LTC4155 - Strange behaviour of ICHARGE[3.0]

Hi!

We've been evaluating the LTC4155 for the past week and so far it's a pretty impressive part to pair with USB Type C powered designs. However we've come to a strange issue:

When changing the ICHARGE[3.0] register contents, the devices acknowledges the change (reading back the register returns a correct value) BUT battery charge current goes wild when the contents of USBILIM[4.0] are changed...

TEST CONDITIONS:

- CLPROG1 shorted to CLPROG2 -> resistor to ground (1.21K)

- RPROG = 470R

The contents of ICHARGE[3.0] are written to the device after a POR event. Right now charge current should be limited to 0.638A ((1.2/470)*1000)*0.25 <- (ICHARGE[3.0] = 0b0011xxxx)), however we're measuring currents as high as 1.1A.

So far, when there's a valid VBUS device has an input limit of 100mA. Our software is capable of increasing the current limit in three discrete steps (500mA, 1500mA and 3000mA). However when changing the USBILIM contents, the voltage observed at the RPROG pin fluctuates widely, consequently increasing charge current.

The only setting that seems to do what's supposed to do is to disable the charger ICHARGE[3.0] = 0b0000xxxx

Why is this happening, vprog shold be a set value regardless the maximum input current, right?

  • 0
    •  Analog Employees 
    on Jan 28, 2019 5:21 PM

    Hello!

    I'll answer your last question first: no, actually, VPROG should change depending on how you've configured the ICHARGE[3.0] register. See the VPROG SERVO column in Table 11:

    So, with your ICHARGE[3.0] = 0011 value, you should see 0.3V on the PROG pin at full current. That said, your calculation was still correct because you multiplied the full-scale voltage by the full-scale current percent (1.2V * 25% = 0.3V). With your settings as described, I would expect the same 0.638A charge current limit.

    It would help me if I understood what the PROG pin looks like under these conditions. If it is truly all over the place, might there be an assembly issue? I also wonder if the act of measuring the VPROG voltage might be disturbing the operation (adding capacitance, maybe noise).

    Does this happen across multiple units?

    Regards,

    Zack