Post Go back to editing

ADT7470 holds PWM output high every second

Hello,

We are using an ADT7470 to control 4x fans in one of our products. We are observing an issue with operating the fans in "low-frequency" mode (bit 6 of 0x40 = 1) whereby every second the PWM output is held HI for approximately 35ms (see attached trace). This is independent of automatic or manual control mode, or of the PWM duty cycle, and causes audible variation in the running fan speed.

More confusingly, if the tachometer is disabled using 0x74, the output is held HI for 750ms instead (attached). And if fast tach readout is enabled (bit 5 of 0x40) then this occurs every 250ms instead of every 1s.

I therefore conclude that this is related to the tachometer readout in low-frequency mode, as this is the only thing in the datasheet that occurs once per second.

Is there a workaround for this issue?

Cheers,

Martijn

attachments.zip
Parents
  • Hi, Martjin.

    Unfortunately there is no workaround to disable pulse stretching in low freqeuncy fan drive mode, it is only disabled when high frequency fan drive mode is selected.  

    Regards,

    Karen

  • Hi Karen.

    I have the same problem.
    However, the pulse stretching cycle occurs every 1 second and the fan vibrates at the same cycle.
    In my experiment, this 1 second period is linked with the Fan Speed Measurement Rate (Register 0x40 Bit[5]).
    In the evaluation board I have, pulse stretching occurs every 22.6ms, so the fan does not vibrate.

    My board is using one daisy-chained TMP05 and two 3-wire fans.
    And as a singularity, TACH3 and TACH4 are connected to GND.

    Ask for help.

    Regards,
    JSKIM.

Reply
  • Hi Karen.

    I have the same problem.
    However, the pulse stretching cycle occurs every 1 second and the fan vibrates at the same cycle.
    In my experiment, this 1 second period is linked with the Fan Speed Measurement Rate (Register 0x40 Bit[5]).
    In the evaluation board I have, pulse stretching occurs every 22.6ms, so the fan does not vibrate.

    My board is using one daisy-chained TMP05 and two 3-wire fans.
    And as a singularity, TACH3 and TACH4 are connected to GND.

    Ask for help.

    Regards,
    JSKIM.

Children
No Data