• State
  • Locked Locked
  • Replies 23 replies
  • Subscribers 65 subscribers
  • Views 6033 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.

bq40z60 - Charge Overtemperature (COT)?

Charley A.
Intellectual 430 points
Other Parts Discussed in Thread: BQ40Z60, BQSTUDIO, CSD18504Q5A, CSD17308Q3

Hello,

I am seeing COT and CTOS safety alerts on my bq40z60 charger.  The reference manual states:

The charger has an independent temperature comparator with a trip point defined by TSHUTDOWN in the
datasheet. Once the overtemperature condition is detected, the charger will be disabled. The charger will
not be re-enabled until the temperature is below TSHUTDOWN – THys for COT:Recovery Delay.

What temperature is this comparator monitoring?  My cell temperatures remain below 26 C and my Int/Fet temperature remain below 70 C according to bqStudio.  

The datasheet states that TSHUTDOWN as 135 C.  Both temperatures are well below this limit.

Thanks,

Charley

  • 0
    TI__Guru 73830 points
    Charley
    There is a second internal temperature sensor located near the charger FETs to shut them off if the temperature exceeds the Tshutdown threshold. This temperature cannot be read. Do you see it move to Safety Status and shutdown the charger?
    Tom
  • 0 in reply to ThomasCosby
    Intellectual 430 points
    The charger does not shutdown. Both XDSG and XCHG remain clear. It discharges normally but does not charge when COT is set.
  • 0 in reply to ThomasCosby
    Intellectual 430 points
    Hi Thomas,

    After some investigation, we found the the overheating issue was partly due to insufficient thermal pad to ground plane connection. But mainly due to our 40v high and low side FETs. Since our adapter voltage is 24V, we needed to upgrade the 30V FETs in the reference design to 40V. Unfortunately, the rise time of these FETs was measured to be 10x longer than the 30V FETs due to their higher gate charge requirements. This was stressing the internal charge FET.

    Do you have a recommendation for 40V Fets that will be compatible with the bq40z60?

    Thanks,
    Charley
  • 0 in reply to Charley A.
    Intellectual 375 points

    I'm very interested in this as well - we are also upgrading to 40V FETs.


    Thanks.

  • 0 in reply to Jim OSullivan
    TI__Guru 73830 points

    I have not tried a 40V FET, but I will find an acceptable alternative to recommend.

  • 0 in reply to ThomasCosby
    Intellectual 375 points

    Would TI CSD18504Q5A be a suitable 40V alternative to the 30V FET in the EVM?

    Thanks.

  • 0 in reply to Jim OSullivan
    TI__Guru 73830 points

    Jim

    I was looking at that FET earlier today, but the gate charge is double that of the 30V FET. That will increase the slew rate and losses in the charger, but I do not know how much heating will occur. The spec show that all of TI's 40V FETs have at least this much charge. You could try it. I have not had a chance to look outside TI for a candidate.

    Tom

  • 0 in reply to ThomasCosby
    Intellectual 430 points
    Unfortunately we have not yet found a 40V FET with low enough gate capacitance to stay below the temperature safety threshold. Our adapter voltage is 24V and we have been experiencing about a 10% failure rate in our circuit with 30V high and low side charging FETS. After these failures we measure about a 100 ohm impedance from gate to source in these FETS. We also see burned PCB traces on the voltage sense inputs – specifically VC3.

    We are using the TI CSD17308Q3 (N-CH 30V 13A 11.8mOhm 5.1nC 8SON).

    We have also tried these 40V FETs:
    • Fairchild FDMC8327L
    • Vishay SIS434DN-T1-GE3
  • 0 in reply to Charley A.
    Intellectual 375 points
    Thomas,

    Could you offer guidance on the voltage sense input traces - how much current should these be sized for?

    Thank you.
  • 0 in reply to Jim OSullivan
    TI__Guru 73830 points

    The voltage sense traces are just used for sensing and cell balancing. If you use internal cell balancing, then the current is in the 10mA to 20mA range.

  • 0 in reply to ThomasCosby
    Intellectual 375 points
    Thanks Tom. That's what I had thought but then I saw Charley's mention of burned traces on the voltage sense inputs and just wanted to understand what that was about.
  • 0 in reply to Jim OSullivan
    Intellectual 430 points
    The burned traces occurred after the FETS failed and damaged multiple components on the board.
  • 0 in reply to ThomasCosby
    Intellectual 430 points
    Tom, would this type of gate driver allow the bq40z60 to interface with 40V FETS?
    www.ti.com/.../description
  • 0 in reply to Charley A.
    TI__Guru 73830 points
    Charley
    The device was not designed to support that type of switch. You will need to use n-channel FETs.
    Tom
  • 0 in reply to ThomasCosby
    Prodigy 50 points

    Hi Thomas,

    I work with Charley. We have been investigating inserting the UCC27211ADRM gate driver between the BQ40z60 and 40V NFETs. The voltage level on HIDRV output pin of the BQ40z60 is not directly compatible with this interface. Can we interface the HIDRV output pin on the BQ40z60 to the HI input pin of the gate driver by grounding the PH pin on the BQ40z60 and removing the bootstrap cap between the PH pin and BTST pin on the BQ40z60? It looks like the LODRV output pin of the BQ40z60 can drive the LI pin of the gate driver without any changes. If this can't work due to other functions associated with the PH and BST pins-  is there another simple method like inserting a discrete or two to get fast voltage translation on the HIDRV pin output?  Or is there another gate driver that can be used to drive the NFETs? 

  • 0 in reply to Ron Whitney
    TI__Guru 73830 points
    Ron,
    The designer does not recommend removing the PH and BTST from charger control. He pointed out that the gate charge on the 40V FETs may not be the problem. The device only supports up to 26V for VACP and at 1MHz switching, the AC losses (such as the diode reverse recovery charge) means input voltages much above 20V are going to generate a good deal of heat and the 30V FETs should be able to support these requirements. Have you tried reducing the adapter voltage to see if this reduces the temperature?
    Tom
  • 0 in reply to ThomasCosby
    Prodigy 50 points

    Hi Thomas,

    We have observed that the BQ40z60 runs much hotter when the input voltage is above 20V. In light of all the issues at 24V input to this circuit we have decided to reduce the input voltage to 18V on this circuit by adding a step down converter our 24V. 

  • 0 in reply to Ron Whitney
    Prodigy 50 points
    Hi Thomas,
    We have two BQ40z60 circuits that are diode-ored with the ac adaptor using three power path controllers (LTC4412) so when a battery is being charged no load current is supplied from the battery charger circuit- it is supplied by the AC Adaptor through the external power path controller. We are experiencing some 30V NFET failures in the charger circuit. Would like to prevent a high discharge current from the battery into the shorted 30V NFET when this occurs. Referring to the figure 9 schematic on pg 32- can we insert a 3A schottky diode between the node of the 10mOhm, 2.2uF cap and the node labeled System Load? We have a zero ohm resistor in this location at the moment for troubleshooting any problems on a board in this area. We can change this to a 3A schottky diode. I can send you a PDF of our schematic showing this if you tell me how. If we have already sent it to you previously it's A-R1 for battery 1 & F-R11 for battery 2. Still investigating the 30V NFET failures. Would like to discuss this as well.
  • 0 in reply to Ron Whitney
    TI__Guru 73830 points

    Ron
    Are you still having issues with this configuration. I can contact you offline to review the schematics. 

    Tom

  • 0 in reply to ThomasCosby
    Prodigy 50 points

    Hi Thomas,

    Yes, would like to take this offline as we are still having issues mostly related to our layout. We are doing a new layout and would like you to review it. Also would like discuss additional simple protective measures for single faults including failures in the BQ40z60RHB circuitry. I sent you an email with the schematics about this on 8/22/2016 at 4:52 PM.

  • 0 in reply to Ron Whitney
    TI__Guru 73830 points

    Ron,

    I found the schematic in my Inbox, but I have not reviewed it yet. I will review it and respond.
    Tom

  • 0 in reply to Ron Whitney
    Prodigy 50 points

    Hi Tom,

    When do you plan to contact me offline to review the schematic and layout?

    Ron 

  • 0 in reply to Ron Whitney
    TI__Guru 73830 points
    Ron
    I will review your schematic within the next couple of days.
    Tom