Post Go back to editing

ADT75 internal clk frequency

Hi


I would like to know internal clock frequency, and data like PSRR if it can.

I have a problem.

ADT75 is mounted on a small board and connected to main board.
Main board supply the power.
It must be wire has some registance, I think.
So power line is a little noisy.
Noise level is like 10mV , 350KHz and triangle wave.
(sorry I can't show the actual picture)


My problem is output data sometime jumped up about 1 deg C.

Bypass capacitor is installed. 0.1 and 1uF near terminals.
I think triangle wave come from ADT75 internal clock, and bad low frequency power line noise make worse the ADT75 featue.

Could I have a comment?

Regards
moto

  • Hi Moto,

    PSRR is usually 0.1°C/V @ Ta = 25°C.

    It is possible that this noise is coming from the IC but the power supply would be a more probable source of noise.

    Try increasing the capacitor values to filter out the noise and ensure that they are placed as close to the ADT75 as possible.

    The best combinations we have seen are a 0.1uF Ceramic capacitor in parallel to a 10uF Tantalum capacitor.

    Regards,

    Rob

  • Dear Mr. Rob

    Thank you for your suggestion.

    Added 10uF capacitor (ceramic) and checked.

    Then it calmed down!

    Great !

    But the power line noise still remain.

    PSRR is usually 0.1°C/V @ Ta = 25°C.

    In 1uF capacitor case , peak noise level is about 0.3V, so it should be no problem.

    In 10uF case ,noise is not so much changed.

    I think it need to reduce the low frequency noise.

    correct?

    ADT75's internal clk is 350KHz?

    temperature conversion time is 60ms.

    To reduce under 16Hz noise is important ?

    Could I have a comment?

    Regards

    moto

  • Hi Moto,

    The internal oscillator frequency is 130KHz. Are there any other devices in the system that could be creating this noise? Are you using a switch mode power supply to supply power?

    Try isolating the ADT75 and operating it from a clean power supply.

    Regards,

    Rob

  • Dear Mr. Rob

    Thank you very much for good suggestion.

    There are no device on this extra board.

    The circuit current is very low , so it may OK to use long cable, I think.

    power source is dcdc converter.

    I have to suspect power source noise.

    and also, I2C pull up 10K resistor is on this extra board.

    It's might be one of noise source.

    Could I have tips or cautions to use this device , layout and circuit?

    Your comment is very helpful for me.

    Regards

    moto

    このメッセージは次により編集されています:  Motoyasu Yoshino Mainly, I would like to know what noise frequency will have an influence on this device. To improve, It should not put 1uF , but 10uF? Regards moto

  • Hi Moto,

    The long lines and noisy power supply are more than likely the source of the noise. Try to isolate the each to locate the source.

    I would recommend you to read some application notes from HERE<http://www.analog.com/library/analogdialogue/archives/43-09/linear_circuit_design_handbook.html>. This handbook will show the best practices for PCB and sensor layout.

    Regards,

    Rob

  • Dear Mr.Rob

    Thank you very much for your support.

    A sigma delta converter is very robust against noise, I think.

    but power line noise must be different

    I'll study your introduced documents.

    Thank you for your cooperation.

    Regards

    moto