TMP102

02 Jun 2011

Hi !

I design my own PCB to implement the SOT TMP102.

I connected my sda, scl, Vcc et gnd pins on my Mbed, using I2C connector.

I test the HelloWorld program of Donatien but when I implement a Serial pc communication, I don't receive what the sensor need to send.

Anyone have an idea about how to solve the issue ?

02 Jun 2011

An illustration of my pcb

my tmp

02 Jun 2011

Did you mind the Pullup-Resistors on sda and scl? does not look like.

02 Jun 2011

hi Johannes !

nop ! I haven't already add the pullup resistors. Are they really important to run the tmp102 ?

02 Jun 2011

not the tmp102, but the I2C-Interface really needs them to work, because all the slaves of an i2c-master are open-collector-outputs . look here: http://mbed.org/handbook/I2C

02 Jun 2011

Well, I read the datasheet. The pullup resistor is 1K on scl and sda.

I post my board in photo. My schem is it ok ? My board

I try again to read the temperature, but it is the same problem...

02 Jun 2011

no ! It is wrong ! I need to connect my pullup resitor on +5V....

02 Jun 2011

I try this : board schema

02 Jun 2011

It's work but not really fine. my serial connextion send me the good temperature but half the tima I have 0.0000. Is a frequency issue ?

02 Jun 2011

In general, the pullups I have used are around 5 to 50 K. And should pull up to 3.3 volts, unless of course your device is on th 5 volt rail.

Onto possible reson for 00000 reading, is not enough gap between samples.

If all else fails read the datasheet ;) Ceri

04 Jun 2011

hi !

I solved this issue in connecting not used pins at gnd !

20 Aug 2012

which unused pins did you connect to gnd ? i made my wiring after your photo and i still get 0.000000 in teraterm.

20 Aug 2012

Nice a reply after one year .... lol I remerber it was one of my second year project. Each ten second, data transmitted was false. A mean to solve this issue was to connect unused pin to GND. According to http://www.ti.com/lit/ds/symlink/tmp102.pdf , just search the value you need to put on ADD0 pin. No need to connect on GND other pin connect SDA, SCL, VDD, GND. No GND for Alert.

BR

20 Aug 2012

/media/uploads/razvancob/_scaled_jpg.jpg

what am i missing ? the code is the one from the cookbook only that p9 is p28 and p10 is p27 in my case.

20 Aug 2012

ok, please post your code. I think your connection pins is okay. Have you tried for the P9 and P10. You have may be a hardware issue with your mbed board.

20 Aug 2012
  1. include "mbed.h"
  2. include "TMP102.h"

TMP102 temperature(p28, p27, 0x90); A0 pin is connected to ground

Serial pc(USBTX, USBRX); tx, rx to PC

int main()

{

while(1) {

pc.printf("Temperature: %f\n", temperature.read());

wait(1.0);

}

}

it's the same thing for pins 9&10 i read about them that they are connected to the SD slot so i moved to 28 and 27. could there be a problem with my breadboard or wires ? i tried to force the resistors as hard as i could into the slots without bending them too much, i think the wires go kind of hard inside, maybe because the breadboard is new. I tried to import other ppl's codes on and the result is the same 0.000000000000 i googled it but didn't find too much info. Thanks for trying to help.

20 Aug 2012

According to Sparkfun board schema, he pull-up resistor have beel already connected on the tmp102. Try to take off your Resistor in SDA and SDL pins.

Source : http://www.sparkfun.com/datasheets/Sensors/Temperature/TMP102_Breakout-v11.pdf

20 Aug 2012

i started without placing them at first, the result was the same, then i read the second comment from this page:

Johannes Sedlmeier

  1. 02 Jun 2011 Did you mind the Pullup-Resistors on sda and scl? does not look like.

so i placed these ones (2.2kohm) and still no improvements, do i need to use the 5V output ?

20 Aug 2012

Gerald is right, you dont need to use external pull up Rs with this breakout board. Have you checked your breadboard for connectivity between the different 5 hole islands on the long sides. In most cases they are not not connected and you need short wires linking one to the next. In your given wiring picture that would mean the tmp sensor doesn't get 3V3 and Gnd. In addition the pull-up Rs are floating. Measure for connectivity without powering up or measure for 3V3 on the breakout module.

21 Aug 2012

Thank you for the advice, i will check everything out as soon as i arrive home on the 29th as now i am in vacation at another location and i don't have a voltmeter here. Thank you all for your help, i'll return with a post after everything is checked. Have a nice day !

edit: i made another experiment with a potentiometer and a led and the "islands" are connected from one way to another. i think the fact that i'm placing the wires straight through the sensor without soldering them with a pin or smth is providing an unreliable connection