• State TI Thinks Resolved
  • Locked Locked
  • Replies 11 replies
  • Answers 1 answer
  • Subscribers 46 subscribers
  • Views 1165 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

LMP90077: LMP90077 drawing excessive current (20mA) but still works

Nick Bennett
Prodigy 120 points
Part Number: LMP90077

I have a batch of prototype board using the LMP90077, eight boards draw minimal power but two are drawing about 20mA but still operate correctly otherwise (the LMP90077 is drawing the current).

  1. The data sheet says to leave the thermal pad floating (but it's grounded), is this connected internally? could this have damaged the chip?
  2. To save current we are switching the analog supply off (digital supply remains and the chip in shutdown mode), could this have damaged the chip?

  • 0
    TI__Guru* 92715 points

    Nick,


    Thanks for the post. I'll just answer the questions as they were posed.

    1. The thermal pad should be isolated from the device. Often, the thermal pad is electrically connected to the lowest voltage on the device. However, In this device, the die is attached with a non-conductive epoxy and there should be no electrical contact. Looking through the E2E, I've seen posts saying that the device works fine with the thermal pad grounded. That being said, we still recommend that the thermal pad is left floating. This is the way that the device was characterized.

    2. I don't know if the device would be damaged with the analog supply turned off. One condition depends on the connections to the analog input pins. If the analog supply voltage is shut off, the reference voltage and the analog input voltages must be shut off as well. Note that in the absolute maximum ratings for the device (page 5 of the datasheet), these voltages must be kept between -0.3V to VA+0.3V or there may be damage to the device. I would note that there aren't any ratings that connect VIO to VA.

    I don't know of any extra current draw associated with shutting down just the VA supply. Can you put the device into standby or power-down to reduce the current instead? Are there any other cases where the supply draws a large amount of current during operation? When the device starts to draw excess current, does the device continue to draw excess current after the supply is restored?


    Joseph Wu

  • 0 in reply to Joseph Wu
    Prodigy 120 points

    Hi Joseph,

    Thanks for the very informative reply and further to that:

    The analog supply and sensor are both powered from the same switchable supply, the sensor is ratiometric with no significant capacitance to give rise to higher voltages at the analog input than the analog supply.

    Putting the device into power down mode reduces the current draw by a couple of milliamps, but still in the order of 20ish mA. I've tried putting series resistors on the D6/DRDY and SDO lines to see if there was a conflict, but found none. Changing the chip solved the problem!

    The device was drawing excessive power when the board was first powered up (on two boards), the other eight boards draw low power in any mode of operation.

    Nick Bennett

  • 0 in reply to Nick Bennett
    TI__Guru* 92715 points
    Nick,


    I'm still not sure where the 20mA is coming from. One thought that I had would be that this could be problem with the soldering process and a bad connection. To test this, you would need to remove a bad device and then solder it back on the board. Do you have a schematic or layout that you'd be able to share?

    Another thought would be eliminating the connection to the grounded thermal pad. The datasheet does recommend floating the thermal pad. While I don't know of any problems connecting it to ground (and others have done this), we might as well check to see if this could be a problem. Is there a way for you to cut out the thermal pad on the board so that it remains floating? Or do you drop vias in to the pad connection to ground and then run the connection to pin 16?

    Let me know what you think. I'm checking with the original design group with any thoughts they might have.


    Joseph Wu
  • 0 in reply to Joseph Wu
    Prodigy 120 points

    Hi Joseph,

    I'm not sure how a bad connection would cause excesive current, when the AFE can be calibrated and readings taken?

    Please see below for a circuit snippet. VA (pin 1) connects to a switched 5V supply.

    I've already tried disconnecting the thermal pad from ground and it still draws 20mA!

    Perhaps it's just a rogue chip? I have another small batch due in next week, I'll be watching each one carefully for current consumption.

    Best Regards,

    Nick Bennett

  • 0 in reply to Nick Bennett
    TI__Guru* 92715 points
    Nick,


    I think you forgot the snippet. Use the "Insert Code, Attach Files and more..." to the lower left of the entry and use the picture icon to insert the picture. You won't be able to copy and paste a picture into the post.

    I mentioned the bad connection because it's possible that the solder bridged something to ground that wasn't supposed to be connected. This could cause extra current draw. I was hoping to see the schematic and layout to see what might be nearby.

    Also, I don't think that it's a rogue chip. Each device is tested and if the device came from a distributor or regular sales channel, we can track the device lot code and when it was tested and shipped.


    Joseph Wu
  • 0 in reply to Joseph Wu
    Prodigy 120 points

    Hi Joseph,

    That's misleading, the pasted circuit was on my screen!

    I've tried again but using 'paste from word' option.

    Cheers,

    Nick

  • 0 in reply to Nick Bennett
    TI__Guru* 92715 points
    Nick,


    Can you please try again? The circuit still didn't come through. Again I'd use the "Insert Code, Attach Files and more..." link to the lower left of the entry and then use the picture icon to put the image into the post.


    Joseph Wu
  • 0 in reply to Joseph Wu
    Prodigy 120 points

    Hi Joseph,

    Third time lucky!

    Nick

  • 0 in reply to Nick Bennett
    TI__Guru* 92715 points
    Nick,


    I was hoping there would be more of the circuit to look over. At this point, I don't see any red flags about the layout. Any extra solder to cause bridging would likely draw more than 20mA of current.

    If you disconnected the landing for the thermal pad it looks like the only problem would be to find a new connection for the pins for VREF1N and VIN7. While there are other clusters of pins, I don't see any of them that could be mistakenly bridged without seriously impacting the operation of the device.

    Out of curiosity, have you determined if the extra 20mA current draw is coming in from the positive supply (or could it be coming another pin). Also, is it possible that the devices have seen some sort of over-voltage condition?


    Joseph Wu
  • 0 in reply to Joseph Wu
    Prodigy 120 points

    Hi Joseph,

    Sorry for the lack of circuit, our company is a little over protective when it comes to IPR (even on standard circuits).

    As I said there is a switched 5V regulator feeding VA and everything left of VA (all other inputs, refs, etc), to the right is all 3V3.

    I did check very carefully for shorts, but as you say this would probably be above 20mA. The current is going into VIO, but not out of any output (D6 or SDO lines), I've not checked CLK, SDI or /CS as they are all inputs.

    It shouldn't have seen any over-voltage condition.

    I suspect we may not get to the bottom of this one, I've another batch of board arriving next week, hopefully they'll all be OK.

    Thanks for all you help, it's good to have the reasurance I'm not doing anything silly with it.

    Nick

  • 0 in reply to Nick Bennett
    TI__Guru* 92715 points
    Nick,


    I'd still like to determine if the units are damaged or of this is a board build issue. I would hang on to the units that were unsoldered off of your board in case you want to take this further.

    I'll close this thread for now, but if you have problems with the next build, post back and we can pursue this problem further.


    Joseph Wu