LEDs about pibat

[Andyzug]

Hi! LEDs are always ON when it's in use

[KlaMa]

Hi Andy,
this is a super cool and usefull HAT. Many thanks. I want to use it for a tele-medicine project. My other hat, an audio grabber hasn't arrived yet and I hope that your power supply will be strong enough to power the hat too.

Anyway, I was wondering if there is any way to change the state of the LEDs through an I2C bus command, e.g. make them flash, turn on only one LED, etc. to have a visual feedback on the state of operation of my device. In particular, I would like to somehow highlight if there is a problem, e.g., no connection with the wifi, no ssh tunnel forming, etc. Utilizing your LEDs to warn about such states would be a great help because it would not require adding any more components to the device.

I have another question: I noticed that the battery and lights remain on after I sent the device a "sudo shutdown -h" command. Is there a way to turn off the battery other than pushing the switch or draining the device of power? Even if it means that pi zero is cut from electricity? I was thinking of an I2C command...

Best regards,
Klaas

[Andyzug]

Hi Klaas,
Sorry for replying so late, I didn't receive any notifications from github.
How is your project going? Is the PiBat doing its job?
Unfortunately I2C is only on ADC (analog to digital converter) and it only reads the battery voltage, it doesn't accept any commands.

[KlaMa]

Hi Andy,
My project is coming along very well. With a small modification, the PiBat is capable of powering the PiZero W with header pins soldered on and an audio grabber HAT for almost 5h 50 min. The device was continuously connected to a WIFI network and sending audio data to a remote server. The Voltage was at 3.4 V, and the power level was at 11.2 % before the device became unresponsive.

• I initially encountered a problem after the battery used for about 40 to 60 minutes (Voltage= 4.0 V, Level= 97%), even when using the PiZero without the audio grabber HAT my project utilizes. The zero kept repeating to reboot without ever becoming responsive again, also when forcing a restart; only attaching an external power supply and a full recharge could resolve this matter. The problem was reproducible without and with the second HAT attached. I managed to overcome this problem by connecting a capacitor between 5V and Ground pins of the PiZero. For testing purposes, I chose a 10,000 µF capacitor, which is physically quite large and would not easily fit into the intended enclosure. I will try smaller ones when they become available to me, which I intend to solder to the PiBat directly. Before adding the capacitor, the device was making an audible high-frequency noise. After adding the capacitor, the sound disappeared, both while running on PiBat and while recharging. This sound does not refer to the signals coming from the audio grabber HAT, which captured high-quality sound under all conditions tested.

• I gave up this idea of a software power-down since the PiBat power supply cannot be switched off by an I2C command. As an alternative, I will use PiBat's switch for shutoff, which appears to be safe as long as a read-only file system shadowed in RAM is used.

• For signaling the state of operation, I will attach an LED to the device's GPIO ports since the PiBat LEDs are not accessible through the I2C bus.

• I will consider using PiBat for other projects that need external power or an uninterrupted power supply (UPS). As mentioned here, using a large enough capacitor will prevent the device from rebooting after the (charging) power supply is disconnected. I did observe uninterrupted operation as well, when I used the 10,000 uF capacitor in my project. This would allow protecting RPIs that need to be continuously powered to enable uninterrupted data collection or prevent file- corruption, etc., at a very competitive price. I only wished for enclosures that could accommodate the PiBat-mounted RPis better, as well as a PiBatversion that could be connected to RPis other than Zero, while still allowing for I2C communication. In the meantime, the soldering iron will probably help to do the job.

• Finally, it would be helpful if one could determine by an I2C command if the RPi is powered through the external power source or the PiBat. The current script can determine the voltage and power level, but it takes about 1h to change from 100% to 97.5%; in the time between it is not possible to determine if the device is powered internally or externally.

Thanks a lot for your feedback and this great little power source.
Klaas

Hi Klaas,
Sorry for replying so late, I didn't receive any notifications from github.
How is your project going? Is the PiBat doing its job?
Unfortunately I2C is only on ADC (analog to digital converter) and it only reads the battery voltage, it doesn't accept any commands.