Yamaha YZ450F

[j bjork]

Im taking it a little easy with the start welding part. I want to make sure I dont have to change anything, and trying to decide how everything should be held together.
The cell holders are watercut.

Testing bus bars, everything seems fine, but some of the little tabs for the balance wires will be in the way.

There is a lot I havent decided yet, like how I will build the box. Where I should put the discharge connector, what connector to use.
What bms to use.

I would like something integrated in the vesc system, but I dont know what to go for.
There is the trampa bms:

VESC Based BMS - Balance charges up to 18s

Trampa Boards : Custom mountainboards, kiteboards and parts.

trampaboards.com

It is 18s, and I dont see anything about running two in series or something like that, but I know I have seen it done.
12A charging?!
2 for each battery wont be cheap.

ENNOID - XLITE-V4

XLITE-V4 is an updated version and almost a direct drop-in replacement for the XLITE, XLITE-V2 & XLITE-V3 Reminder: Monolitic BMS are not recommended for split packs. We recommend using Master-XLITE and slaves for split pack or using the XLITE-12+ with a 24S slave (see wiring diagram for XLITE-12+

www.ennoid.me

Ennoid XLITE-V4
There is a 32s version, but it says "beta, GH cable set not available yet"
So, beta probably means trouble, not sure what GH cable set means.
Charging 20A with alu cover, didnt see one available on the site.
That is ok for normal charging, but what if I want to charge fast sometimes?
35mA balancing? Lets hope the batteries wont need balancing then, it would take a while..

Maybe it would be possible to let the charge port control a relay for the charger, then charge trough the discharge connector?
If the bms sense something wrong the relay opens and the charging stops. But it will not sense the current going to the battery.
Wonder what will happen to the state of charge display?

I think there should be some more bms:es that work with vesc?

 

[rivvs]

Im taking it a little easy with the start welding part. I want to make sure I dont have to change anything, and trying to decide how everything should be held together.
The cell holders are watercut.

Testing bus bars, everything seems fine, but some of the little tabs for the balance wires will be in the way.

There is a lot I havent decided yet, like how I will build the box. Where I should put the discharge connector, what connector to use.
What bms to use.

I would like something integrated in the vesc system, but I dont know what to go for.
There is the trampa bms:

VESC Based BMS - Balance charges up to 18s

Trampa Boards : Custom mountainboards, kiteboards and parts.

trampaboards.com

It is 18s, and I dont see anything about running two in series or something like that, but I know I have seen it done.
12A charging?!
2 for each battery wont be cheap.

ENNOID - XLITE-V4

XLITE-V4 is an updated version and almost a direct drop-in replacement for the XLITE, XLITE-V2 & XLITE-V3 Reminder: Monolitic BMS are not recommended for split packs. We recommend using Master-XLITE and slaves for split pack or using the XLITE-12+ with a 24S slave (see wiring diagram for XLITE-12+

www.ennoid.me

Ennoid XLITE-V4
There is a 32s version, but it says "beta, GH cable set not available yet"
So, beta probably means trouble, not sure what GH cable set means.
Charging 20A with alu cover, didnt see one available on the site.
That is ok for normal charging, but what if I want to charge fast sometimes?
35mA balancing? Lets hope the batteries wont need balancing then, it would take a while..

Maybe it would be possible to let the charge port control a relay for the charger, then charge trough the discharge connector?
If the bms sense something wrong the relay opens and the charging stops. But it will not sense the current going to the battery.
Wonder what will happen to the state of charge display?

I think there should be some more bms:es that work with vesc?

I would suggest to stay away from Ennoid. So far I have had:
Master bms bricked during first setup
Master bms fried during precharge
Fried 2 slave board.
No I finally have a master and slave that works but it will not allow charging at all.
Using master xlite

 

[chuyskywalker]

Ennoid XLITE-V4

I have one of the V3's (I think) of these and it's...ok. It works just fine, but it doesn't add a ton of value over many other BMS's other than the big, primary advantage that you can bypass power going through a BMS set of mosfets/contactor/etc yet still properly react to parallel group LVC and other common BMS protections. If you have room for a regular BMS that can handle the amp load you want, I'd probably still recommend going that route.

Maybe it would be possible to let the charge port control a relay for the charger, then charge trough the discharge connector?

I wired up something like this for my BMS bypass. My controller (a spintend vesc) has an "Ignite" port (they call it) which, when you send battery + voltage to it, the controller turns on. Take way the B+, and the controller goes off. I use the P+ port from the BMS to activate a relay that allows B+ through to the ignite port. By doing this, I get exactly what you're interested in, but without needing a contactor. BMS detects any of the things wrong you might imagine, turns off, relay opens, ignite port goes dead, controller turns off. Roughly the same level of protection, very few moving parts.

You can see some of it in the wire mess of the second picture here: Hill Scooters' Scooter

 

[j bjork]

Thanks for the replays
rivvs, that sure is something that worries me. I suppose it is a different model, and maybe you had bad luck or something in the setup that contributed to the problems. I dont think I have seen more about these kind of problems, but then I havent seen that much more about ennoid at all actually.

chuyskwwalker, the main reason is that I want a bms that talks to the controller. I expect to be able to set different discharge rates at different battery temps and things like that. Also be able to log cell voltages and temps. Instead of cutting out if something starts to go wrong I will limit discharge. Not that it has been a problem so far, but it is always interesting with development

About the charge port, I think you missed my point. I already have what you describe on my other bike. The contactor is mainly because I prefer not to have 110V in a wet end muddy connector when I pull it out with sweaty hands, and as an anti spark when I plug it in again after I changed battery.

I know it will work to charge through the discharge port, to use the relay was just so the bms can turn off the charge if necessary.
The question is what will happen with the "state of charge" indicator? Will it be happy with the voltages, or will it show way off because it doesent record the current going to the battery?

 

[bunya]

With regards to the SOC, the ennoid LV Master uses the shunt on the negative to measure current. As such it is independent to whether the discharge or charge contactor is open or closed on the positive. If current goes in or out of the battery it is accurately calculated.
Not sure if this is the same as the X-LITE

 

[j bjork]

Thanks, that sounds good. I guess the best is just to ask them if it is the same. I thought it didnt measure discharge and left that to the controller. But then it needs reliable info from the controller, so it is probably better if it has its own shunt.

I have been working on the batteries:

I made some separators to mount between the cell holders.

Has done some spotwelding.

But this worries me. When I did the first weld on this cell i heard a little "hiss" sound. I guess it can be something else, but my first thought was that I punctured the cell. I cant see anything wrong, no moisture or anything. I also went back and did the second weld, I didnt notice anything strange or any sound then.

I have one more just like it in the other battery half I have done so far..
I am welding at 45J with k-weld

 

[harrisonpatm]

When I did the first weld on this cell i heard a little "hiss" sound

I can't see anything from the picture, so it's up to you if you want to disassemble to check. It is possible for a puncture to not leak. For what it's worth, I've found that puncturing the cell is more likely to happen on the negative end rather than the positive end.

 

[chuyskywalker]

chuyskwwalker, the main reason is that I want a bms that talks to the controller. I expect to be able to set different discharge rates at different battery temps and things like that.

You'll likely need to get into some of the custom VESC programming ecosystem for that, FYI. Last time I interacted with this setup, it was a very basic "at low voltage, roll down power till cutoff" type control and that was about it.

 

[A-DamW]

But this worries me. When I did the first weld on this cell i heard a little "hiss" sound. I guess it can be something else, but my first thought was that I punctured the cell.

Have you ever smelled Li-ion electrolyte? it has a VERY distinctive smell.

Try sniffing closely around the weld, you'll know.

 

[rivvs]

If you want to read more about the Ennoid you need to look on the many electric skate and one wheels forum. I have used the master lv and the master xlite. Also the xlite need to have amp going through it to stay on and keep the charge Port on. If there is no voltage/amp it will shut down the charging.

 

[j bjork]

I tried sniffing around the cells, but couldnt smell anything. I also tried a cell that has been punctured a long time after removing spotwelds, but couldnt really smell anything there either. I suppose it has to be when there is a gas or something getting out.
Anyway, I dripped on some thin glue. Hoping it would seal small leaks if there were any, then measured and noted all the cell group voltages.
Hopefully I can then see if there is a problem before I finish mounting the batteries.

I also finished the other 2 battery half's, but I lowered to 40J

I'm waiting a little with the + and - connections though, I dont know how I want them yet.
Probably just folded 90 degrees and with a few holes, but I can just as well wait a little.

 

[j bjork]

I ordered some plastic separators to put between the battery halves, but as usual it takes time..
In the meantime I have been looking at controller placement.
This is what I have come up with:

I think i can spare the space when I am going for relatively small batteries.
I think it is a good placement, with centered weight and relatively short wires to the battery and motor.
I hope cooling will be good there, but it will be turbulent so not that easy to say maybe.

 

[j bjork]

Not much have been done, but some parts have arrived at least:

Got some flat wire to use as balance wires.
I havent decided on what bms to use yet, so I might as well solder on wires on the bus bars for now.
Just so I can continue putting the batery halves together.

I also got this rubber foam double sided adhesive thing I planned to mount between the battery halves.

It is less rubber and more foam than I hoped, so I am a little hesitant
It is not very hard to rip off a peace.
The battery halves will be screwed together with spacers between, so it shouldent be that much strain on the foam I suppose.
And there isn't any sharp edges really, that could dig in to the foam.

But still, it would feel safer with something between the cells that would be difficult to penetrate with a screwdriver

 

[chuyskywalker]

But still, it would feel safer with something between the cells that would be difficult to penetrate with a screwdriver

I've used think HDPE before and really liked it. I also see PCB board stuff (FR4, I think it's called?) used a bunch as well. Thin plastic, generally, work pretty great.

 

[PK2000]

I'm using battery paper and some 0.5mm fibreglass sheet.

 

[bananu7]

I think the best material would be "epoxy fiberglass", in 0.5mm or similar. Its main use is a as a high-voltage insulator and is very puncture resistant precisely for this application. I put it under BMSes in two ebike batteries i built so far and plan to use as padding for the upcoming enduro battery in alu box.

 

[j bjork]

I decided to go for a layer of fish paper and the foam rubber:

I did this a few days ago, and today I got a mail from the cutter saying that my plastic sheets are finished

I have started shaping the battery box, I think I will make it in fiberglass.
But I really have to make some decisions very soon.

I cant decide what bms to run.. Should I go for something vesc compatible?
Maybe run 2 cheaper bms:es that control 15s each?
Go for a ant bms, or maybe a daly?

Then there is the connector thing. Andersson power pole or something else?
Maybe I should buy or make something that connects when I slide in the battery, so I dont have to connect anything?
I would be interested if someone have some input on that

 

[bananu7]

I have just done some research on the battery boxes, as if my build materializes, it will be the biggest thing I'll have to finish.

For material, so far i still think that 2mm aluminium is a the best choice in my area, but carbon fiber or glass fiber laminates are very good alternatives. I have built my most recent ebike battery from PC, and while it has excellent properties all across the board, I'm not sure if it will fit the bill for an enduro motorcycle. A thin shell with a secondary PC deflector layer (I'm thinking loose pebbles, branches, rock edges) could work quite well too. The idea being that the springy PC can absorb the impacts without breaking, is easy to replace being on the outside, is cheap and able to be formed into a nice flowing shape, while the internal rigid box keeps the battery solidly mounted, protecting the cells mostly from humidity, dirt and moving around.

For connectors, I saw quite an interesting offer on Aliexpress for quick detach 500A connectors:


Link: Link

They are about 60 eur for a pair, supposedly waterproof. At the same time i keep thinking that the controller still has regular, exposed terminals... so it's not like it's very easy to make the main connection waterproof, so maybe just mirroring what the controller has is a cheaper, more reliable option?

 

[chuyskywalker]

I cant decide what bms to run.. Should I go for something vesc compatible?
Maybe run 2 cheaper bms:es that control 15s each?
Go for a ant bms, or maybe a daly?

You can always kick the can down the road a bit if you make the BMS an "external component". A route I've taken before in which I just make up the whole battery, grab all the leads out, and feed them out of the case in a bundle. That way the BMS is external and instead of most setups where you just run a pos/neg off the "battery", this would have the same pos/neg, but also a bms bundle wire. This kind of DB25 works out really nicely because it can carry 21 leads for the balance wires + 4 leads for temp probes (so, 2 probes).

Edit: see this build I did: Fiido Q1S Upgrade; 200A, VESC, Copper-Nickel 52v Molicel

 

[harrisonpatm]

You can always kick the can down the road a bit if you make the BMS an "external component". A route I've taken before in which I just make up the whole battery, grab all the leads out, and feed them out of the case in a bundle. That way the BMS is external and instead of most setups where you just run a pos/neg off the "battery", this would have the same pos/neg, but also a bms bundle wire. This kind of DB25 works out really nicely because it can carry 21 leads for the balance wires + 4 leads for temp probes (so, 2 probes).

This is the way. If there's a problem with the BMS, its exposed and you can easily deal with it. If you want to access the balance wires, they're more easily accessible.

 

[j bjork]

It has to be simple and quick to connect/disconnect, that is why I consider someting that slides in when I mount the battery.
The anderssons take a little wiggling when connecting or disconnecting, but they dont get stuck from mud or sand.
I think the one linked by bananu7 will get stuck and hard to get off when it is dirty, and it would take 2.
So at least 2 connectors to fiddle with every time I´m changing battery.

I have looked a little at another, dual pole connector that looks more interesting (used on electric cars).
The locking mechanism can be removed on that, but I think there is still too many surfaces that could get stuck in sand or mud.

I dont think external bms sounds very interesting, it could be possible to just use it for charging. But it would take a massive connector and I am not sure how the bms would handle being pulled out and plugged in all the time

I do consider having bms displays on the chargers, and using qs10 connector or something with the 4 extra, small pins for a charge port. Then I can at least see the cell voltages when charging, if I dont have a bms that communicates with vesc.
(And have a vesc compatible display on the bike)

 

[bananu7]

Getting dirt in the connectors isn't something that has crossed my mind; my assumption was that they would be covered by something and then the waterproofing on them would take care of the very last bits of moisture. That being said, i like that they have a locking feature. I have so many devices with sliding batteries where that slide contact is a failure point and they require an occasional knock or padding to keep the contacts running. Now imagine the same but with a couple hundred A through it and the sparking from vibrations. If you want to have the battery be contacting on slide in, you probably should make the entire box have guides that make it stay solid with regard to the other part on the bike (or at the very least have the connector have some play from the box so that it can self-align).

Incidentally, I have a design that I have been working on for a licence plate holder for enduro bikes that would connect on slide:



I have a couple of connectors already, but while they seem fine for turn signals or perhaps a small ebike battery, I am not sure I would trust them for the power you're likely to be putting through them. Oh, and on ebikes, that connection is the place where the water always seems to accumulate...

As far as external BMSes go, I wouldn't go that route personally. Setting aside that it would be virtually impossible to pass UL or any sort of cert without a BMS solidly inside, it just doesn't seem to make much sense to add extra complexity of getting extra couple dozen wires out of the battery. If you add a connector, that's yet another failure point. And those wires typically don't get any protection on them, which means that a short on any of them turns them into nice heaters for the cells inside - ask me how I know.

Instead of assuming that a battery without a BMS will never fail, I'd plan for periodic inspection of the battery with a removable lid. That way a faulty BMS can be replaced if/when it decides to fail, but the battery is always protected, doesn't require any special consideration when charging and exposes minimum necessary to the outside world via just the output terminals. The only thing I'm not sure about is whether the positive side fuse should be inside or outside of the battery. Putting it outside technically exposes a circuit protected only by the BMS, which can be a bit slow to react in case of a dead short, but putting it inside requires opening the case to replace it.