Im curious about your charging current. some people have told me that anything over 4 amps charging will roast the bms.
1 Like
I have been using/testing a 16A TC/Elcon charger for charging my 20S packs with an XLITE-24. Heating is there but nothing to damage the board. BMS is able to manage the end of the charging cycle with the throttling functionnality over CAN bus. BMS can ask the charger to throttle down once the battery is almost full. By the way, end of cycle charging normally becomes a problem without throttling with a high current charger with a relatively small battery pack. The BMS is not good at it without CAN bus. So better slowly charging unless you have an elcon/TC charger with CAN bus.
There are two possibilities if the BMS generate a lot of heat during charging:
-
Either the board is defective/badly soldered at some point so that the charge FETs on the charger side does not engage and the charge current circulate in the FET internal body diode. 4 to 5A becomes the maximum in this scenario.
-
There was a time last year that I had to use higher RDson mosfets (not as good as the normal ones) due to the part shortage so that the board charging capability was limited below 10A
2 Likes
Thanks for this reply. I have your xlite 24s bms and will be using it on a 14s4p pack and a 8A charger and Im just really nervous about breaking stuff bc thats usually how things go for me.
So i extended my setup now with 2 x Ennoid Vesc and xlite24 BMS all connected via CAN Bus. In my test setup everything seems to work. All devices are recognized in the Vesc Tool. Still waiting for my new Board with the limited space for Vesc so I can do a real live test.
6 Likes
Hey @ENNOID, I have a little issue. I canât update the firmware from 5.2 to 5.3 because the HW-Name got changed from one tool to the other.
How can I solve this?
Manual upload with .bin file which is available on the website on XLITE page in the bottom table.
I made both XLITE-12 and XLITE-24 firmware separate now to improve user experienceâŚat least from 5.3 and above. A few parameters are set different by default and some users were confused after a firmware update
OK, thanks.
Iâve been looking on GitHub for the compiled firmware but didnât think of checking the website.
@ENNOID bug report:
with firmware v5.4 my 15S-SS starts always in charging operation because the charge control FETs switch on for some reason causing pack voltage to appear on the charge port and getting falsely detected as âcharger connectedâ
It also loops on and off in that state.
Turn on, detect voltage on charger input, go into charging operation, stay there for a bit, turn off, turn onâŚ
I downgraded to v5.3, no issues there.
2 Likes
Weird, I will have a look at that issue this week and will come back later. Thank you for reporting it.
2 Likes
I did some test with EBMS-tool 5.04 today.
I found what the issue wasâŚA few extra line of codes were not commented as it was supposed.
So, Iâm now updating the github release section with the new 5.04 firmwares.
Sorry for the trouble. Hope it works fine for all ENNOID hardware now.
2 Likes
9 Likes
Will this HW version have fixed the issue where the cells on separate balancing ICâs donât balance to each other, only within the group attached to the same IC?
1 Like
Is the V2 going to have slightly longer mounting holes which is going to mess up existing designs?
Yes V2 is a complete PCB redesign, so the mounting holes wonât fit with the previous V1 Sorry, but I tried to keep them identical in size, but could not. Also, the balance connectors order are different so it is a complete new board. Iâm still planning to keep producing V1 boardsâŚ
âŚbut about separate IC that do not balance to each other, well it does in fact, but it is more related to your settings in EBMS-tool, there is the âcell management->Balancing->cell balance difference thresholdâ and also the âcell management->SOC->mismatch thresholdâ that must be both manually set below 10mV if you want them to balance more evenly.
The small mismatch in cell voltage reading are caused by three things:
-
The balance cable length/size, in fact balance cable resistance has a small effect on reading. The second balance IC is more sensible to balance cable resistance.
-
The resistance of the internal traces on the board also create some offset between channels. I tried to lower the trace resistance on V2 by making them wider and shorter, so it has less effect on readings.
-
Iâm using LTC6804 balance IC due to the part shortage and they are not as good (precision wise) as the original LTC6811 that I was using previously. Traces and balance cables resistance affect more the measurement accuracy with the LTC6804 IC
4 Likes
can I offer you money to open source v2?
to pay for R&D and future support
You can buy the product to contribute to the R&D and future support.
Iâm now doing this for a living:
- Hardware design for other companies
- Selling VESC & BMS on my website
Making the hardware completely open source is just a free pass for others to copy and kill future R&D+ support.
5 Likes
Hereâs a few comments regarding v2 applications to OneWheel (which this is like the only smart bms that will fit inside the battery box currently):
- âPower button connector for waking up & resetting the BMS with an external ON/OFF button switchâ
This is how the factory BMS works in the OneWheel. i.e. Using a momentary button to bridge ground to a pin that wakes up the BMS which supplies power to the controller. Not sure how useful that is being charge only though.
- âOutput switch connector for controlling an external shield board. The XLITE can now be scaled to a full BMS with external add-on || shield board.â
Interesting. Any specs on space dimensions for the discharge shield board addon? 
- âNow with buzzer, built-in humidity & temperature sensorâ
It would be handy if we could also control the buzzer from the VESC via CAN message to relay other error messages from the VESC.
1 Like
too bad iâm not software inclined to actually use any of these features,
The new features for V2 are for opening up the XLITE beyond a single use case âcharge BMSâ thus making it more flexible.
It is true that you donât really need a power button for a charge only BMS, but it becomes necessary if you plan to use it as a discharge BMS when using a discharge shield. The extra connectors are all located on the opposite side of the BMS and can be left unpopulated if not needed.
The I2C connector can be used for having an I2C display showing most important datas and having the power button allow to shuffle between the different displayed set of datas. This is an important feature that was missing to XLITE IMO. I really missed that display present on my other discharge BMS.
The discharge shield boards are yet to be defined. Any shape or power is possible for future design.
For the buzzer, it is something that can be done eventually.
1 Like
