Post Go back to editing

Transient Issue with MAX44284

Category: Hardware
Product Number: MAX44284

Issue encountered with MAX44284 (SOT23 Package). The device generally works as expected, other than when I generate a transient load condition (step change current consumption). Same behaviour seen in multiple systems.

Setup details:
- device powered (VDD) from 3V3,  with 100nF and 1uF ceramic capacitors.

- Voltage powering load is 12V

- I am using a 50mR sense resistor, with 330R 'in-line' RIN resistors between RS+/RS- and the sense resistor.

- I am using 10nF filtering cap between Rs+ and Rs-. No other filtering caps.

- The output of the amplifier generally behaves as expected, with the output voltage being proportional to the current.

- I have left the output of the amplifier floating (other than oscilloscope probe) to avoid potential issues with stability / capacitive loading.

The issue appears when I simulate a load transient, effectively creating a step change in the current consumption. The device output seems to get 'stuck' at around 0.6V for 8 seconds or so, and then eventually rises to the expected value.

I have tried the following with no results:

- Increasing the filtering cap from 10nF to 100nF with no change.

- Increasing the 'in-line' RIN resistors from 330R to 1k, with no change.

- Adding more decoupling caps on VDD, no change.

- Adding filtering caps between RIN and GND on both sides RS+ and RS-, no change.

Any help would be greatly appreciated.
Hugo

  • Hi  ,

    Good day. Can you provide the schematic diagram of your setup? Also, can you please share with us the output you've encountered? Thank you.

    Regards,
    Gilbeys

  • Hi Gilbeys,

    Thank you very much for your prompt response. Please find below the relevant section in the schematic.


    VDD_USER is set to 12V. Error condition arises when VDD_USER_CONN is connected to a small load (3R or so), causing a step change in the current.
    Thanks again for your time.

    Best,
    Hugo

  • Hi Gilbeys,

    I have now taken some scope traces of the behaviour I am seeing.

    Images are pictures of oscilloscope screen, so apologies in advance for the screen reflections.




    Described behaviour where output of amplifier goes to roughly 600mV, stays there for a few seconds, then rises to the expected value.


    Zoomed in capture of the first edge from 0V to 600mV.


    Zoomed in capture of the second edge from 600mV to expected value.

    Any help would be greatly appreciated.
    Best,
    Hugo

  • Hi  ,

    Good day. Sorry for missing the thread. I'm trying to simulate your setup using LTSpice. Can you expound more on how you perform the load transient test? Thank you.


    Regards,
    Gilbeys

  • Hi Gilbeys,
    Thank you very much for following up. 
    The load transient test was performed physically on the board. With VDD_USER set to 12V, I physically connect a 3 Ohm load directly to GND. 
    This is usually enough to trigger the error condition described above.

    I have noticed that the probability of the erroneous behaviour showing up seems to depend on the VDD_USER voltage. I was not able to trigger the issue with VDD_USER at 5V, but it seems to have ~90% chance of triggering with VDD_USER at 12V. (with some sort of probability transition between 5V and 12V)
    Unfortunately, I get the impression that the issue may not appear in simulation, since it seems to be potentially more of a latch-up kind of problem.

    Looking forward to your findings!
    Best,
    Hugo

  • Hi  ,

    Good day. I'll try to perform bench testing using the available EV Kit. I'll get back to you once I have the data. Thank you.

    Regards,
    Gilbeys

  • Hi Gilbeys,

    Just wanted to check if you have managed to replicate the issue in your setup.
    Thank you once again,
    Hugo

  • Hi  ,

    Good day. I'm still waiting for the EV kit. Thank you.

    Regards,
    Gilbeys

  • Hi  ,

    I had the EV kit and will try to test your setup. Can you provide these details:

    - Is the Vdd_User_Conn an external voltage source? If yes, what was the value of the source?

    - Can you provide the part of the circuit that is connected to IN+ and IN-?

    -Also I'm curious about how you perform the measurement. If the Vdd_User_Conn is an external voltage, which source do you open first?

    Regards,
    Gilbeys

  • Hi Gilbeys,

    Great to hear you have the EV kit.

    - VDD_USER_CONN is not connected to an external voltage source. This node is where the load to GND is connected. i.e. 3R to GND as a transient.
    - The + side of IN+ (VDD_USER) is just 12V DC, which is enabled after VDD_3V3. IN- (VDD_USER_CONN) is just the node for the load to GND.

    The power-up sequence is VDD_3V3 comes up first. Seconds later VDD_USER is enabled to 12V. These are the only driven voltages in the circuit.

    Seconds later I perform the test where I connect the 3R load onto VDD_USER_CONN, providing the transient load.
    Hope this helps and really looking forward to knowing if you manage to replicate the behavior on your side.

    Best,
    Hugo