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.

Step up regulator LMR62014 as input to BQ24250

Other Parts Discussed in Thread: BQ24250, LMR62014

Hi Everyone,

I'm using a boost converter LMR62014, to generate the input voltage for the battery charging PMIC BQ24250. This configuration is chosen as the minimum input for the BQ24250 to function is 4.2V (given the VDPM value is updated in the respective register) and my source voltage is around 3.7V. For this reason I use the LMR62014 to step up the source voltage (VSYSTEM) to 4.5V as shown in my schematic below:

The problem is when I bring the BQ24250 out of HIZ mode, the voltage at the input crashes, causing the internal registers of the BQ24250 to be reset. When bringing my system up, I do the following steps:

  1. Enable HIZ mode on the BQ24250, no current drawn. At this stage the LMR62014 is disabled and the input to the BQ24250 is sitting just below VSYSTEM at around 3.6V (due to D4)
  2. Enable the watchdog to fire every 10s  - this ensures the device remains in host mode
  3. Enable the LMR62014 - this brings the input to the BQ24250 up to 4.5V (stable)
  4. Adjust the BQ24250 VDPM setting to 4.2V - The impact of this is verified in register 5 (0x04) of the BQ24250 as the device is no longer in the VDPM regulation loop
  5. At this point all registers are checked and verified
  6. Bring the BQ24250 out of HIZ mode and set the current limit to 100mA - at this point all registers appear to reset.

A trace of the VSYSTEM and input to the BQ24250 shows that VSYSTEM is stable at 3.7V (CH1 - Yellow), and the input to the BQ24250 (CH2 - Blue) is crashing from around 4.5V to around 2.5V as soon as the device is taken out of HIZ mode and allowed to draw current. From this we can conclude that bringing the BQ24250 out of HIZ mode is crashing the LMR62014. I have tried increasing the input capacitance, to damping the transient load from bringing the BQ24250 out of HIZ mode - however this doesn't resolve the problem.

So following this, I have a two questions.

  1. What is the in-rush current when the BQ24250 is taken out of HIZ mode?
  2. Could the switching noise from the LMR62014 be causing problems for the BQ24250 (or vice versa)?

Any advice on this issue would be really appreciated! In particular if the team at TI E2E could take the evaluation modules for these two devices and try replicate this scenario, that would be awesome!! ;)

All the best!

  • I'm going to get an EVM of LMR62014 and try it with bq24250 on bench. Will let you know what I find.

    Thanks,

    Wenjia

  • Thanks Wenjia! I'm looking forward to see if you can replicate this issue - let me know if you need additional information regarding my set up.

  • Hi Wenjia,

    Did you have any luck with the EVM today? I had a look at it as well, it seems like the module is fixed at 20V, so you'll have to change the resistors on the feedback network to get the 4.5V output in my use case.

    Cheers,

    Michael

  • Michael,

    It takes ~3 days for the LMR62014 EVM to deliver to me. Will update you once I have looked at it on bench.

    Thanks,

    Wenjia

  • Michael,

    Do you have battery connected? What is the VBAT and ICHG when you saw registers reset?

    What is the load on system?

    Thanks,

    Wenjia

  • I duplicated your setup on bench today and was able to re-produce the step 1 to step 5 in your previous description. However, I didn't see the drop on VIN and all register value still held.

    Thanks,

    Wenjia

  • Hi Wenjia,

    The battery is not connected, and ICHG is at its default value.

    If the battery is in the circuit, I would imagine the regulator wouldn't be hit so hard by the inrush current...

    Could you check without the battery? What is the schematic of your setup at the moment.

    Cheers,

    Michael

  • Michael,

    I've changed the resistors on the LMR part EVM and was able to output the same 4.5V voltage to VIN of bq24250. As you were doing the test in host mode, the external resistor setting on my bq24250 EVM doesn't matter so I didn't change them to match yours.

    I didn't connect the batter in my test either, and you should see the same as following.

    - With no battery connected, I saw battery detection routine on VSYS as soon as I brought the IC out of Hiz state.

    - With Hiz enabled, there should be no voltage output on VSYS.

    Do you have a load on VSYS of bq24250? With 100mA input ILIM, if the load is too big, it is possible that VIN will drop to below VIN_DPM.

    Below is the waveform I saw when I wrote 0 to Hiz_mode bit. I had ~300mA load on VSYS.

    Thanks,

    Wenjia

  • Michael,

    Can you capture the signal of the input to your boost converter when you bring the IC out of Hiz mode?


    Thanks, 

    Wenjia

  • Hi Wenjia,

    Thanks for the replies, I have been doing some investigating today to find the source of this issue. The trace of the input to the boost converter (i.e. VSYSTEM) is shown above in yellow (CH1) in the first post on this thread - it appears to stay stable during the crash.

    I have being doing some tests today with the BQ24250 in HIZ mode, that show when I place a 50Ohms resistor between the output of the boost converter and ground, 90mA flows as expected (4.5/0.05), however the signal on the scope appears to be on the verge of becoming unstable (sinusoidal peak-to-peak fluctuations of between 0.3-0.4V).

    Tomorrow I will drop this resistance further to see at what load the boost converter crashes (if it crashes). The boost converter circuit was designed and evaluated in the TI WEBench, and I want to now manually recheck the calculations for the coil and output capacitor and find out the theoretical output current limit.

    When you changed the resistors on the EVM for the boost converter, did you change anything else? What are values/part numbers of the inductor/output capacitor(s) in use?

  • Hi Michael,

    I didn't touch anything else on the boost converter EVM except for R1 and R2 (I changed them according to R7/R11 in your schematic). The other BOM is in the below user's guide.

    http://www.ti.com/lit/ug/snva500a/snva500a.pdf

    Thanks,

    Wenjia

  • Hi Wenjia,

    I had a look for that user guide / manual before but was unable to find it. I would say thanks for posting the link, but it seems that you have posted instead a link to some property / a villa somewhere in Texas... A copy/paste error perhaps, or are you suggesting I move to Texas?

    Cheers,

    Michael

  • Hi Michael,

    I'm so sorry!!! must be the cache...Here's the link

    http://www.ti.com/lit/ug/snva500a/snva500a.pdf

    Welcome to Dallas, however maybe not at this moment:)

    Thanks,

    Wenjia

  • Hi Wenjia,

    So after a lot of testing and various experiments in many configurations - I have concluded the instabilities that I am seeing, are a result of a rookie mistake.

    For what ever reason, when designing this board, I have not put in place anywhere near enough decoupling capacitors. My theory for the regulator voltage dropping and some situations (such as when I have the battery connected to the BQ24250) even spiking, is that the switching noise from the other circuitry is being detected on the feedback pin, and since the feedback pin is responsible for regulating the output, this is the reason I am seeing drops and spikes occurring at various periods.

    It appears as I draw more current the switching noise increases (the source of which is not exactly known - there are several locations which should contain decoupling capacitance which don't) which in turn causes the regulator to become more and more unstable. This instability when manifested as a voltage drop is what was causing the BQ24250 to lose power and configuration.

    Do these conclusions sound reasonable to you?

  • Hi Michael,

    This can very likely be the cause. Thanks for the follow up, please let me know if you have other questions.

    Thanks,

    Wenjia