BlueNRG-1 completed the challenge (standby current finally dropped below 10uA)

gs001588

BlueNRG-1 completed the challenge (standby current finally dropped below 10uA) [Copy link]

  BlueNRG-1 completed the challenge (the standby current finally reached below 10uA). I worked until 4am and was almost beaten by my wife. Finally, I figured out the low power consumption of the accelerometer. I feel relieved! Send a broadcast every 10 seconds. The accelerometer only switches X, Y, and Z on a timed basis. All channels of the gyroscope are always off. The environmental sensor switches on a timed basis. 0x0C40 - 0x0BB4 = 0x8C = 140 (mV), which took 510 seconds. 0x0BB4 - 0x0B4E = 0x66 = 102mV, 500 seconds 0x0B4E - 0x0ADD = 0x71 = 113mV, 600 seconds. (One of the pictures was lost, I uploaded it again when I got home.) The experiment was not completed. Basically, the power was lost for 2mV in 10 seconds. The voltage before the power was tested was 1.439V, so it was calculated as 1.45V. If you calculate it this way, (3000mV - 1450mV) * (10s / 2mv) = 7750 seconds, about 129 minutes. I will do another complete test tonight to see how long it can last. There is room for optimization, and the average power consumption should be lower after optimization. The SPI port of the acceleration and gyro sensor (corresponding to DIO0 to DIO3) has been processed. The processing method is the same as that of cruelfox, DIO1 connects 51k to VBLUE, and DIO0,2,3 connects 51k to GND. DIO1 is the chip select signal of SPI. As long as the chip select is stable, the other signals should not have any problems. mV unit, voltage display 3000mV version test firmware: BLE_Beacon_gs001588_0201.bin (89.62 KB, downloads: 1)

2018-2-2 03:24 上传

点击文件名下载附件

V unit, voltage display 3.0V version test firmware: BLE_Beacon_gs001588_0201_3V.bin (89.63 KB, downloads: 0)

2018-2-2 03:24 上传

点击文件名下载附件

----------------------- 2018-02-01 Evening test data------------------------------------------- Tested tonight, something went wrong. Started the test, USB Dongle timing count was reset. There was a problem at 87 minutes. The time of the 525th frame is 0x004fff32, which is 5242674ms; the time after the 526th frame is obviously wrong. Because the program did not handle it, the timer overflowed by 5242879. The final freeze time is 7593 seconds, which is 126.5 minutes. The lowest voltage is 0x0599, which is 1433mV. The firmware has been modified: BLE_Beacon_gs00158820180202.bin (89.56 KB, downloads: 2)

2018-2-2 03:36 上传

点击文件名下载附件

BLE_Beacon_gs00158820180202V3.bin (89.56 KB, downloads: 2)

2018-2-2 03:36 上传

点击文件名下载附件

bkn1860

The host is mighty   

gs001588

bkn1860 posted on 2018-2-1 10:19 Awesome OP

Thanks for your support! I was just interested and worked till 4am

bkn1860

gs001588 posted on 2018-2-1 11:50 Thank you for your support! I was interested in it until 4 am

What tool did you use to measure the current?

littleshrimp

This version is good, 10 seconds per packet, average current 33uA

gs001588

littleshrimp posted on 2018-2-1 12:51 This version is good, one packet per 10 seconds, average current 33uA

Thanks to Brother Xia for helping me test: handshake The standby current should be able to be reduced a little bit, and the default state of all IOs has not been configured yet. From your current graph, it seems that there is not much room for the normal working current to be reduced. If there is, it may be that I added a delay between the recovery time of LSM6DS3 and the reception of the interrupt. Without the delay, the interrupt will be queried immediately and continuously, and the program will wait here. Now I forgot the frequency of the acceleration output, is it 104Hz or 1.66KHz?

gs001588

bkn1860 posted on 2018-2-1 12:30 What tool did you use to measure the current?

I directly connected the multimeter in series. My meter is pretty good among the poor price, VC980+.