This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

AMC6821: Killing Intelligent Temperature Monitor and PWM Fan Controller

Part Number: AMC6821

I am controlling a Sanyo Ace 97BMB33 4-wire fan with in-built PWM Controller, so I'm just sending the PWM signal to the fan (the TACH signal is connected but I'm not doing anything with it).  ie. I just set a PWM value on the AMC6821 IC which then controls the fan.

See https://e2e.ti.com/support/data_converters/precision_data_converters/f/73/p/701340/2611991#2611991 for the circuit diagram I used.

I have set the Registers to the following values and in the following order. 

Config Register 1 = 0x08 (Software-DCY Mode)

Config Register 2 = 0x03 (Enable PWM)

Config Register 3 = 0x83 (leave as default)

Config Register 4 = 0x40 (Set DCY Duty Cycle)

However "it seems" if we power up the AMC6821 IC with these settings WITHOUT having the fan connected the IC gets very hot very quickly and I think we destroy it - we have now destroyed THREE ic's this way!!

I can't understand how this can happen, because the AMC6821 is just driving a logic level 25kHz PWM square wave into the fan - it is NOT controlling any power, so why does it get so hot and die?

If I have the Sanyo fan always connected when I power up the AMC6821 I don't have this issue.

Is it because I have setup the Config registers in-correctly?

Please HELP, because I can't keep on blowing these chips up.

Yours in complete frustration Chris

  • <<Update>>
    I also tried writing the Config Register 1 last in the above sequence, but that made no difference.
    I can also confirm that the above values I write to the AMC6821 registers are correctly updated in the registers, but there is no 25kHz square ware on the PWM Output.
    Chris
  • Hi Chris,

    I am currently looking into this and will get back to you.

    Best regards,
    Megan
  • Hi Chris,

    Another item that will help with the debug is a schematic of your current setup. This will allow us to verify if everything is connected well for correct operation. The issue doesn't sound like a configuration issue, but a possible connection issue and the above would help to verify.

    Edit: Please include the connection scheme for the different fans during operation.


    Best Regards,
    Matt

  • Thanks for the reply Matt,

    There is a link to my circuit diagram in my first post.

    The fan we are controlling outputs a logic level TACH signal and is controlled by a logic level output from the PWM-OUT pin of the AMC6821.

    We think perhaps when we re-connect (or is it disconnect?) the AMC6821 from the fan that perhaps the GND line is not connected and perhaps the 24V power to the fan somehow goes through this IC?

    The TACH output of the AMC6821 is tied high, but the Sanyo Fan has it's own tie up on the PWM-OUT line so we do not populate the R14 tie up resistor.

    Chris

  • Hi Chris,

    The AMC6821 requires the power supply to be off when connecting or disconnecting the fan. If you connect or disconnect the fan while in operation, you risk damaging the internal ESD protection which could cause the device to fail.

    Is the IC heating up on its own when powered on without the Fan connected?

    Another thing, I noticed you are writing 0x40 to configuration register 4; however, in the datasheet, page 38, Bit 7 requires you write a 1 to this location.

    Best,
    Megan
  • Hi Megan,

    I will have to do a test again (once are order of more AMC6821 chips comes in) to check if the IC is heating up without the fan connected - I'm assuming it shouldn't regardless of what we set the DCY (Duty Cycle) Register (0x22) to?

    Sorry I made a typo in my first post. I have set Configuration Register 4 to 0x88 (which does set Bit 7 to 1).

    Does the order of writing to these registers on power-up matter?  I saw a post which says that Config Register 1 should be written to last - is that correct?

    My current order of writing to the registers is:

    Config Register 2, Config Register 3, Config Register 4, Config Register 1 and finally DCY (Duty Cycle) Register.

    Chris

  • Chris,

    On page 10 of the datasheet under the "Power-on Reset and Start Operation" section it says, "The device does not perform any monitoring functions until the START bit of Configuration Register 1 is set ('1')." It is probably good practice to write this reg last - even though you are not setting the START bit (your config reg 1 = 0x08).

    Once you complete the test again let me know the results. In the meantime you could measure the Vdd pin on the failed devices and see if it returns an expected voltage value or if they got damaged/shorted.

    Best,
    Megan
  • Hi Megan,

    We measured the Vdd pin and it was measuring the correct voltage (3.3V).  We replaced the IC and it is now working properly.

    However our setup will almost certainly result in the Fan being disconnected from the IC while we have 24V power going to the Fan.

    So is there any sort of external protection we can put on the AMC6821 to prevent it from blowing up?

    Chris

  • Hi Chris,

    It may be difficult to add external protection to your circuit when it is still unclear which pin on the device is causing the problem. You could further investigate each node to see which pin is being shorted. My thought is either the TACH or PWM-Out pin.

    In the meantime, you could look into TVS (transient voltage suppressing) diodes, as well as resettable fuses/eFuses, and PTC resistors as some external protection options.

    Best,
    Megan