BQ24192: How to increase charge current using bq27531-g1 FG (bq27531evm demo board)?

Hey Nitish,

So I do have a couple of questions and comments about your set up.

1) Do you have a load on the SYS pin when running this test?

2) When you set INLIM to 500mA, the input current will be capped at 500mA. To get 1.38A of charging from your laptop, you would have to connect it to a power supply like you later in your post, or a USB port on the laptop that can support at least 1.5A.

3) In regards to changing your charge current, there is a BYPASS option in the gauge to allow the user to program the charge current. I am not as familiar with the gauge portion of this EVM, but I can investigate further. A couple of notes on this: Please make sure you are providing accurate configuration settings (i.e. battery chemistry, proper learning cycle) to properly. You can also upload the .gg file as a reply to this post to see what you gauge is configured for.

4) Lastly, register 1'b011000 is the default setting for charger current (2048mA). However, this is just the setting. If the charger is undergoing IINDPM or VINDPM (i.e., you have a load on SYS pin), then the charge current will decrease to keep the power source within limits. This questions goes hand in hand with question 1).

Hi Joel,

Thanks for the reply!

1. I was using the bq27531evm demo board when I ran this test. And during charging, there was no load connected except from the board itself. I was thinking this was the reason why I was getting 463mA instead of 500mA when I charge using laptop USB. Is my understanding correct?

2. Yes, I tried connecting it to PS of setting 6V, 3A (sluua90a setting) and I got a charging voltage and current of 4.2V, ~463mA. I never saw the charge current went up from this value even after changing the PSEL to low. I also tried connecting it to my portable charger with output current rating of 2.1 A but still the charging current of the battery is ~463mA. Should I go ahead and connect it to a power adapter then? Or to a power supply but with an increased current setting (>3A)? 

3. I was actually doing the learning cycle when I observed the low charge current. I wasn't done with this process yet. But I am attaching my .gg files for your reference. The battery that I'm using is a Panasonic NCR18500A. I'm attaching its datasheet as well. Please do let me know the result of your investigation. But, is the BYPASS mode recommended for production? How do the fuel gauge determine the charge current settings?

4. I see. I will monitor this status. I don't have my demo board setup today but I will get back to you by the end of the week. But like I said, when I tried to charge it, only the demo board itself and the battery were connected to the power. I was assuming I will get an increase in charge current after I connected it to a different power source but still it was capped off at 463mA.

bq27531_calibrated_21Dec - Copy.xlsbq27531_calibrated_21Dec.gg - Copy.xls

Panasonic NCR18500A Data Sheets[2367].pdf

Hi Joel,

The DPM_Stat during the charge is 0. There was also no load attached to the SYS pin.

Today, I tried to charge a battery with Vbat initially at 3.375V. I connected to a USB power adapter with 2A output, PSEL=0, and no load at the SYS pin. The charge current was at 560 mA, while the charging current is at 4.196V. After some time, the charge current decreased as the Vbat increased.

This CC charge current is what we'd like to be increased to 1.35A. According to figure 14 of the bq24192 datasheet, the Vbat = 3.375V should already be in the CC phase when charged.

Let me know your insights.

Thanks,

Hey Nitish,

1) In regards to the comment about the laptop port, that was my initial thought. You would see a current limit (i.e. reduced charge current) if your laptop port cannot handle more than 500mA of output. 

2) At this point, it does now sound like a power supply limitation. You most likely will not see a difference using a different power supply.

3) BYPASS should not have to be used in production if all gauge parameters are correct. However, I recommend using it now in testing to determine whether the gauge or the charger is the limiting the charge current, according to this snippet of the datasheet below: 

And the gauge uses a complex algorithm to calculate the optimal charge current using different parameters such as Qmax, SoC, chemistry, whether in CCCV or MLC charge topology as examples.

Now from your most recent post, it seems like the charging cycle is already in CV mode, where the charge current will steadily decrease until you hit the charge termination limit. What is the SoC the gauge is reporting at the time? When you mentioned the charging voltage was 4.196V, was this measured or reported by the gauge. It would be helpful to have the register values of the gauge at the time of these measurements.

Hi Joel,

Thanks for showing me this table. I will let you know once I tried this command. This is helpful! 

But for now I tried to restart the learning cycle.

I discharged the battery further and relaxed it for 5hrs. The Vbat was at 2.8V after 5hrs (measured using multimeter across the battery). The terminate voltage spec of this batt is at 2.5V

I then started to charge it using 2A output USB power adapter. I observed that the charging current started at 742mA this time (gauge's measurement) with charging voltage at 4206 mV  (gauge's measurement). I recorded the log of the charge cycle and attached it here for your reference.

The Vbat was already at 2.8V when I started to charge it but still it didn't reach the 1.35A charge current. And this quickly started to decrease (see log file). Does this mean that the charging cycle started immediately at CV mode (like you said)? Could this be the reason why I never see it to charge with 1.35A (because it started in the CV mode)? 

I continued to charge it for 3hrs until the SOC is 100% as shown in bqstudio.

This time the current was fluctuating at around 180mA (gauge reading) and charging voltage also fluctuating around 4068 mV (gauge reading). The taper current is set to 190mA. D1 (STAT bit) is also blinking way faster than 1Hz, and the CHRG_STAT bits are continuously switching from 10b to 11b and back to 10b. The FC bit was kept cleared, and VOK remained set (indicative that the gauge didn't recognize the Full Charge state). Does this mean that my learning cycle failed? 

How can I make it to start charging in a similar profile shown in Figure 14 of slusaw5b (Pre-charge, CC, CV)? Could this be because the learning cycle isn't finish yet? Or is there anything wrong in what I'm doing?

Would the learning cycle fail given the above circumstances (i.e. start to charge at CV mode)?

Thanks!!!learning_cycle_cell_2 - Copy.txt

Hi Joel,

Also, is there a way for the host to configure the INLIM and override the setting derived from PSEL and OTG states?

In our board, PSEL is eternally tied to 1. So, with OTG=1, the INLIM setting will be 500mA. In production, we want the firmware to update the INLIM bits depending on the power source connected. Can we do so via i2c commands? Actually, I have tried it but it get resets to the default of 500mA every WDT expiration (I think..).

Hey Nitish,

To your questions:

Does this mean that the charging cycle started immediately at CV mode (like you said)? Could this be the reason why I never see it to charge with 1.35A (because it started in the CV mode)?

It certainly looks like that. From the log file you provided, the first plot point for the battery voltage is 4.206V, already near the regulation voltage.

I suggest testing with a battery simulator if you have one (Keithley source meter, Kepco bidirectional power supply, etc) at a fixed voltage (3.2V < VBAT < 3.9V)

Does this mean that my learning cycle failed? 

You said STAT pin is blinking faster than 1Hz. This sounds to me like a battery not present condition. Make sure you battery is connected or that indeed it is a good battery. Also suggest discharging the battery and holding it at a constant voltage (i.e. 2.6V and holding it at that voltage). Then try to charge on your own by applying a pre-charge current (200mA is fine) until the battery voltage is above 3V, then apply 2A of charge and see what the battery voltage measures.

The above test I described is without the gauge+charger. This will tell us if the battery is good or not. 

How can I make it to start charging in a similar profile shown in Figure 14 of slusaw5b (Pre-charge, CC, CV)? Could this be because the learning cycle isn't finish yet? Or is there anything wrong in what I'm doing?

It could be related to a bad battery, a poor battery connection, a poor learning cycle with an inaccurate Qmax. How are you performing the learning cycle?

Regards,

Joel H