What’s the practical reason behind this?
If it’s done charging and balancing, the connected charger doesn’t really do anything…
Tipsy is somehow right. Forgetting the charger plugged-in overnight is not a big deal for the XLITE and the pack in general. It is more of a concern with some dumb BMS that bleed the pack indefinitely if the charger is left plugged-in.
I honestly think that low voltage alarm might be more convenient because some VESC don’t have auto-shutdown and require a manual shutdown by the user. Some user reported forgetting to turn OFF their VESC and drained their battery pack due to this. The buzzer will make it worse for sure, but on the other hand, if the buzzer is heard on time it might save the pack. Some people are even adding a buzzer to their VESC for that reason, but the XLITE already has one so…
1 Like
Some chargers, like mine, will turn off voltage if they don’t see an amp draw after a few minutes. I need to power cycle the charger if i need to use after it has hit this cutoff - if there was an alarm on the xlite for a connected voltage it wouldn’t go off for long 
I’m using an Ennoid LV BMS and encountering an issue during the initial test connections phase in the terminal. It’s indicating that my 11th cell has failed. My setup includes two 12s slave modules and a 20s battery pack, which I’ve divided into two sets of 10s each. Despite being able to read the voltage of all 20 cells, I’m puzzled by what the “starting test connections” entail and why it’s flagging the 11th cell as problematic.
Check your balance connection around c11. It is probably floating. Otherwise a blown-out balance resistor around c11 will still read cell voltage properly but the balance function wont be possible, so it tells you.
Hi.Can you share the design for version ss-15s? I want to repair my bms.
PM me for a repair
@ENNOID Couple questions about the v3 xlite. I added a 120mm by 100mm by 15mm heatsink to the v3 via the bottom of the board secured with screws and good quality thermal pads on the underside in .5mm and .75mm to ensure it has good contact with all the components. Even in free air im hitting thermal limits on the xlite at about 1000w (13a at 18s) and max of about 9.5a or 700w with the heatsink mounted and enclosure secured.
Id like to improve the performance so that i can use increase the charge current to 14a or aprox 1200w while keeping the cooling passive but it seems like the heatsink isn’t saturated. I’ve adjusted the thermal contact twice to very small improvements. How would i go about getting this thing closer to the 20a spec? What methods did you intend to keep from cooking this baby
Second, i have this integrated via can with the vesc and occasionally it will show a ghost charge current in the mAmp range which isn’t an issue except that in the integration with megan it switches the display to the charging page unnecessarily. Is there a threshold setting to adjust to prevent this? I did a little digging around but nothing seemed to standout or I overlooked it. Not a big deal just wanted to tweak it so it’s a little more refined
2 Likes
Swapping with mosfets with a lower rdson is one option. Increasing the default 65C thermal limit is also an option.
The mosfets of the XLITE-V3 were 150V rated 7.5mOhm, but most people have a 18S or 20S pack so most good quality mosfet rated 100V in a TDSON-8 package would work. 5mOhm rdson or lower. V4 use 120V mosfet with lower rdson for that reason
I would ask the developer for the megan display to put a minimal current trigger value. The XLITE has a setting for minimal current trigger, but it is used internally. The current reported is the current measured.
1 Like
Hi. are you thinking of using the KA49522A from Nuvoton instead of ltc6811. They have built-in current monitoring and a mosfet driver in one chip.
Garbage chip compared to LTC68XX and others similar IC, either from NXP, TI, Infineon etc.
There are many similar IC to the KA49522A on the market. Problem with them is that they can’t scale. Those IC are good for creating a cheap BMS for a very specific product, but the trend is always to go higher than just 22S
I also prefer to keep the design as flexible as possible. The LTC68XX are quite good at this. You can have isolation between slave and use as much as you want without to worry about timing, delays, limitations etc.