LTC4020 (based on DC2134A demo board) charger troubleshooting.

Hello, Team

 

We are evaluating the LTC4020 demo circuit D2134A for a Lithium Polymer 11S battery charger design.  We are modified this demo board to adapt it to our charger requirements:

 

Vin = 46 to 51V

Vout = 39,26 to 50,81V

Vfloat = 46,2 V

Iout max = 10 A

Icharge = 4,2 A
Battery: LiPo 11S1P (40,7V,  4200 mA/h)


Here are the relevant components modified on the demo board for our charger design:

R8 = R9 = 187kΩ 0.1%

R10 = R11 = 10,7kΩ 0.1%

C14, C18, C27, C28: C5750X7S2A106M230KB (2220-X7S-10 uF±20%-100V) TDK

C15: EXR 63V 470uF±20% (Hitano)

D7: BZT52-B6V2J (6,2V 0,1A zener) Nexperia

M2, M3, M4, M7: NVTFS010N10MCL (100V, QG(TOT) = 13 nC) ON Semiconductor

RCBAT1: WSL2512R0120FEA (2512, 12 mΩ±1%, 1Вт) Vishay

Switching power supply frequency: 250kHz (no changed)

JP1 set to CC/CV position

 

The charger works, the battery is charging, but sometimes the LTC4020 IC burns and fails for no apparent reasons. Thus, the battery can be charged normally 3-4 times, and the fifth time some kind of failure occurs and the IC fails.

Voltage waveforms observed at the gates of transistors M2 (yellow), M4 (green), M7 (blue), M3 (red) during battery charging are shown in Figure 1 (Vin = 50V).

Thus, the charger based on LTC4020 IC works very unstable.

We tried to replace transistors M2, M4, M3, M7 with NTTFS5C673NL (On Semiconductor) transistors with a lower Qg(tot) = 4,5 nC, but with them the LTC4020 chip failed even faster, although in theory it should have worked more stable.

We tried to change the resistors RCBRT1 and RCBRB1 to 5mΩ WSL25125L000FEA (Vishay) resistors (Imax = 10A) and the inductor to IHLP8787MZER220M51 (Vishay), but the IC failed again.

I would like to understand what could be the reasons of IC fails and get recommendations on the external components of the LTC4020 IC.

Picture 1

Is there any recommendation for step-by-step debugging of the LTC4020 IC-based charger design to test all it’s modes of operation. We tried to connect the NTC-thermistor to the NTC pin and simulate the heating of the battery, but when a thermistor resistance achived 5 kOhm, the IC failed for some reason. In this regard, there are questions:

1) is it possible to shorts the NTC contact during battery charging to check the operation of the NTC-mode?

2) can a fast change in resistance of the thermistor lead to the failure of the LTC4020 IC?

3) how to check the operation of the NTC-mode correctly and safely for the IC?

 

Can you help with the values of the ITH (R15, C6) and VC (R16, C8) pin components based on our charger requirements?

What other components need to be changed to ensure the stable operation of the LTC4020 IC based charger according to our requirements?

What is the relationship between the maximum output current Iout of the converter and the current of the inductor, since the choice of the correct value of the current sense resistors is not clear?

 

Is there a more detailed description of the operation of this IC, since the slightest mistake leads to the failure of the IC? As a result of debugging the operation of the charger, about 10 ICs have already failed, and always under different circumstances. After the failure of the IC voltage level at the INTVcc pin instead of 5 V is about 2 V and the IC heats up a lot. At the same time, no one external components failed.

Although the M1 transistor failed once when hot plugging the battery, when the charger was power on.

One time IC failed when we ON/OFF it by SHDN signal when battery charged process was almost finished.

 

Is there a calculator for calculating the values of external LTC4020 IC components and operating parameters, or an example of calculating an LTC4020 IC based charger?

 

Thanks in advance for any help.

Kind Regards,

Sergey

Parents
  • +1
    •  Analog Employees 
    on Jun 8, 2021 9:36 PM

    I have studied your component substitutions, and they are all good.  So, it is unknown why your circuit is failing.

    Also, the gate waveforms look OK.

    Concerning your questions about the NTC,

    1. It is OK to short NTC pin to GND while the converter is operating.  This is a good way to simulate end-of-charge behaviour.  M1 PFET will turn off, and VOUT will regulate through R9, R11 so that VFBMAX=2.75V.
    2. There is no problem to have a fast change of NTC resistance or to short NTC pin to GND.
    3. See following discussion.

    One case of failure you described was when NTC=5kΩ.  The NTC pin sources 50µA, so the voltage at NTC pin will be 0.25V.  This is very near the low-resistance threshold of 0.3V (datasheet, p. 4) that will cause charging to terminate.  Is it possible that R7-R9-R11 voltage divider from VOUT to VFBMAX pin is not properly connected?  If it is not connected the LTC4020 will command an increasing output voltage until the LTC4020 fails.

    Other ideas for troubleshooting are

    • Check to see whether the converter is oscillating.  I suggest to look at inductor current or VOUT voltage with a horizontal sweep rate of ~2ms/div.
    • Probe the high-voltage LTC4020 pins - 1, 2, 6, 7, 30-38.  Do the waveforms look OK?
    • Also probe INTVCC for stable 5V output.

    I will offer further help as required, but first I want to fix the circuit so it does not fail.

Reply
  • +1
    •  Analog Employees 
    on Jun 8, 2021 9:36 PM

    I have studied your component substitutions, and they are all good.  So, it is unknown why your circuit is failing.

    Also, the gate waveforms look OK.

    Concerning your questions about the NTC,

    1. It is OK to short NTC pin to GND while the converter is operating.  This is a good way to simulate end-of-charge behaviour.  M1 PFET will turn off, and VOUT will regulate through R9, R11 so that VFBMAX=2.75V.
    2. There is no problem to have a fast change of NTC resistance or to short NTC pin to GND.
    3. See following discussion.

    One case of failure you described was when NTC=5kΩ.  The NTC pin sources 50µA, so the voltage at NTC pin will be 0.25V.  This is very near the low-resistance threshold of 0.3V (datasheet, p. 4) that will cause charging to terminate.  Is it possible that R7-R9-R11 voltage divider from VOUT to VFBMAX pin is not properly connected?  If it is not connected the LTC4020 will command an increasing output voltage until the LTC4020 fails.

    Other ideas for troubleshooting are

    • Check to see whether the converter is oscillating.  I suggest to look at inductor current or VOUT voltage with a horizontal sweep rate of ~2ms/div.
    • Probe the high-voltage LTC4020 pins - 1, 2, 6, 7, 30-38.  Do the waveforms look OK?
    • Also probe INTVCC for stable 5V output.

    I will offer further help as required, but first I want to fix the circuit so it does not fail.

Children
  • Pin2_BST1

    Pin2_BST1

    Vout

    Vout voltage (2ms/div)

    Pin6_SENSTOP

    Pin6_SENSTOP

    pin7_SENSVIN

    Pin7_SENSVIN

    pin32_SW2

    Pin30_BST2

    pin34_INTVcc

    pin34_INTVcc

    pin35_PGND

    pin35_PGND

    pin38_SW1

    pin38_SW1

    pin36_PVin

    pin36_PVin

    I don't repeat M2 (pin1_TG1), M4 (pin37_BG1), M3 (pin31_TG2) and M7 (pin33_BG2) gate waveforms, since it has not changed.

    Voltage divider R7-R9-R11 from Vout to VFBMAX pin is properly connected, and 3-5 thimes we were charge the battery succesfully from start to end, but at the next time LTC4020 IC was failed by unknown reason (when SHDN On/Off, when power on/off, when we tested NTC work and etc.)

    Thanks for help.