Shop OBEX P1 Docs P2 Docs Learn Events
LTC1298 reading LM35 — Parallax Forums

LTC1298 reading LM35

ArchiverArchiver Posts: 46,084
edited 2003-05-21 15:17 in General Discussion
Hi,

I'm making progress now with my datalogger (*THNKS ALL, see BS2-USB
message).

But now i'm walking to someting strange!!

If i measure voltage on my Vout of the LM35CZ (Yes, i'm from Holland
so it is in Celcius)i take a reading from 0,2445V let's say 245mV.
If i take a ADC reading with y LTC1298 (single ended channel 1) a get
a readout from 190!! I now i miss someting but, can anyone explane it
to me..

De voltage to the LTC1298 and the LM35 are 5,002V reference voltage.

Thnks in advance

Willem

Comments

  • ArchiverArchiver Posts: 46,084
    edited 2003-05-16 15:32
    Willem,

    That sounds about right. The ADC will not output millivolts in. The number
    it generates is in relation to the reference voltage you give it. It can be
    a pain to calibrate if the LM** isn't in a waterproof package, as raising
    air temp with any accuracy is hard. So, use a multiple turn pot instead. As
    the LM** outputs 10mV per degree, use the pot to feed a voltage of .3445V
    and read the ADC. Take the result and subtract the 190 from it. You now know
    what one degree equals on the ADC and can have the Stamp do the math and
    turn it back into degrees.

    A 5V reference won't make for a very accurate measurement. If you are
    measuring normal air temps, I would recommend getting a 1.5V precision
    regulator or a zener diode, depending on the resolution you want. This would
    make your max temp that you can read 150 degrees.

    One last thing, the LM35z is the least accurate. If you want to make sure of
    precise output, check the LM35z versus a good thermometer.

    Hope this helps,

    Jonathan
    www.madlabs.info

    Original Message
    From: "wrep" <basic@P...>
    To: <basicstamps@yahoogroups.com>
    Sent: Friday, May 16, 2003 6:29 AM
    Subject: [noparse][[/noparse]basicstamps] LTC1298 reading LM35


    > Hi,
    >
    > I'm making progress now with my datalogger (*THNKS ALL, see BS2-USB
    > message).
    >
    > But now i'm walking to someting strange!!
    >
    > If i measure voltage on my Vout of the LM35CZ (Yes, i'm from Holland
    > so it is in Celcius)i take a reading from 0,2445V let's say 245mV.
    > If i take a ADC reading with y LTC1298 (single ended channel 1) a get
    > a readout from 190!! I now i miss someting but, can anyone explane it
    > to me..
    >
    > De voltage to the LTC1298 and the LM35 are 5,002V reference voltage.
    >
    > Thnks in advance
    >
    > Willem
    >
    >
    >
    >
    >
    >
    > 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.
    >
    >
    > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
    >
    >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2003-05-16 15:40
    >I'm making progress now with my datalogger (*THNKS ALL, see BS2-USB
    >message).
    >But now i'm walking to someting strange!!
    >If i measure voltage on my Vout of the LM35CZ (Yes, i'm from Holland
    >so it is in Celcius)i take a reading from 0,2445V let's say 245mV.
    >If i take a ADC reading with y LTC1298 (single ended channel 1) a get
    >a readout from 190!! I now i miss someting but, can anyone explane it
    >to me..
    >De voltage to the LTC1298 and the LM35 are 5,002V reference voltage.
    >Thnks in advance
    >Willem

    Hi Willem.

    The LTC1298 is almost right. The reading should be:

    245/5002 * 4096 = 200

    The output count would be directly in millivolts only if you use a
    4.096 volt reference, because it is a 12 bit converter with 4096
    counts. With a 5.002 volt reference, you get
    5002/4096 = 1.2212 millivolts per bit.

    To convert the reading you are getting back into millivolts, you
    have to multiply times 1.2212, and on the Stamp you do that like this:

    millivolts = reading ** 14496 + reading

    The ** 14496 gives a good approximation to .2212 (14496/65536).

    The reading of 190 when the input is 245 is still too low, though.
    Check the power supply for good bypassing and a good ground
    connection.

    -- regards,
    Tracy Allen
    electronically monitored ecosystems
    mailto:tracy@e...
    http://www.emesystems.com
  • ArchiverArchiver Posts: 46,084
    edited 2003-05-21 09:23
    Hi Jonathan

    Thanks for your reaction. But a answer give more question's. So..

    1.)Why feed the ADC with 0.3445V?
    2.)You say to use a voltage reference of 1,5V to get a max. output
    for 150 degrees (10mV/Grade) but is the output depending on the
    reference voltage? i tought the output is 10mV/degree UNdepending of
    the reference! it could be that the LM35CZ give's a 10mV/degree full
    span? i think i'm lost now help!!

    By the way I do have a stable signal (bypassing the power supply and
    i'm only 1 digit wrong, THANKS TRACY as well!!)

    willem

    --- In basicstamps@yahoogroups.com, "Jonathan Peakall"
    <jpeakall@m...> wrote:
    > Willem,
    >
    > That sounds about right. The ADC will not output millivolts in. The
    number
    > it generates is in relation to the reference voltage you give it.
    It can be
    > a pain to calibrate if the LM** isn't in a waterproof package, as
    raising
    > air temp with any accuracy is hard. So, use a multiple turn pot
    instead. As
    > the LM** outputs 10mV per degree, use the pot to feed a voltage
    of .3445V
    > and read the ADC. Take the result and subtract the 190 from it. You
    now know
    > what one degree equals on the ADC and can have the Stamp do the
    math and
    > turn it back into degrees.
    >
    > A 5V reference won't make for a very accurate measurement. If you
    are
    > measuring normal air temps, I would recommend getting a 1.5V
    precision
    > regulator or a zener diode, depending on the resolution you want.
    This would
    > make your max temp that you can read 150 degrees.
    >
    > One last thing, the LM35z is the least accurate. If you want to
    make sure of
    > precise output, check the LM35z versus a good thermometer.
    >
    > Hope this helps,
    >
    > Jonathan
    > www.madlabs.info
    >
    >
    Original Message
    > From: "wrep" <basic@P...>
    > To: <basicstamps@yahoogroups.com>
    > Sent: Friday, May 16, 2003 6:29 AM
    > Subject: [noparse][[/noparse]basicstamps] LTC1298 reading LM35
    >
    >
    > > Hi,
    > >
    > > I'm making progress now with my datalogger (*THNKS ALL, see BS2-
    USB
    > > message).
    > >
    > > But now i'm walking to someting strange!!
    > >
    > > If i measure voltage on my Vout of the LM35CZ (Yes, i'm from
    Holland
    > > so it is in Celcius)i take a reading from 0,2445V let's say 245mV.
    > > If i take a ADC reading with y LTC1298 (single ended channel 1) a
    get
    > > a readout from 190!! I now i miss someting but, can anyone
    explane it
    > > to me..
    > >
    > > De voltage to the LTC1298 and the LM35 are 5,002V reference
    voltage.
    > >
    > > Thnks in advance
    > >
    > > Willem
    > >
    > >
    > >
    > >
    > >
    > >
    > > 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.
    > >
    > >
    > > Your use of Yahoo! Groups is subject to
    http://docs.yahoo.com/info/terms/
    > >
    > >
    > >
    > >
  • ArchiverArchiver Posts: 46,084
    edited 2003-05-21 15:17
    Willem,

    Sorry if I made it all more confusing. I'll see if I can make it more clear
    or if I make it worse ;-0

    I'll answer question 2 first:

    2) I assume you are reading air temps or something else that is in the range
    of less than 150 degrees C. If this is true, using a reference voltage of
    1.5V will give you better accuracy than a reference of 5V.

    The way the ADC works is it compares the sensor input voltage against the
    reference voltage and produces an number between 0 and 4096 based on the
    difference. So, with a 5V reference, as Tracy pointed out, you get about
    1.2mV per bit. With a 1.5V reference, you get .366mV per bit, which is more
    accurate. So, with a 5V reference, each degree = 8.33 on the ADC, and with a
    1.5V reference, each degree = 27.32 on the ADC.

    The reason it limits your sensor to 150 Degrees this way is that at 150
    degrees the sensor will output 150mV, which is the same as the reference
    voltage. At this or any higher temperature, all the ADC will return the
    maximum value of the ADC, in this case with a 10 bit LTC1298, it would be
    4096. A temp of 150 degrees = 4096, and so does 151, 152, 153 etc. It isn't
    that it limits the sensor's output, it limits the ADC's output.

    So, to get the best resolution of the sensor, you want a reference that is
    just above or equal to the highest temperature you want to measure. I said
    1.5V because that is an easy value to find in a precision zener diode or
    precision voltage regulator to use as the reference voltage.

    1) This is about calibration. As you found, real life often doesn't work
    perfectly. So, although an input of 245mV *should* give you a result of 200
    from the ADC. But you got 190, an error of 10. So if you feed the ADC
    255mV, and get the result from the ADC, you can check the error to get a
    factor to use in correcting the results.

    Using a pot as I suggested, you can check the error along the whole range of
    temps you want to read, average the error, and use that to correct your
    readings.While the ADC will probably be consistent, this allows you to make
    sure. Sounds like you have this straightened out, a error of 1 on the ADC
    isn't anything to worry about.

    Hope this didn't make it worse. If it does, let's get Tracy to explain in
    ;-)

    Jonathan

    www.madlabs.info






    Original Message
    From: "wrep" <basic@P...>
    To: <basicstamps@yahoogroups.com>
    Sent: Wednesday, May 21, 2003 1:23 AM
    Subject: [noparse][[/noparse]basicstamps] Re: LTC1298 reading LM35


    > Hi Jonathan
    >
    > Thanks for your reaction. But a answer give more question's. So..
    >
    > 1.)Why feed the ADC with 0.3445V?
    > 2.)You say to use a voltage reference of 1,5V to get a max. output
    > for 150 degrees (10mV/Grade) but is the output depending on the
    > reference voltage? i tought the output is 10mV/degree UNdepending of
    > the reference! it could be that the LM35CZ give's a 10mV/degree full
    > span? i think i'm lost now help!!
    >
    > By the way I do have a stable signal (bypassing the power supply and
    > i'm only 1 digit wrong, THANKS TRACY as well!!)
    >
    > willem
    >
    > --- In basicstamps@yahoogroups.com, "Jonathan Peakall"
    > <jpeakall@m...> wrote:
    > > Willem,
    > >
    > > That sounds about right. The ADC will not output millivolts in. The
    > number
    > > it generates is in relation to the reference voltage you give it.
    > It can be
    > > a pain to calibrate if the LM** isn't in a waterproof package, as
    > raising
    > > air temp with any accuracy is hard. So, use a multiple turn pot
    > instead. As
    > > the LM** outputs 10mV per degree, use the pot to feed a voltage
    > of .3445V
    > > and read the ADC. Take the result and subtract the 190 from it. You
    > now know
    > > what one degree equals on the ADC and can have the Stamp do the
    > math and
    > > turn it back into degrees.
    > >
    > > A 5V reference won't make for a very accurate measurement. If you
    > are
    > > measuring normal air temps, I would recommend getting a 1.5V
    > precision
    > > regulator or a zener diode, depending on the resolution you want.
    > This would
    > > make your max temp that you can read 150 degrees.
    > >
    > > One last thing, the LM35z is the least accurate. If you want to
    > make sure of
    > > precise output, check the LM35z versus a good thermometer.
    > >
    > > Hope this helps,
    > >
    > > Jonathan
    > > www.madlabs.info
    > >
    > >
    Original Message
    > > From: "wrep" <basic@P...>
    > > To: <basicstamps@yahoogroups.com>
    > > Sent: Friday, May 16, 2003 6:29 AM
    > > Subject: [noparse][[/noparse]basicstamps] LTC1298 reading LM35
    > >
    > >
    > > > Hi,
    > > >
    > > > I'm making progress now with my datalogger (*THNKS ALL, see BS2-
    > USB
    > > > message).
    > > >
    > > > But now i'm walking to someting strange!!
    > > >
    > > > If i measure voltage on my Vout of the LM35CZ (Yes, i'm from
    > Holland
    > > > so it is in Celcius)i take a reading from 0,2445V let's say 245mV.
    > > > If i take a ADC reading with y LTC1298 (single ended channel 1) a
    > get
    > > > a readout from 190!! I now i miss someting but, can anyone
    > explane it
    > > > to me..
    > > >
    > > > De voltage to the LTC1298 and the LM35 are 5,002V reference
    > voltage.
    > > >
    > > > Thnks in advance
    > > >
    > > > Willem
    > > >
    > > >
    > > >
    > > >
    > > >
    > > >
    > > > 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.
    > > >
    > > >
    > > > Your use of Yahoo! Groups is subject to
    > http://docs.yahoo.com/info/terms/
    > > >
    > > >
    > > >
    > > >
    >
    >
    > 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.
    >
    >
    > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
    >
    >
    >
    >
    >
Sign In or Register to comment.