• State Resolved
  • Locked Locked
  • Replies 13 replies
  • Subscribers 63 subscribers
  • Views 3262 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.

BQ24172 trouble to turn it on

Henrique Kuehne
Prodigy 190 points
Other Parts Discussed in Thread: BQ24172

Hi,

I'm using the BQ24172 very close of its EVM (circuit attached).

When I insert the 12V in the input, the PVCC enters in a switching mode and the average voltage is low. If I reduce the Vin until around 11V its output goes to a continuous voltage, and then, I can vary the input voltage from 8V up to 17V and it continues working fine.

Can someone help me to solve this?

Please, feel free to ask for details.

Thanks in advance,

Henrique.

4667.08 - BATTERY.pdf

  • 0
    TI__Expert 3160 points

    Henrique,

    I don't immediately see any problems with your schematic.  Do you have some waveforms you can share? (ACDRV, BATDRV, SW in particular) What voltage is on REGN?

    Under what conditions are you trying to start-up with?  (is a battery connected, approximate system load)

    Your input current limit is set to 4A and you have 2-cell battery with charge current is limit of 1.5A, do you expect a very large system load?

    Best Regards,

    Rachel 

  • 0 in reply to Rachel Moore
    Prodigy 10 points

    Hi Rachel,

    Below the waveforms:

    ACDRV:

     

    BATDRV:

     

    SW:

    REGN: 

    I'm trying to start-up it with no battery connected but with a 10k resistor assembled in R111 position. As load I have a switched regulator in the output (PWR_IN) to 5V and I'm using an 8 ohms resistor in this output, so the total load consumes around  3W. Maybe I will need around 5W with all board connected.

    Thank you for attention and if you need more details feel free to contact me.

    Best regards,

    Henrique.

  • 0 in reply to Henrique Tardiole Kuehne
    TI__Expert 3160 points

    Henrique,

    It would be better to have a scope plot with both ACDRV and BATDRV on it, but It looks like something is causing the IC to switch the power path selection from VIN to BAT and back.  I'm wondering if there is a leakage path from system to battery?

    If you can look at BATDRV, ACDRV, AVCC and SRN on the same plot, do you see if maybe the sleep comparator is being triggered (SRN voltage approaches AVCC voltage)?  You would see SRN rise until ACDRV and BATFET switch and then SRN voltage would fall until the FET drivers switch again and recharge SRN.  

    Best Regards,

    Rachel 

  • 0 in reply to Rachel Moore
    Prodigy 190 points

    Rachel,

    Below the waveform with the signals:

    CH1 = BATDRV; CH2 = ACDRV; CH3 = AVCC; CH4 = SRN.

    These waveforms were captured with no load in the PVCC and Vinput = 12.9V.

    Notice that the SRN signal is very low level and don't get the level of ACDRV or AVCC.

    Below I got the waveforms with 8 ohms load in PVCC.

    CH1 = BATDRV; CH2 = ACDRV; CH3 = AVCC; CH4 = SRN.

    I inserted the cursors to help in reading the levels.

    As you did, I also can't see any leakage path in the circuit.

    Best regards,

    Henrique.

  • 0 in reply to Henrique Kuehne
    Prodigy 190 points

    Rachel,

    I've captured waveforms of start-up.

    For the waveforms below we have:

    CH1 = AVCC; CH2: ACDRV; CH3 = BATDRV; CH4 = SRN.

    First with 11.5V at input:

    And then with 13V at input:

    Best regards,

    Henrique.

  • 0 in reply to Henrique Kuehne
    TI__Expert 3160 points

    Hi Henrique,

    Thanks for the detailed captures! It looks like OVP is being triggered (spike on AVCC). 

    I believe that you should remove CR9 to resolve the problem.  Reverse blocking protection is provided by Q16 so CR9 shouldn't be needed anyway.  

    Let me know if this helps!

    Best Regards,

    Rachel 

  • 0 in reply to Rachel Moore
    Prodigy 190 points

    Hi Rachel,

    I've tested removing the CR9 and it looked like it was working fine. But when I connected the load it stoped to work and return to switched mode.

    After this I removed the capacitors C91 and C92 from Q15 and Q16 and now it seems to be working even if I have load connected or not. Do you see any problem if removing these capacitors?

    Thanks for attention and help =).

    Best regards,

    Henrique.

  • 0 in reply to Henrique Kuehne
    TI__Expert 3160 points

    Hi Henrique,

    Sorry for the delay, I wanted to check this on the bench myself. 

    Unfortunately, I would not recommend removing C91 and C92.  They are important for controlling turn on time which is important and must be changed with BATFET timing in mind to prevent shoot-through.  You can modify the values as long as you keep the timing of BATFET in mind and verify you are maintaining a break-before-make situation between input and battery FETS.

    To really solve the problem, we need to find the cause of the voltage spike you are seeing on AVCC.  I checked on our EVM and this is not normal. With a 15-17V supply I measure a 100-200uSec ~2-3V maximum spike. 

    The fact that your AVCC voltage is jumping up to almost 2x of the input voltage is very troubling. Please try to capture the input current and voltage on AVCC and see if that tells us more. Also it would be good to probe the other pins while this is happening to see if we can find the cause. 

    In addition, you should add a 330pf Cap in parallel to R108 (labeled as C4 in the EVM schematic).  

    Thanks for your patience!

    Rachel

  • 0 in reply to Rachel Moore
    Prodigy 190 points

    Hi Rachel,

    Sorry for the long delay. I was waiting a new build from the board to test it. In this new board I did some improvements in layout and it seems to be working correctly now.

    The first layout I did not respect some rules for the current sensors and some tracks had inductance greater than it should have due to be too thin.

    Now, I'm testing and I have another issue. As you can see in the schematic, I set the ACSET to allow up to 1.5A to the system. I'm using a load that is sinking 2.2A and the controller is not acting. Can you see the reason to this?

    Thanks again for your support.

    BR,

    Henrique.

  • 0 in reply to Henrique Kuehne
    TI__Expert 3160 points

    Hi Henrique,

    The ACSET pin doesn't actually limit the current going to the system but rather it sets the total current for Battery Charging Current + System Current.  This will allow you to have an adapter with a relatively smaller current capability and the IC will dynamically adjust/reduce the charging current to ensure the system is fully supported during periods of peak power draw.  In effect, it only will reduce the charging current and will not attempt to restrict the system current.  

    The reasoning behind this is that if the IC were to limit the system current it would immediately kill/brown out the system. However, by allowing the system to continue to pull more current, the adapter may be able to support this for a short time and the peak power load of the system might pass before the adapter voltage drops and kills the system.  

    The assumption is that the user will adjust the ACSET limit to be just safely under the current capability of their adapter and that their maximum system load should be less than this, or the adapter will not be able to support the system.  In a practical sense, your typical system load + fast charge current (ISET) should be less than or equal to the current your adapter can provide or your battery will charge slowly.  

    The Bq24172 datasheet on page 19 describes the function as follows:

    Please let me know if this explanation helps, or if you still have questions or concerns about the DPM function.

    Best,

    Rachel

  • 0 in reply to Rachel Moore
    Prodigy 190 points

    Hi Rachel, how are you?

    I hope fine.

    We are doing tests with the BQ24172 and we realize that many times it does not start the load up when it is powered with voltage above 11V at input.

    We have also saw that when the load is disconnected (doing Q8 cutted) the BQ starts to work and the level of PVCC reaches its correct value. Then we can connect the load (seting Q8) and the system is powered up.

    My question is, if I remove the C92 I don't have problema with this timing, but you advise me to don't remove it because of break-before-make logic it implements with both NMOS and the C91. So, how could I calculate a tighter time to this circuit? I mean, a circuit that can be faster to connect the Vin to the PVCC, avoiding the over-voltage protection to be activated?

    Thank you in advance,

    Henrique.

  • 0 in reply to Henrique Kuehne
    TI__Expert 3160 points

    Hi Henrique,

    Can you share an updated schematic?  I know we had made some changes, but I want to make sure I have this right, I don't see a Q8 on the version I have. 

    I'm not clear why this would be related to triggering over-voltage protection. Are you doing this test without a battery present? 

    The default values for C92, C91 and the Cgs of BATFET from the EVM, typically work fine, but you can change these values to tune the performance on your PCB.  You want to ensure that the RBFET turns off before the BATFET turns on, and that ACFET controls inrush current.  

     

    Thanks!

    Rachel 

  • 0 in reply to Rachel Moore
    Prodigy 190 points

    Hi Rachel,

    Sorry the delay.

    Yes, I'm doing this test without the battery.

    The datasheet said that the BQ has internal circuit for break-before-make and that it has 10us timing in such logic. It also explain that if we need to increase this time we can use a capacitor between gate and drain of mosfet. So, I think it should have a way to calculate this increasing time or even to use only the internal 10us.

    1261.08 - BATTERY.pdf