stamp likes things in 5V, you might want to just throw a
    resistor across that guy there. I know I've heard someone say as much
    before.
    You might actually want to do a search through the archives.

    Just take 5volts and divide by 20mA and you should get a resistor value.
    That's the best I can remember!


    Original Message
    From: "ikenniemandanders" <tld-nv@s...>
    To: <basicstamps@yahoogroups.com>
    Sent: Thursday, July 08, 2004 1:50 AM
    Subject: [noparse][[/noparse]basicstamps] Pressure sensor 4-20mA


    > Hi guys & girls,
    >
    > I am working on a HVAC project. I have to read in an 4-20mA pressure
    > sensor (between -0,5 and 7 BAR brand DIXELL). I've searched a lot on
    > internet and didn't find a proper solution to connect this sensor to
    > my BS2P-24. I've already made an easy circuit wich contains an 12BIT
    > ADC and a resistor (150ohms) so I can get max 3V over my adc inputs.
    > When I read the ADC in, I get VERY different values. My power supply
    > is extremly stable (checked with an oscilloscope). Is there
    > somebody who has created a simular project and could give me a
    > better way to connect the pressure sensor with my stamp?
    >
    > Kindly regards
    > Tim
    > Belgium
    > Student ICT
    >
    >
    >
    >
    > To UNSUBSCRIBE, just send mail to:
    > basicstamps-unsubscribe@yahoogroups.com
    > from the same email address that you subscribed. Text in the Subject and
    Body of the message will be ignored.
    >
    > Yahoo! Groups Links
    >
    >
    >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2004-07-08 12:45
    I did "not" look at the particular sensor you mention, but 4-20 ma to 1-5
    volt is easy.....

    Typically, you would use a precision 250 ohm resistor for that conversion.
    You convert the 4-20 ma to voltage 1 - 5vdc be inserting a resistor from the
    sensor output signal back to ground (single ended inputs on the adc).

    Why does it start at 4 ma or 1 volt? Sensor error / Wiring detection.

    Here is an ASCII schematic.... Hope it comes out well.


    Sensor Power source ----> Sensor --- Sensor Output
    > 250
    ohm|
    > ADC + Input

    |

    |
    Ground

    >ADC - Input


    Hard to tell, but the resistor is mounted "between" sensor output and
    ground.
    Remember to subtract 1 volt when you do your scaling for zero since you
    start at 4ma / 1 volt for a zero reading or whatever the lower scale point
    is of the sensor.

    One other note, 250 ohm resisters can be hard to find or expensive. Use 249
    ohms and compensate in software or use a couple of 120 ohms with a 10ohm in
    series, or a couple of 500 ohms in parallel. it's your preference.

    "If" anyone has a good source for reasonably priced 250 ohm resistors let me
    know please. $2.00 a peice is not reasonable for me.

    Richard







    Original Message
    From: "SB" <steve.brady@r...>
    To: <basicstamps@yahoogroups.com>
    Sent: Thursday, July 08, 2004 5:50 AM
    Subject: Re: [noparse][[/noparse]basicstamps] Pressure sensor 4-20mA


    > I'm unsure of hte coding sequence of these guys unless they're an analog
    > measurement.
    >
    > But, since the stamp likes things in 5V, you might want to just throw a
    > resistor across that guy there. I know I've heard someone say as much
    > before.
    > You might actually want to do a search through the archives.
    >
    > Just take 5volts and divide by 20mA and you should get a resistor value.
    > That's the best I can remember!
    >
    >
    >
    Original Message
    > From: "ikenniemandanders" <tld-nv@s...>
    > To: <basicstamps@yahoogroups.com>
    > Sent: Thursday, July 08, 2004 1:50 AM
    > Subject: [noparse][[/noparse]basicstamps] Pressure sensor 4-20mA
    >
    >
    > > Hi guys & girls,
    > >
    > > I am working on a HVAC project. I have to read in an 4-20mA pressure
    > > sensor (between -0,5 and 7 BAR brand DIXELL). I've searched a lot on
    > > internet and didn't find a proper solution to connect this sensor to
    > > my BS2P-24. I've already made an easy circuit wich contains an 12BIT
    > > ADC and a resistor (150ohms) so I can get max 3V over my adc inputs.
    > > When I read the ADC in, I get VERY different values. My power supply
    > > is extremly stable (checked with an oscilloscope). Is there
    > > somebody who has created a simular project and could give me a
    > > better way to connect the pressure sensor with my stamp?
    > >
    > > Kindly regards
    > > Tim
    > > Belgium
    > > Student ICT
    > >
    > >
    > >
    > >
    > > To UNSUBSCRIBE, just send mail to:
    > > basicstamps-unsubscribe@yahoogroups.com
    > > from the same email address that you subscribed. Text in the Subject
    and
    > Body of the message will be ignored.
    > >
    > > Yahoo! Groups Links
    > >
    > >
    > >
    > >
    > >
    >
    >
    >
    > To UNSUBSCRIBE, just send mail to:
    > basicstamps-unsubscribe@yahoogroups.com
    > from the same email address that you subscribed. Text in the Subject and
    Body of the message will be ignored.
    >
    > Yahoo! Groups Links
    >
    >
    >
    >
    >
    >
    > [noparse][[/noparse]Scanned by AwesomeNet Anti-Virus]
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2004-07-08 12:55
    Sorry, I see you wanted only 3 volts, so if the 150 ohm resistor is used you
    should see 0.6 v for your low end and 3v for you high end. You have to
    subtract the initial .6v from your readings.
    4ma * 150 = 0.6 v 20ma * 150 = 3 v then 3v - 0.6v = 2.4 volts total
    range

    Original Message
    From: "SB" <steve.brady@r...>
    To: <basicstamps@yahoogroups.com>
    Sent: Thursday, July 08, 2004 5:50 AM
    Subject: Re: [noparse][[/noparse]basicstamps] Pressure sensor 4-20mA


    > I'm unsure of hte coding sequence of these guys unless they're an analog
    > measurement.
    >
    > But, since the stamp likes things in 5V, you might want to just throw a
    > resistor across that guy there. I know I've heard someone say as much
    > before.
    > You might actually want to do a search through the archives.
    >
    > Just take 5volts and divide by 20mA and you should get a resistor value.
    > That's the best I can remember!
    >
    >
    >
    Original Message
    > From: "ikenniemandanders" <tld-nv@s...>
    > To: <basicstamps@yahoogroups.com>
    > Sent: Thursday, July 08, 2004 1:50 AM
    > Subject: [noparse][[/noparse]basicstamps] Pressure sensor 4-20mA
    >
    >
    > > Hi guys & girls,
    > >
    > > I am working on a HVAC project. I have to read in an 4-20mA pressure
    > > sensor (between -0,5 and 7 BAR brand DIXELL). I've searched a lot on
    > > internet and didn't find a proper solution to connect this sensor to
    > > my BS2P-24. I've already made an easy circuit wich contains an 12BIT
    > > ADC and a resistor (150ohms) so I can get max 3V over my adc inputs.
    > > When I read the ADC in, I get VERY different values. My power supply
    > > is extremly stable (checked with an oscilloscope). Is there
    > > somebody who has created a simular project and could give me a
    > > better way to connect the pressure sensor with my stamp?
    > >
    > > Kindly regards
    > > Tim
    > > Belgium
    > > Student ICT
    > >
    > >
    > >
    > >
    > > To UNSUBSCRIBE, just send mail to:
    > > basicstamps-unsubscribe@yahoogroups.com
    > > from the same email address that you subscribed. Text in the Subject
    and
    > Body of the message will be ignored.
    > >
    > > Yahoo! Groups Links
    > >
    > >
    > >
    > >
    > >
    >
    >
    >
    > To UNSUBSCRIBE, just send mail to:
    > basicstamps-unsubscribe@yahoogroups.com
    > from the same email address that you subscribed. Text in the Subject and
    Body of the message will be ignored.
    >
    > Yahoo! Groups Links
    >
    >
    >
    >
    >
    >
    > [noparse][[/noparse]Scanned by AwesomeNet Anti-Virus]
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2004-07-08 15:46
    EXCUSE THE CAPS.

    4-20 MA IS A FAIL-SAFE FEATURE. IF YOU HAVE A BROKEN CONNECTION YOU GET
    O WHICH IS A FAULT CONDITION.

    YOU SCALE YOUR READING TO 4 MA = O AND 20 MA = FULL SCALL SO IF A DEVICE
    WAS 4-20 MA AND 150 VDC, THEN 4 MA WOULD BE 0 VDC AND 20 MA WOULD BE 150VDC.

    73
    SPENCE
    K4KEP

    essage: 24
    Date: Thu, 8 Jul 2004 06:45:34 -0500
    From: "Richard" <rwskinner@a...>
    Subject: Re: Pressure sensor 4-20mA

    I did "not" look at the particular sensor you mention, but 4-20 ma to 1-5
    volt is easy.....

    Typically, you would use a precision 250 ohm resistor for that conversion.
    You convert the 4-20 ma to voltage 1 - 5vdc be inserting a resistor from the
    sensor output signal back to ground (single ended inputs on the adc).

    Why does it start at 4 ma or 1 volt? Sensor error / Wiring detection.

    Here is an ASCII schematic.... Hope it comes out well.


    Sensor Power source ----> Sensor --- Sensor Output
    > 250
    ohm|
    > ADC + Input

    |

    |
    Ground

    >ADC - Input


    Hard to tell, but the resistor is mounted "between" sensor output and
    ground.
    Remember to subtract 1 volt when you do your scaling for zero since you
    start at 4ma / 1 volt for a zero reading or whatever the lower scale point
    is of the sensor.

    One other note, 250 ohm resisters can be hard to find or expensive. Use 249
    ohms and compensate in software or use a couple of 120 ohms with a 10ohm in
    series, or a couple of 500 ohms in parallel. it's your preference.

    "If" anyone has a good source for reasonably priced 250 ohm resistors let me
    know please. $2.00 a peice is not reasonable for me.

    Richard







    Original Message
    From: "SB" <steve.brady@r...>
    To: <basicstamps@yahoogroups.com>
    Sent: Thursday, July 08, 2004 5:50 AM
    Subject: Re: [noparse][[/noparse]basicstamps] Pressure sensor 4-20mA



    >> I'm unsure of hte coding sequence of these guys unless they're an analog
    >> measurement.
    >>
    >> But, since the stamp likes things in 5V, you might want to just throw a
    >> resistor across that guy there. I know I've heard someone say as much
    >> before.
    >> You might actually want to do a search through the archives.
    >>
    >> Just take 5volts and divide by 20mA and you should get a resistor value.
    >> That's the best I can remember!
    >>
    >>
    >>
    Original Message
    >> From: "ikenniemandanders" <tld-nv@s...>
    >> To: <basicstamps@yahoogroups.com>
    >> Sent: Thursday, July 08, 2004 1:50 AM
    >> Subject: [noparse][[/noparse]basicstamps] Pressure sensor 4-20mA
    >>
    >>
    >
    >>> > Hi guys & girls,
    >>> >
    >>> > I am working on a HVAC project. I have to read in an 4-20mA pressure
    >>> > sensor (between -0,5 and 7 BAR brand DIXELL). I've searched a lot on
    >>> > internet and didn't find a proper solution to connect this sensor to
    >>> > my BS2P-24. I've already made an easy circuit wich contains an 12BIT
    >>> > ADC and a resistor (150ohms) so I can get max 3V over my adc inputs.
    >>> > When I read the ADC in, I get VERY different values. My power supply
    >>> > is extremly stable (checked with an oscilloscope). Is there
    >>> > somebody who has created a simular project and could give me a
    >>> > better way to connect the pressure sensor with my stamp?
    >>> >
    >>> > Kindly regards
    >>> > Tim
    >>> > Belgium
    >>> > Student ICT
    >>> >
    >>> >
    >>> >
    >>> >
    >>> > To UNSUBSCRIBE, just send mail to:
    >>> > basicstamps-unsubscribe@yahoogroups.com
    >>> > from the same email address that you subscribed. Text in the Subject

    and

    >> Body of the message will be ignored.
    >
    >>> >
    >>> > Yahoo! Groups Links
    >>> >
    >>> >
    >>> >
    >>> >
    >>> >
    >
    >>
    >>
    >>
    >> To UNSUBSCRIBE, just send mail to:
    >> basicstamps-unsubscribe@yahoogroups.com
    >> from the same email address that you subscribed. Text in the
    Subject and

    Body of the message will be ignored.

    >>
    >> Yahoo! Groups Links
    >>
    >>
    >>
    >>
    >>
    >>
    >> [noparse][[/noparse]Scanned by AwesomeNet Anti-Virus]
    >>
    >>
  • ArchiverArchiver Posts: 46,084
    edited 2004-07-09 14:33
    --- In basicstamps@yahoogroups.com, "ikenniemandanders" <tld-nv@s...>
    wrote:
    > Hi guys & girls,
    >
    > I am working on a HVAC project. I have to read in an 4-20mA
    pressure
    > sensor (between -0,5 and 7 BAR brand DIXELL). I've searched a lot
    on
    > internet and didn't find a proper solution to connect this sensor
    to
    > my BS2P-24. I've already made an easy circuit wich contains an
    12BIT
    > ADC and a resistor (150ohms) so I can get max 3V over my adc
    inputs.
    > When I read the ADC in, I get VERY different values. My power
    supply
    > is extremly stable (checked with an oscilloscope). Is there
    > somebody who has created a simular project and could give me a
    > better way to connect the pressure sensor with my stamp?
    >
    > Kindly regards
    > Tim
    > Belgium
    > Student ICT


    Hi Tim,

    with a high pressure, (75 to 100 psi) I assume you might be using
    this on a refrigeration compressor ?

    In that event, the compressor will generate pulsations that would
    make the needle shake on a meter. I have a collection of meters
    needles that have fallen off.

    A Bourdon Tube meter will have a small capacity and mechanical
    resistances so will not reveal the high speed noise that electronics
    will. Think mA meter vs. O-Scope.

    Determine the time domain you need for a reading and add some filter
    that will keep your reading accurate in that time.

    Considder a highest reading or lowest reading, or averaging filter.
    And, considder all three. A mechanical pulse eleminator as well as a
    hardware filter and software filter.

    As far as resistors go, don't worry about expensive precision
    resistors as you have already lost the lower 20% of the scale by
    converting from 4-20mA. (and as SPENCE poiunted out, gained a broken
    wire failure mode)

    The resistors will just be for scaling to your 3 volts. Calibraton
    can be in software and not hardware. Also, if you add a pot for
    adjustment, you just wiped out the benifit of a precision resistor.

    Other problems for very noisy ADC input signals are things like a
    fluxuatoing power supply on the ADC. Use a Scope and see how steady
    the power supply is.

    Add a small capacitor on the ADC inputs to ground. This will help to
    remove some of the noise. add an R/C filter of larger values to
    reduce noise.

    Possibly add filter caps on the powr to the devices that are
    switching.

    And, test a seperate value. create a 12mA signal at the pressure
    device and connect that to the wires instead of the pressure device.
    if a steady state signal is still noisy, then look for the problem in
    the board.

    Dave
  • ArchiverArchiver Posts: 46,084
    edited 2004-07-10 14:44
    --- In basicstamps@yahoogroups.com, "ikenniemandanders" <tld-nv@s...>
    wrote:
    > Dave, how do i determine my capacitor values? Could you be so kind
    > to draw the schematic with ADC convertor? I have no idea how to
    > connect the capacitor with m
  • ArchiverArchiver Posts: 46,084
    edited 2004-07-10 21:15
    I'm just testing my program with a manometer, the pressure sensor
    and a bottle of R22 or 404 (it doesn't mather) so a very variable
    pressure is not possible. Perhaps i'm not using a very good ADC
    (texas instr). I might be able to compensate the fluctuating values
    by using a FIFO register with for example ten values, then divide
    the sum of the values trough ten so I get an probiately stable value.
    I've already changed my DIXELL sensor by an DANFOSS sensor because
    the sensor i'm using goes only to -0.5BAR and I must be able to
    measure total vacuum for my project.

    Greets,


    Tim


    --- In basicstamps@yahoogroups.com, "Dave Mucha" <davemucha@j...>
    wrote:
    > --- In basicstamps@yahoogroups.com, "ikenniemandanders" <tld-
    nv@s...>
    > wrote:
    > > Dave, how do i determine my capacitor values? Could you be so
    kind
    > > to draw the schematic with ADC convertor? I have no idea how to
    > > connect the capacitor with m
  • ArchiverArchiver Posts: 46,084
    edited 2004-07-10 21:40
    I would be strange if i don't manage to process the values to a
    stable value because the DIXELL and ELIWELL regulators (on doorpanel
    mounting modules wich regulates the evaporator fan and reads in the
    temperature sensors.



    --- In basicstamps@yahoogroups.com, "ikenniemandanders" <tld-
    nv@s...> wrote:
    > I'm just testing my program with a manometer, the pressure sensor
    > and a bottle of R22 or 404 (it doesn't mather) so a very variable
    > pressure is not possible. Perhaps i'm not using a very good ADC
    > (texas instr). I might be able to compensate the fluctuating
    values
    > by using a FIFO register with for example ten values, then divide
    > the sum of the values trough ten so I get an probiately stable
    value.
    > I've already changed my DIXELL sensor by an DANFOSS sensor because
    > the sensor i'm using goes only to -0.5BAR and I must be able to
    > measure total vacuum for my project.
    >
    > Greets,
    >
    >
    > Tim
    >
    >
    > --- In basicstamps@yahoogroups.com, "Dave Mucha" <davemucha@j...>
    > wrote:
    > > --- In basicstamps@yahoogroups.com, "ikenniemandanders" <tld-
    > nv@s...>
    > > wrote:
    > > > Dave, how do i determine my capacitor values? Could you be so
    > kind
    > > > to draw the schematic with ADC convertor? I have no idea how
    to
    > > > connect the capacitor with m
  • ArchiverArchiver Posts: 46,084
    edited 2004-07-10 22:32
    Hi Tim,

    I often connect 4-20ma instruments to an ADC and Stamp, tensiometers
    that monitor soil moisture suction in -90 to +20 kPa range. Here is
    a URL:

    http://www.emesystems.com/OWL2face.htm#4-20%20ma

    Be sure you have the grounds connected properly. Take Dave's
    suggestion and look at what you read on a multimeter connected in
    parallel with your 150 ohm termination resistor. If that is steady,
    with little AC component, then the Stamp reading should be steady too.

    -- Tracy


    >I am working on a HVAC project. I have to read in an 4-20mA pressure
    >sensor (between -0,5 and 7 BAR brand DIXELL). I've searched a lot on
    >internet and didn't find a proper solution to connect this sensor to
    >my BS2P-24. I've already made an easy circuit wich contains an 12BIT
    >ADC and a resistor (150ohms) so I can get max 3V over my adc inputs.
    >When I read the ADC in, I get VERY different values. My power supply
    >is extremly stabled (checked with an oscilloscope). Is there
    >somebody who has created a simular project and could give me a
    >better way to connect the pressure sensor with my stamp?


    >I'm just testing my program with a manometer, the pressure sensor
    >and a bottle of R22 or 404 (it doesn't mather) so a very variable
    >pressure is not possible. Perhaps i'm not using a very good ADC
    >(texas instr). I might be able to compensate the fluctuating values
    >by using a FIFO register with for example ten values, then divide
    >the sum of the values trough ten so I get an probiately stable value.
    >I've already changed my DIXELL sensor by an DANFOSS sensor because
    >the sensor i'm using goes only to -0.5BAR and I must be able to
    >measure total vacuum for my project.
    >
    >Greets,
    >
    >
    >Tim
    >
    >
    >--- In basicstamps@yahoogroups.com, "Dave Mucha" <davemucha@j...>
    >wrote:snip
    >
    > > From your first post, your readings were extreem. If you had a
    >> little bouncing it would be easier to pin down a cause.
    >>
    >> What type of thing are you reading ? are you reading water
    >> pressure ? air conditioning pressure ?
    >> The process you are reading may offer more clues. Water pressure
    >> would remain faily constant while Freon pressure in an air
    >> conditioner system would be very noisy.
    >>
    > > Dave
    >
  • ArchiverArchiver Posts: 46,084
    edited 2004-07-11 03:05
    --- In basicstamps@yahoogroups.com, "ikenniemandanders" <tld-nv@s...>
    wrote:
    > I'm just testing my program with a manometer, the pressure sensor
    > and a bottle of R22 or 404 (it doesn't mather) so a very variable
    > pressure is not possible. Perhaps i'm not using a very good ADC
    > (texas instr). I might be able to compensate the fluctuating values
    > by using a FIFO register with for example ten values, then divide
    > the sum of the values trough ten so I get an probiately stable
    value.
    > I've already changed my DIXELL sensor by an DANFOSS sensor because
    > the sensor i'm using goes only to -0.5BAR and I must be able to
    > measure total vacuum for my project.
    >
    > Greets,
    >
    >
    > Tim

    Hi Tim,

    If you expect the pressure to be constant and it varies wildly, it
    may be a sensor location. it may be a poor connection or many other
    things. It may even be a bad chip.

    put a small, 0.1 or 0.01uF cap to ground from your input line. See if
    that helps.

    try a 10k resistor on the input line and a 0.01uF after that to
    ground and see if what that does.

    Ther are a few software filters, but they are not always the best way
    to handle noise. but are usually very easy.

    The one I use is to read the input
    add the input to a variable.
    divide the variable by some value, say 10. This is the average.
    then subtract that result from the variable.
    it will keep the average rolling and allow the value to respond to
    your inputs.

    a second duplicate filter could take the result of the divide by ten
    as the input. add that to a second variable, divide that by 10 and
    then subtract that average from the variable. that winds up being
    second filter that will smooth the readings even more.

    If you find that adding the values to a total of 10 will exceed the
    memory for a variable, you can subtract a constant from each input,
    then add it back to the result to get the number.

    Dave
  • ArchiverArchiver Posts: 46,084
    edited 2004-07-12 15:34
    Also, something to mention. We use pressure transmitters / transducers on
    high pressure equipment all the time, however we always install a small
    orifice (Pressure Dampner) before the sensor. There are many sensor
    manufacturers that already supply these in the sensor if you specify it.
    They can be bought for under a buck.

    It really smooths out the pulsation spikes. Example, a vacuum gauge on an
    engine has a pulse for each time a cylinder makes it thru it's cycle.
    Without the dampner installed, the needle or readings would be pretty crazy.
    With the dampner, they are nice and steady. In this application, most
    gauges you buy off the shelf already have it installed.


    Also, I just had the same problem as your mentioning a few weeks ago and it
    happened to be my grounding. I missed a wire and had an incomplete STAR
    ground. My readings were all over the place. Easy to do when playing with
    multiple supplies.

    We use Danfoss sensors all the time so I know it probably isn't a sensor
    problem.

    Good Luck,
    Richard




    Original Message
    From: "Dave Mucha" <davemucha@j...>
    To: <basicstamps@yahoogroups.com>
    Sent: Friday, July 09, 2004 8:33 AM
    Subject: [noparse][[/noparse]basicstamps] Re: Pressure sensor 4-20mA


    > --- In basicstamps@yahoogroups.com, "ikenniemandanders" <tld-nv@s...>
    > wrote:
    > > Hi guys & girls,
    > >
    > > I am working on a HVAC project. I have to read in an 4-20mA
    > pressure
    > > sensor (between -0,5 and 7 BAR brand DIXELL). I've searched a lot
    > on
    > > internet and didn't find a proper solution to connect this sensor
    > to
    > > my BS2P-24. I've already made an easy circuit wich contains an
    > 12BIT
    > > ADC and a resistor (150ohms) so I can get max 3V over my adc
    > inputs.
    > > When I read the ADC in, I get VERY different values. My power
    > supply
    > > is extremly stable (checked with an oscilloscope). Is there
    > > somebody who has created a simular project and could give me a
    > > better way to connect the pressure sensor with my stamp?
    > >
    > > Kindly regards
    > > Tim
    > > Belgium
    > > Student ICT
    >
    >
    > Hi Tim,
    >
    > with a high pressure, (75 to 100 psi) I assume you might be using
    > this on a refrigeration compressor ?
    >
    > In that event, the compressor will generate pulsations that would
    > make the needle shake on a meter. I have a collection of meters
    > needles that have fallen off.
    >
    > A Bourdon Tube meter will have a small capacity and mechanical
    > resistances so will not reveal the high speed noise that electronics
    > will. Think mA meter vs. O-Scope.
    >
    > Determine the time domain you need for a reading and add some filter
    > that will keep your reading accurate in that time.
    >
    > Considder a highest reading or lowest reading, or averaging filter.
    > And, considder all three. A mechanical pulse eleminator as well as a
    > hardware filter and software filter.
    >
    > As far as resistors go, don't worry about expensive precision
    > resistors as you have already lost the lower 20% of the scale by
    > converting from 4-20mA. (and as SPENCE poiunted out, gained a broken
    > wire failure mode)
    >
    > The resistors will just be for scaling to your 3 volts. Calibraton
    > can be in software and not hardware. Also, if you add a pot for
    > adjustment, you just wiped out the benifit of a precision resistor.
    >
    > Other problems for very noisy ADC input signals are things like a
    > fluxuatoing power supply on the ADC. Use a Scope and see how steady
    > the power supply is.
    >
    > Add a small capacitor on the ADC inputs to ground. This will help to
    > remove some of the noise. add an R/C filter of larger values to
    > reduce noise.
    >
    > Possibly add filter caps on the powr to the devices that are
    > switching.
    >
    > And, test a seperate value. create a 12mA signal at the pressure
    > device and connect that to the wires instead of the pressure device.
    > if a steady state signal is still noisy, then look for the problem in
    > the board.
    >
    > Dave
    >
    >
    >
    >
    > To UNSUBSCRIBE, just send mail to:
    > basicstamps-unsubscribe@yahoogroups.com
    > from the same email address that you subscribed. Text in the Subject and
    Body of the message will be ignored.
    >
    > Yahoo! Groups Links
    >
    >
    >
    >
    >
    >
    > [noparse][[/noparse]Scanned by AwesomeNet Anti-Virus]
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2004-07-12 16:49
    --- In basicstamps@yahoogroups.com, "Richard Skinner"
    <rwskinner@a...> wrote:
    > Also, something to mention. We use pressure transmitters /
    transducers on
    > high pressure equipment all the time, however we always install a
    small
    > orifice (Pressure Dampner) before the sensor. There are many sensor
    > manufacturers that already supply these in the sensor if you
    specify it.
    > They can be bought for under a buck.


    > > And, considder all three.
    > > A mechanical pulse eleminator as well as a
    > > hardware filter and software filter.


    This is what I meant by a mechanical filter. A not here is that the
    snubber you mention can be had in a few varieties. one is actually
    an isolation diaphram (diaphram seal), another a small hole or
    orifice.

    The farther away from the transmitter, the greater the capacity in
    the tubing between the snubber and the transmitter. This should also
    decrease the noise. And, you can use multiple units to get even more
    filtering. (anyone here thinking, tiny orfice = resistor and large
    capacity = capacitor ?)

    Dave
  • ArchiverArchiver Posts: 46,084
    edited 2004-07-12 19:11
    Thanks to anyone who has provided me usefull information (allen,
    dave and richard). I installed an orifice (i'm in the refrigeration
    branche so i have some orifices in my possesion (I used the smallest
    hole). My pressure value comes in real stable so it works. I tested
    it with an 2x16 lcd display and i'm now able to see de pressure in
    bars en millibars on the screen. One important part of my project is
    finished. You're wright about the vacuumpump noise because the value
    became very unstable buth with an orifice it stabled a lot.


    Kindly regards,


    Tim
    student ICT @ Belgium





    --- In basicstamps@yahoogroups.com, "Dave Mucha" <davemucha@j...>
    wrote:
    > --- In basicstamps@yahoogroups.com, "Richard Skinner"
    > <rwskinner@a...> wrote:
    > > Also, something to mention. We use pressure transmitters /
    > transducers on
    > > high pressure equipment all the time, however we always install
    a
    > small
    > > orifice (Pressure Dampner) before the sensor. There are many
    sensor
    > > manufacturers that already supply these in the sensor if you
    > specify it.
    > > They can be bought for under a buck.
    >
    >
    > > > And, considder all three.
    > > > A mechanical pulse eleminator as well as a
    > > > hardware filter and software filter.
    >
    >
    > This is what I meant by a mechanical filter. A not here is that
    the
    > snubber you mention can be had in a few varieties. one is
    actually
    > an isolation diaphram (diaphram seal), another a small hole or
    > orifice.
    >
    > The farther away from the transmitter, the greater the capacity in
    > the tubing between the snubber and the transmitter. This should
    also
    > decrease the noise. And, you can use multiple units to get even
    more
    > filtering. (anyone here thinking, tiny orfice = resistor and
    large
    > capacity = capacitor ?)
    >
    > Dave
  • ArchiverArchiver Posts: 46,084
    edited 2004-07-13 03:31
    good, the hole we use in our orifice is about 0.015" which is pretty small.
    I guess AC Orifices come that small but smallest ac orifice I had laying
    around was a 0.025.

    Glad you got it going.

    Richard
    Original Message
    From: "ikenniemandanders" <tld-nv@s...>
    To: <basicstamps@yahoogroups.com>
    Sent: Monday, July 12, 2004 1:11 PM
    Subject: [noparse][[/noparse]basicstamps] Re: Pressure sensor 4-20mA


    > Thanks to anyone who has provided me usefull information (allen,
    > dave and richard). I installed an orifice (i'm in the refrigeration
    > branche so i have some orifices in my possesion (I used the smallest
    > hole). My pressure value comes in real stable so it works. I tested
    > it with an 2x16 lcd display and i'm now able to see de pressure in
    > bars en millibars on the screen. One important part of my project is
    > finished. You're wright about the vacuumpump noise because the value
    > became very unstable buth with an orifice it stabled a lot.
    >
    >
    > Kindly regards,
    >
    >
    > Tim
    > student ICT @ Belgium
    >
    >
    >
    >
    >
    > --- In basicstamps@yahoogroups.com, "Dave Mucha" <davemucha@j...>
    > wrote:
    > > --- In basicstamps@yahoogroups.com, "Richard Skinner"
    > > <rwskinner@a...> wrote:
    > > > Also, something to mention. We use pressure transmitters /
    > > transducers on
    > > > high pressure equipment all the time, however we always install
    > a
    > > small
    > > > orifice (Pressure Dampner) before the sensor. There are many
    > sensor
    > > > manufacturers that already supply these in the sensor if you
    > > specify it.
    > > > They can be bought for under a buck.
    > >
    > >
    > > > > And, considder all three.
    > > > > A mechanical pulse eleminator as well as a
    > > > > hardware filter and software filter.
    > >
    > >
    > > This is what I meant by a mechanical filter. A not here is that
    > the
    > > snubber you mention can be had in a few varieties. one is
    > actually
    > > an isolation diaphram (diaphram seal), another a small hole or
    > > orifice.
    > >
    > > The farther away from the transmitter, the greater the capacity in
    > > the tubing between the snubber and the transmitter. This should
    > also
    > > decrease the noise. And, you can use multiple units to get even
    > more
    > > filtering. (anyone here thinking, tiny orfice = resistor and
    > large
    > > capacity = capacitor ?)
    > >
    > > Dave
    >
    >
    >
    > To UNSUBSCRIBE, just send mail to:
    > basicstamps-unsubscribe@yahoogroups.com
    > from the same email address that you subscribed. Text in the Subject and
    Body of the message will be ignored.
    >
    > Yahoo! Groups Links
    >
    >
    >
    >
    >
    >
    > [noparse][[/noparse]Scanned by AwesomeNet Anti-Virus]
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2004-07-20 21:15
    >But, you do need to figure out why the value swings.

    I got in on this thread kind of late, so please excuse me if this has
    already been covered. If you are measuring pressure from a pump or
    something which is a pulsating device, I would say you should buffer the
    pulses by running the pressure output into an integrating vessel of some
    sort. For high pressure (few hundred PSI) you could use a length of pipe.
    This would smooth out the pulsations. If this isn't enough, go to a larger
    buffer volume, and/or add a simple RC electronic filter to the input of
    your ADC.
    steve
  • ArchiverArchiver Posts: 46,084
    edited 2004-07-20 23:32
    --- In basicstamps@yahoogroups.com, Steve Roberts <sroberts@s...>
    wrote:
    >
    > >But, you do need to figure out why the value swings.
    >
    > I got in on this thread kind of late, so please excuse me if this
    has
    > already been covered. If you are measuring pressure from a pump or
    > something which is a pulsating device, I would say you should
    buffer the
    > pulses by running the pressure output into an integrating vessel of
    some
    > sort. For high pressure (few hundred PSI) you could use a length of
    pipe.
    > This would smooth out the pulsations. If this isn't enough, go to a
    larger
    > buffer volume, and/or add a simple RC electronic filter to the
    input of
    > your ADC.
    > steve

    Since pressure travels at the speed of sound, there are certain
    mechanical problems to be aware of.

    One is the capacitance of the sensor. if there is little or no
    capacitance, it will respond with little delay to the to signal.
    Also if there is no mechanical resistance aka hysteresis or sticktion
    then the reading will closely follow the actual pressure.

    In refrigeration, they have capillary tubing and a foot or so of that
    will dampen quite a lot of the pulsing. having a capacity tank after
    that tubing and before the sensor will do even more.

    But, having a stabile input is MUCH better than trying to filter the
    noise once it is in the electronics.

    Dave
Sign In or Register to comment.