• State Resolved
  • Locked Locked
  • Replies 1 reply
  • Subscribers 46 subscribers
  • Views 611 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

LMH6518 runs warm, unable to communicate via SPI

Ali Lown
Prodigy 10 points
Other Parts Discussed in Thread: LMH6518

I have created an analog test board for our current project, which uses an LMH6518 as setup in the reference design in the datasheet (Figures 6,12,13,15).

At the moment when power is applied to the IC, the device's temperature rises rapidly to stabilise at ~68 degrees C (ambient 28 degrees C), in the default state (aux output on). Note that the device is not heatsinked, and is attached to the top of a 2 layer PCB.

This seems to be quite warm for the device when idling, is this the sort of temperature you would expect from it (assuming datasheet power consumption of 1.1W)?

When I attempt to initiate communications with the LMH6518 via the SPI interface, as soon as /CS is taken low, the device starts attempting to draw large quantities of current along the SDIO line. If I deliberately limit the current it can draw from the master on the SDIO line, the LMH6518 starts increasing in temperature gradually. When the current is not limited (tie SDIO directly  to the 3V3 rail), the device rapidly increases in temperature (takes less than 1 second to reach 80 degrees C), and continues to increase until I remove power.

I have inspected the board, and have confirmed that there are no shorts between the device pads as measured at the outside of the device, that the device is soldered in the correct orientation, and that the board schematics match the reference design schematics.

Does anyone have any ideas regarding why the LMH6518 appears to be attempting to draw large quantities of current via the SPI interface?

Thanks.

  • 0
    TI__Mastermind 22825 points

    Hi Ali,

    The LMH6518 EVAL board is a 4 layer board (inner power and ground layers) with several vias to copper areas on the other 3 layers for good thermal conductivity. Does your test board have these provisions for good power dissipation? An indication of this is to feel the heat rise on the board / copper area while the device is operating. If the board copper stays cool while the device package gets hot to the touch, there is room for improvement in the board layout.

    Here is the inner GND plane of the EVAL board for reference to show the copper area (Red area) devoted to heatsinking:

    CS\ Causing excess current draw: You have not noted the increase in 3.3V supply current in your post. However, there should not be such an increase which can register as case temperature rise! SPI communication lines are low power logic lines and these themselves cannot draw much current from the supply (but you might be changing the operating point and affecting supply current).

    Have you checked to make sure the circuit is not unstable (oscillating)? Often times, supply current increase can be attributed to device instability.

    Have you considered doing your evaluation on the LMH6518 EVAL kit available for customers? Done this way, you'd start from a known-good operating point where board layout and heat-sinking characteristics have been checked-over and given the OK.

    Here is the link if you are interested:

    http://www.ti.com/tool/lmh6518sqevk

    Please review and let me know if you have questions?

    Regards,

    Hooman