Linux: BQ27220 - How to enter lowest power consumption mode

Hi HB
The device wakes every 20 s to take voltage, current and temperature measurements and performs SOC calculations to maintain accuracy so that is inevitable.

Are you running your tests with your board or an external EVM? What are your pull up resistors pulled up to? VDD or some external voltage source?

thanks
Onyx

Hi Onyx,

>>The device wakes every 20 s to take voltage, current and temperature measurements and performs SOC calculations to maintain accuracy so that is inevitable.

Do you mean we cannot stop wake-up??

But in 

TRM (SLUUBD4.pdf) 4.2.6 Wake-Up Comparator

" Setting both [WK_TH1] and [WK_TH0] bits to 0 disables this feature.", >>> what does this mean?? not to disable Wake-up?

We test on our board, not EVM.

I2C is pull high to 3.3V, but system shutdown will be 0v. VDD is 0v when system shutdown.

Only battery (max 4.2V) will be used at this moment to measure sleep current.

For enter sleep mode 9uA current, we don't need to set any additional configurations?

HB

Hi Hb,
You cannot prevent the gauge from waking up to carry out measurements when in sleep mode. You can disable the gauge from exiting sleep mode due due to the wake comparator, which triggers if the current exceed the limits set while the gauge is still within the 19 secs out of 20 that it is in sleep mode. In this case the gauge will only exit sleep mode if in the 1s out of 20 it wakes and the current is above the sleep current threshold.

thanks
Onyx

Hi Onyx,

I set 

1) Set Dsg Relax Time to zero (0x922E=0x0)

2) disable the wake comparator (Operation Config A [WAKE_EN, WK_TH0, WK_TH1] all zero) (0x9206=0x480)

Can above two settings help achieve lower sleep current?

or actually we don't need to set any configurations, sleep current is already the lowest value?

thanks

HB

note:

Here is our measurement result that shows Fuel Gauge BQ27220 sleep current in 60 seconds:

already enter sleep mode and have lowest power consumption??

Battery=3.6V Average (60sec)= 12.54 　 Time I_BQ27220 (uA) 1  15:20:59.4   14.20 2  15:21:00.4   14.50 3  15:21:01.4   14.60 4  15:21:02.4   13.60 5  15:21:03.4   12.30 6  15:21:04.4   11.00 7  15:21:05.4   10.10 8  15:21:06.4   9.30 9  15:21:07.4   9.20 10  15:21:08.4   8.90 11  15:21:09.4   8.90 12  15:21:10.4   14.40 13  15:21:11.4   26.60 14  15:21:12.4   48.20 15  15:21:13.4   12.50 16  15:21:14.4   11.30 17  15:21:15.4   9.20 18  15:21:16.4   10.00 19  15:21:17.4   9.20 20  15:21:18.4   8.90 21  15:21:19.4   8.90 22  15:21:20.4   8.90 23  15:21:21.4   14.70 24  15:21:22.4   14.20 25  15:21:23.4   13.10 26  15:21:24.4   11.60 27  15:21:25.4   9.20 28  15:21:26.4   9.40 29  15:21:27.4   9.20 30  15:21:28.4   9.30 31  15:21:29.4   8.90 32  15:21:30.4   8.90 33  15:21:31.4   8.90 34  15:21:32.4   48.10 35  15:21:33.4   13.10 36  15:21:34.4   11.80 37  15:21:35.4   10.80 38  15:21:36.4   10.00 39  15:21:37.4   9.20 40  15:21:38.4   9.30 41  15:21:39.4   8.90 42  15:21:40.4   8.90 43  15:21:41.4   8.90 44  15:21:42.4   8.90 45  15:21:43.4   8.90 46  15:21:44.4   12.10 47  15:21:45.4   10.90 48  15:21:46.4   10.00 49  15:21:47.4   9.30 50  15:21:48.4   9.30 51  15:21:49.4   8.90 52  15:21:50.4   8.90 53  15:21:51.4   8.90 54  15:21:52.4   48.20 55  15:21:53.4   8.90 56  15:21:54.5   12.40 57  15:21:55.5   11.10 58  15:21:56.5   9.20 59  15:21:57.5   9.30 60  15:21:58.5   9.20 61  15:21:59.5   9.40 62  15:22:00.5   8.90 63  15:22:01.5   14.30 64  15:22:02.5   14.70 65  15:22:03.5   14.10 66  15:22:04.6   12.80 67  15:22:05.6   11.70 68  15:22:06.6   9.20 69  15:22:07.6   9.40 70  15:22:08.6   9.20 71  15:22:09.6   9.30 72  15:22:10.6   8.90 73  15:22:11.6   31.70 74  15:22:12.6   47.80 75  15:22:13.6   30.80 76  15:22:14.6   57.00 77  15:22:15.6   54.70 78  15:22:16.6   55.30 79  15:22:17.6   54.00 80  15:22:18.6   53.50 81  15:22:19.6   53.20 82  15:22:20.6   53.20 83  15:22:21.6   61.20 84  15:22:22.6   60.60 85  15:22:23.6   60.00 86  15:22:24.6   59.20 87  15:22:25.6   57.90 88  15:22:26.6   56.50 89  15:22:27.6   54.90 90  15:22:28.7   66.70 91  15:22:29.7   64.50 92  15:22:30.7   63.30 93  15:22:31.7   35.30 94  15:22:32.7   12.30 95  15:22:33.7   13.50 96  15:22:34.7   14.80 97  15:22:35.7   14.70 98  15:22:36.7   14.10 99  15:22:37.7   13.50 100  15:22:38.7   11.40 101  15:22:39.7   9.40 102  15:22:40.7   9.60 103  15:22:41.7   9.50 104  15:22:42.7   12.00 105  15:22:43.7   13.20 106  15:22:44.7   14.10 107  15:22:45.7   14.70 108  15:22:46.7   14.80 109  15:22:47.7   13.10 110  15:22:48.7   11.80

HB

Hi Hb
Do not set the dsg relax time to 0, keep the default. Setting to 0 will not reduce the sleep current. Disabling the wake comparator might reduce the current draw by a little bit.

What exactly is the problem here? Are you unable to enter sleep mode?

thanks
Onyx

Hi Onyx,

Our task is trying to let BQ27220 keep lowest power consumption in sleep mode by using a battery. If we can keep lowest current, our battery can save power more time till user start to use our device.

1) BQ27220 will enter sleep mode automatically, so we don't need to set additional configurations for BQ27220 to enter sleep mode.

    Based on our current measurement result, I think we have already enter sleep mode, right?

2) What we measure the current value of BQ27220 is not static 9uA value, it is floating value due to wake-up. Range is around from 9uA to 60uA. Average current value is around 12uA in 60 seconds.

   Can we reduce more sleep current in BQ27220? or avoid large wake-up current.

3) BQ27220 don't have shutdown mode that I heard from TI FAE, can you confirm this? Because TRM still has shutdown mode flow, but we cannot find register to achieve shutdown mode. We can achieve shutdown mode in BQ27621 that we used before.

thanks

HB

Hi Hb,
1. The bq27220 will enter sleep mode as long as the sleep conditions are met. so yes if you see about 12uA current draw, then yes you are in sleep mode.
2. Yes you will not see static 9uA. The gauge wakes every 1s while in sleep but doesn't take any measurements. It will however take measurements once ever sleep current time and once ever sleep voltage time which i believe is 20s. This time varies based on which gauge. So you seeing 12uA is correct. The gauge has to wake to take measurements to maintain accuracy and when it wakes you will see those current spikes. There is no way to reduce the spikes. You can reduce the overall current consumption in sleep mode by increasing sleep time which is how often the gauge wakes to take measurements. You can double or triple that time from 20s to say 60s. You just have to make sure the guage isn't missing load currents during that increased interval.
3. The device does have the shut down mode, but the draw back is that you will loose all previously stored data on the gauge if you to shutdown. it will go to its original default setting . You will have to load your gmfs file and the learned FCC to the gauge to resume accurate gauging upon wake up.

I hope this helps.
thanks
Onyx

Hi Onyx,
it's more clear, but I have more questions:

1) >>"You can double or triple that time from 20s to say 60s"
Which data memory position can increase sleep time?
I check there is
-) 0x921E Sleep Voltage Time
-) 0x921F Sleep Current Time
But I am not sure I should set which position.

www.ti.com/.../sluubd4.pdf

2) Can you help confirm: based on your idea to increase sleep time, if we adjust sleep time to 60s, sleep current will have large value every 60s, but not in 20s, that can save overall average power consumption, right?

thanks

HB

Hi HB
Both of them. this is one of those gauges that the times were seperated. On our older gauges the sleep time was just one value. So you can play around with the times to see which provides an ideal result for you. Use bqstudio to make the modifications in data memory.
2. As long as your application is not drawing current or drawing very minimal current in sleep mode, you can set the time to about 60s. The reasoning being that your battery will not be exerted upon so the voltage will be relatively constant except due to self discharge, so there will be no need for very frequent voltage or current measurements.

you will need to still evaluate the accuracy performance of the gauge if you increase these sleep times to ensure that not too much error was introduced due to the less frequent measurements.

thanks
Onyx

Hi Onyx,
thank to your suggestions.
I will mark this thread as resolved when I finish final measurement.
plz wait ~

thanks.
HB