One BMS to rule them all

BYqSXt8Z

100 W
Joined
Jul 3, 2014
Messages
126
Location
GTA, Ontario, Canada
Hello everyone, my buddy owns FTEX and he's agreed to work with me on a versatile BMS (4s-14s, ~300A) that will be fully programmable (cutoffs, chemistry, etc.). The goal is to have a made in North-America option in the high-end spec range, at a mid-to-high end price (say $200-$300). We have prototypes we have to flash via canbus at the moment so it's more than a pipe-dream, but we are still working out kinks, so now is the time to make sure we build it right.

What features do you think we should add?

I've already collected these from FB in the last 30 minutes, and I love them!
-use soldering holes / terminal options for the balance instead of JST
-12v regulated output for active cooling options and small accessory loads
-control over what the temp sensor actions do at different thersholds (cutoff, throttle, prevent charge but allow discharge, change charge profile limits, etc.)
-soft start option
-wifi (better for large applications like RV or a house)

Thanks for the input :)
*edit, adding the best FB suggestions as they come*
 
Excellent point, I did not know marine was 60v nominal.

I think the 14s is just because that's where our reference design started (the OEM BMS he produces for some ebike companies), I'll see what we can do about increasing. I'd like to see it go up to a higher s count for sure.
 
Hillhater said:
Remembering that marine packs are often LiFePo,..so 60v would mean 18-20s ?
I assume Bluetooth as well as WiFi.. ?
Data logging ?

Yes, ideally we will be able to push the same design up to 24s, but FETs operate best in certain ranges, so it might have to be split into two products, one for higher voltages, one for lower voltages.

Bluetooth was the next addition, but now I'm wondering if WiFi should be the priority. Either way, that's the plan!
 
Ok.

Copy this. Fer less than 146$. Go for 32s. If there is a hobby market. If the market is 96s, coolio.. but isnt a distributed BMS safer there?

https://www.ebay.com/itm/114978507669?hash=item1ac540cb95%3Ag%3Ad5EAAOSwcGNhORer&LH_All=1

This BMS will easy peak 600A. Who want a metal oxide semiconductor field effect transistor bank in the way. This shunt is .. notalotta moH. Much more reliability than a semiconductor package mounted on the system. A chunk of copper. 300A contin flat Panasonic relay for the DC breaking. .... 3 temps, and the ( relativly stoopid) programming...

When I install them I take the header off and install a Cannon.

Can bus. Who or what do you plan with a CAn implementation?

How much does any hobbyist use CAN? It is for comms with the charger, right? Which charging system?

I got a friend named Julian who is working on the same thing.

Oh yeah gold signal pins.. DB headers to take it.. Screw downs and really.... easily.. weather prooffed / idiot proofed.

Bluetooth, Uart or Serial, and wifi. Why not. All three, easy to reach and swap.

Maybe balance line fusing, if its not distributed. Quality harnesses. Like some gold pins or something.. that will sell.. For sure . I a m pretty sure any money you spend to get away from the "junk design " of the style, getting pumped out of Shanghai and Shenzen. Real harnesses with positive interlocking weatherproofed terminals,.. and a good supply where the DIY user can get them.. Not that clone JST junk.... 24g wire junk headers silly stupid amounts of solder on the wave flowwed mass produced and listed 10000 times on eBay junk.

I would pay 300$ for a BMS like that, out the box. ^^^
 
DogDipstick said:
Ok.

Copy this. Fer less than 146$. Go for 32s. If there is a hobby market. If the market is 96s, coolio.. but isnt a distributed BMS safer there?

Yeah this won't go up to the high voltages. The reference design is from FTEX who made this BMS for an e-bike company in montreal because they were tired of the DOA BMS rate they had with their import batteries. From there I saw a potential for something more versatile if we had access to the firmware through an interface (bluetooth / wifi) because of the TI IC they use at its core. It really opens up the opportunity to tweak it instead of needing various SKUs.

The CAN is actually to communicate with their controllers, they use it to have adaptive regen on ebikes, regen power varies based on battery state of charge to avoid damage and to maximize usable energy. I'm not sure how it would be adapted from there, but if we make the software open sourced (I'll really push for that), having the bus there opens up to mods for better functionality with solar installations.

Sending you a DM as well, I have more questions
 
DogDipstick said:
Maybe balance line fusing, if its not distributed. Quality harnesses. Like some gold pins or something.. that will sell.. For sure . I a m pretty sure any money you spend to get away from the "junk design " of the style, getting pumped out of Shanghai and Shenzen. Real harnesses with positive interlocking weatherproofed terminals,.. and a good supply where the DIY user can get them.. Not that clone JST junk.... 24g wire junk headers silly stupid amounts of solder on the wave flowwed mass produced and listed 10000 times on eBay junk.

Please elaborate on balance fusing!

Yeah, harness is not going to be JST... we need something reliable. I am thinking if we leave the connector port unpopulated, then people can either solder on the port, or solder quality silicone leads directly from the board if they prefer that. Not sure here yet. If you can suggest a good connector format we can piggyback on, I'm all ears! Depending on board layout, we could even have a few unpopulated formats available, so that based on use case the user can use the better option.
 
BYqSXt8Z said:
-use soldering holes / terminal options for the balance instead of JST

I would say don't balance at all unless you can do it actively. Maybe just provide a port for an external balancer to plug into.
 
Back
Top