Beagle board black industrial support full battery range

I am using Beagle broad black industrial grade for a battery operated application which has TPS65217 for battery charging and voltage generation.Battery pack of 3.6v /10Ah is connected to bat and batsense pin along with temperature sense. PMIC is reconfigured using I2C commands commands that need to be executed to set UVLO = 2.73V and LDO2 = LDO4 = 2.8V

I am facing unexpected issues after doing so.

Board is unexpectedly shut down after 3 days after working on solar(Vin) in day time and battery in no sun time. Now it is not getting switched on also. Board shows shorts at VDD_3V3A net at c15 capacitor.

Without proper protection from EMF, EMP and Static you more that likely blow it up. All you can do is review your design and apply best practices. Log your voltage with a logger that has a high dynamic range so you can see the transient responses in correlation to the sun and cloud movement.

10 Ah should be very adequate to wash out the lows, however that assumes the battery is at full capacity.
You have so many issues to consider, what i just stated is not even a drop in the bucket regarding what might have happened.

Check the wiring diagram. There is U4 (TL5209) on the board, that supplies VDD_3V3B rail. It’s powered directly with board input voltage. TL5209 has minimum voltage dropout of 0.5V. That means BBB requires Vbat of >= 3.8V.

Besides, there is VDD_3V3A to VDD_3V3B voltage leakage issue which was widely discussed long time ago.

As for PMIC configuration, check the Octavo guidelines

I would not waste my time with that, looks like they are a bunch of hucksters chumming for clicks. Went to that site and they wanted me to pay $5 to see the whole page, * them. If that is how they operate look for another source for your project.

Another thing is don’t believe much of what you find on the internet, much of it is useless. Might be better off designing your own powersupply for that. Chip makers will have a basic evaluation circuit so just start with that and add what you need.

Beagle board works fine with solar panel input on VDD lines. We have issue with solar +battery operation( which needs to be charged by TPS65217).
Stops working after some days but works in lab condition

Had followed the same document for PMIC configuration.
What is voltage leakage issue with VDD_3V3A and VDD_3V3B? there is nothing which can be changed for TLV5209.
It seems that working with full battery range(3.0V to 4.2V) for beagleboard is not feasible

All you can do is monitor your energies that are stored and produced. 10-12 years ago we did a feasibility study of solar power in our area. Not sure where all those solar companies got their data, what they stated is not even close to the truth.

If your storage and production capacity exceeds your steady state power demand, by several orders of magnitude the solar is fine. In other words, if you have a huge truck battery and a couple of good panels you would be fine. The other major short coming is the seasonal issues, summer is pretty strong and the rest of the year is weak to bad. If your device is located in the dessert you will have very good luck with solar.

It’s very old story. The problem of inter-rail leakage in SoC was discovered by users soon after the BBB release date. Back at that time, it was discussed on Google Groups with the beagleboard guys so they’re well aware of the defect. Eventually, in an attempt to get it more organized, the issue was opened on github in 2021. With zero effect, unfortunately. The new HW revision with the problem fixed was never released. Nobody cares. All the technical nitty-gritty and the proposed solutions can be found on github (see issues ##11 and 13).

Yes, the achievable capacity utilization ratio is quite poor. But it’s much worse to operate BBB without any battery, because of another problem. The SoC has multiple power rails. According to TI, on each power-up and power-down, the respective sequences must be obeyed otherwise the SoC’s reliability is not guaranteed. That’s why there is PMIC in the design. The problem is that this particular PMIC was designed for battery-backed applications. When no battery is present, the lockout can happen inside the PMIC during unexpected power down resulting in removal of all the power voltages simultaneously, violating the power-down sequence. The problem was acknowledged by TI and there are some workaround guidelines for the TPS65217 chips.

Anyway, BBB is not a new design so all the information is available. Just help yourself.