Unisual ballancing behavior

[Sasa Duric]

I have noticed this behavior before, and now again.. As it can be seen in the screenshot, my cells are well ballanced and there is no solar production, but delta jumps to 180mV and it looks sbms is ballancing even it should not. 

 

Home Assistant history, shows ballancing was rightfully happen when battery was charging around noon, but late afternoon as well. That strange ballancing is active for quite a while, and than ot stops.

Acoording to sbms settings ballancing should be active just when charging, but even if that is not the case, few mV is not a reason for ballancing. What is happening here? 

[Dacian Todea (electrodacus)]

Sasa,

I know about that issue. It is not cell balancing but occasional reading error. You can see that is 60mV, 120mV and it seems you even get occasional 180mV  This happens when cells are at 3.300V as there are a lot of zeros in the data. I will try to fix that in the next firmware if I get the time. It is just occasional error display reading it does not affect any other functionality. The ISL94203 has the correct values and sends the correct values but the microcontroler reads those wrong on occasion over the I2C interface.  

I knew about this for a few years and just started to annoy me recently when I started to data logging everything :)

[Sasa Duric]

Thanks for the clarification Dacian. Yeah, it is annoying. I log bunch of stuff myself in HA. I hope we'll have that corrected soon. And, do you plan to do diversion metrics too? it looks gimmicky to have the screen without actual support for it. Is there OTA update possible or we need to use a programmer?

[Dacian Todea (electrodacus)]

Yes the diversion metrics is the first on the list to be implemented.

It is implemented in the external datalog application and it will be implemented the same way in the SBMS0 so need the user to provide the resistance of the hearing elements to calculate current and power from that.

OTA is only possible for the ESP32 firmware. 

The main microcontroller requires a cheap ST-Link V2 programmer. I use one of this $3 USB stick type ST-Link and 3 wires are needed SWD , SWK and GND and the SBMS0 needs to be powered and the BOOT0 (pin5) connected to 3V (pin 11) before powering the SBMS0 so it enters programing mode.  

 

[sailingharry]

The special USA price is double that.  Still cheap.  But before I run out and buy one, a few questions:

*  When do you anticipate the next software update?

*  What are some high priority changes you are planning?  I don't use diversion, so won't bother to upgrade if that's what's done, and probably won't bother with the logging issue either.

But if you add a low temp charge cutoff (type 1) instead of a low temp disconnect (type 5), I'd be first in line for an update!

Harry

[Dacian Todea (electrodacus)]

Harry,

The price should not be anywhere close to double at least not as of now. The SBMS0 is made in Canada so still under the free trade agreement.

Not sure exactly likely in about 2 to 3 months there will be a new software.

The main update is related to diversion monitoring and fixing the issue described above. I will add support for SBMS2 but that will not be available this year as I need to build a prototype and requires a lot of testing before I start to produce that. And main change to that is adding a second PV shunt so it will have PV1 and PV2 measurement as the older SBMS40, SBMS120

I could add a separate temperature limit that only disconnects charging but you will never want to get there and eventually the load will also be disconnected as it is unlikely for the battery to get above set limit once temperature is to low. Best option is to have the battery in an insulated and heated enclosure.