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.

TPS92515: Repeatable Error After Power Cycle

Part Number: TPS92515

Hey,

Our TPS92515 LED driver circuit is exhibiting a power output drop after some power cycles. The power drop is a pretty stable 20mA below what we desire and have observed the circuit provide given our controls. Inputs to the chip such as DAC/IADJ, Vin, PWM, and VCC have been observed to be the same during proper and faulty operational power cycles. After powering down and back up our circuit for 30 times we will see this error show up once or twice and then return to normal operation on the subsequent power cycles. For some of our circuits on the same board we never see this issue arise.  

In the below chart you can see our normal linear operation above the faulty operation. We've noticed that the error is not percent based on our expected output but ~20mA decrease for every point.

In this next chart I've graphed the current outputs over the course of the test. We cycle through 11 different DAC settings for every power cycle. We power cycle in this test 30 times. Note that the highlighted section of dots is the only one to exhibit the power drop error and during the next power cycle it has returned to proper output.

At this point we believe there may be a transient issue upon startup that sets an internal voltage reference incorrectly. However we have been unable confirm this suspicion so far. The Coff pin charges up and down constantly during operation so any issue with it should only be visible for an extremely short time. We also believe that the BOOT and SW pins would exhibit the same instantaneous error if any at all. We would also like to note that our Vin ripple is well below both 2V and 10% of total voltage.

Is there some internal voltage reference we have missed or do not fully understand? Is this a known issue or a sign of a defective part? Is it possible some aspect of the circuit design is failing(it exhibits no other issues)?

Please and Thanks

Tyler

  • Hello Tyler,

    During the test for the graph above, does this behavior change during the test? Are the results different for parts that exhibit this? When I look at the data points I see 12 per column and you have 11 DAC settings so is this during one power cycle and there are 12 data points or is this part of two power cycles? The regulation point changes only after power cycling and changes back after power cycling? I'm just confirming what you stated above.

    Also, what is your current sense resistor value?

    Regards,
  • Hey Irwin,

    The behavior only changes from power cycle to power cycle, never inside of the same power cycle. So if our first point comes up lower than expected we have observed the rest of the points for that power cycle to also come up low 100% of the time. Every power cycle is either a complete success or complete failure at every point within it.
    Excuse my typo, there is actually 12 control points we test during every power cycle. Sorry for the confusion.

    Our current sense resistor is 0.2ohms.
  • Hello Tyler,

    So what is going on is the self calibration of the TPS92515. There are 5 mV adjustments to make the current regulation set-point more accurate. It allows an internal change of 5 mV per step. at full Viadj this would be 5 mV/240 mV or about 0.021%. At one amp average this is about what you are measuring.

    There is another e2e posting regarding this: e2e.ti.com/.../691778

    If that doesn't work you can search TPS92515 current regulation and it will show up.

    Regards,
  • Wow. Where in the datasheet can I find this?
  • Is there any more detailed documentation of the current regulation/self calibration function you described? I've viewed the other thread that you linked and the example within it, but I'd appreciate as much information on this function of the chip as possible.
    Would increasing the capacitor on Vin possibly help stabilize this inaccuracy?
  • Hello Tyler,

    I sent you an email regarding this.

    Regards,