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.

BQ25713: Observed higher voltage at HIDRV1 and HIDRV2. Can it cause damage to the MoSFETs?

Part Number: BQ25713
Other Parts Discussed in Thread: CSD17551Q3A

Here we are seeing MOSFET failure which is driven by HIDRV1 in our board with BQ25713. As per datasheet HIDRV and LODRV pins are driven by REGN So, we are expecting output of the HIDRV and LODRV pin should be below or around 6.5V. But in our case we are observing around 20V with spikes up to~24V in the EVM. We have chosen very similar MOSFETs (RQ3E100BNTB) instead of using CSD17551Q3A (one which used incEVM-017). 

Here are some pictures of HIDRV1 & HIDRV2 captured in the EVM: 

When absolute maximum Vgs(Max) mentioned in datasheet of N-MOS is +/-20V, signal coming from BQ25713 can damage if it is more than absolute maximum. But we did not understand how MOSFET used in EVM is handling this voltage.

If you have solution or better explanation please write back to us. 

RQ3E100BNTB

  • We were measuring across PGND and TP13 to know Vgs, It was wrong. When we observed signal across Vgs of HIDRV1 mosfet we got around 8.4V in EVM. But in our custom design we are not able to write to 'Charging current register', when we read it back it shows 0000 (Rest all registers set and we can read them back). 

    Schematic : 

    CHG_IN is a supply with 12V input. We observed short between drain and source of Q6 after the first power on. And we are observing audible clicking noise of few Hz from our circuit, we're investigating from where. Before programming BQ25713 with charging current there was no issue. Note that we've used NMOS RQ3E100BNTB instead of using CSD17551Q3A (one which used in EVM-017) . Apart from this the schematic is quite close to the EVMs.  

    Here are the signal observed in oscilloscope :

    Red signal is output of LODRV1 and another one is from HIDRV1. 

  • Hi Ravikumar,

        The LSFET will be driven by LODRV which will switch/off between REGN and GND to turn on the low side NFET. The HIDRV is driven by VIN + REGN, as the HIDRV circuitry comprises of BTST voltage. BTST + plate is charged to REGN while - plate is held at SW potential. This allows us to drive the high side NFET with a positive VGS, as S = SW, G = HIDRV and HIDRV = REGN + SW. VGS is limited to the REGN voltage which will be less then rated 20V VGS of MOSFET,as you have to look at gate voltage relative to source voltage and not gate voltage relative to ground.

    I would recommend testing on the EVM first to observe expected behavior.

  • Hello Kedar,

    Thank you for the insights. 

    You are addressing the misunderstanding from our side which we have already resolved. Here issue is High drive Mosfet at Buck side is going bad after we plug in charger to our custom board. Our schematic is exactly like the EVM, with only a change in the MOSFETs. Which behavior should we compare between the EVM and our board?

    Please verify our circuit and if you can address some design issue/change in it would be helpful. Also we can share the layout, please let us know the best format to do so.

    Regards,

    Ravikumar G.

     can you please look into it?

  • Sorry Ravi, I don't support charger products. Please work with Kedar to get the issue resolved.

  • It's okay, Damian. Thank you for your response.

  • Hi Ravikumar,

    •  From your waveform of HIDRV1 and LODRV1, you mention LODRV1 is red? From the waveform it looks like LSFET1 is being driven while HSFET1 is off except for a short pulse, so I would not expect you to observe a short across HSFET1. 
    • Regardless of the output of LODRV1 and HIDRV1 (in boost mode you should expect HIDRV1 to be on except for short pulses where the bootstrap capacitor has to be charged to keep the HSFET1 on), your input voltage of 12V, and output charge regulation voltage of 16.8V makes it much more likely that you  are operating in Boost mode, so you would want to observe the SW2 signal along with HIDRV2 and LODRV2. HIDRV1 and LODRV1 drive the FET of the buck stage. Please refer to Table 2. MOSFET Operation of the datasheet
    • For audio noise, use the bit EN_OOA to set minimum PFM burst frequency to above 25 kHz to avoid audio noise.