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.

DRV425: DRV425 - Quality and Drift performance (over time and temperature)

Part Number: DRV425

Hi.  We've placed these items in a temperature- and magnetic-controlled chamber for 2 weeks, cycling 5C to 38C:

1) 2 each EVM modules from TI

2) Custom boards with DRV425's (not Q1) using non-magnetic materials and conforming to the recommended datasheet circuit

3) Laboratory grade magnetometers to nullify the magnetic field

Each time the temperature returned to 25 C, the EVM modules showed no more than 50 nT offset.  However, the custom boards showed more than 500 nT offset.  We have 3 theories:

1) Our soldering technique required touch-ups with a hand iron, which may have heated the devices,

2) The devices vary in performance depending on the particular day the silicon was made,

3) The EVM circuit is different than the recommended circuit in the datasheet (shouldn't matter, but....),

Do you have any thoughts on what might be causing our boards (which use the same passive components as the EVM, but the recommended circuit in the datasheet) to perform differently than the EVM?  Also, do you believe the automotive quality part would have lower drift over time and temperature?  Thanks.

  • I would like to add 3 more notes:

    1) We have tested with 3.3V and 5 V.

    2) For all tests, and for the EVM, we are configuring R1/R2 in the same way as the EVM: REFout = VDD/2

    3) We have taken steps to verify that the measurement of the output of the EVM and our custom boards is correct equivalent.  The differential sampling of the output is equal between the EVM boards and our custom boards.

    One more question: Do you believe an external reference, or using the internal reference would impact these offset drift results?

  • Hello Charles,

    Here are a few things to check and verify my assumptions.

    • By offset drift you mean you measure the value once at 25C and the delta from that to another time at 25C is 500nT?  The offset specification is 8µT but I assume you mean the delta.
    • You are measuring Vout vs Vref
    • Your VDD is powered by the same supply in both conditions.  PSRR is large on this device for the flux-gate offset.
    • Are you using R1 (Rshunt) with the same value as the EVM.
  • Thank you for the hasty response, Javier.

    1) Yes, by offset drift we mean that initially use our reference laboratory grade magnetometer to remove the field from the chamber, then zero-out the DRV425 (EVM or custom board) by subtraction.  Over temperature cycles that last anywhere from 1 to several days, the EVM boards return to +/- 50 nT offset when the temperature returns to 25 C.  Our boards have offsets that climb (possibly logarithmically, but that's a guess) to 500 nT after 2 weeks of these cycles.

    2) Yes, we are measuring differentially the output of the devices in the same way.

    3) Sometimes yes, other times no, but it doesn't seem to make a difference.

    4) We do not have an Rshunt.  We are connecting R1 directly to VDD.  Might this be a problem?  EDIT: Yes, we are using the same RSHUNT resistor as is used in the EVM: CRCW0805100RFKEA

  • Charles,

    Can you share a schematic?  I am refereeing to the one below of the DRV425EVM User's Guide.  That may be a problem with R1.

  • This is the relevant portion, and also the entirety of the PCB which we have separated for our testing purposes.

    I must also clarify that we are using the same RSHUNT, CRCW0805100RFKEA.  When I responded to you I thought you were asking about the RSEL pullup/pulldown.

  • Charles,

    Nothing stands out.  I see the output filter is different but I don't think that is an issue.  I know the output impedance should not be a large issue as you have a 35.7kΩ and I would assume if this was an issue it would be observed on the initial temperature.  The difference amplifier is connected different but I also do not see an issue with that.

    The only other thing I could say is if there is any external current that are changing near the device that could cause a change of magnetic field near the device.

    Can you share the layout of this section?  It may be a hint but I don't know what else it could be.  You could verify the Vref voltage to make sure you are not having your power supply move like I said the Flux-gate offset does change with the supply voltage as seen in the datasheet.