Recommend smart BMS under 100.00 for 14s Tesla 2170 cells...

rollin76

1 µW
Joined
Apr 12, 2011
Messages
3
As the title says I am looking for a decent reliable smart BMS for a 14S 5P battery using Tesla model 3 2170 cells...
There is so many to choose from I don't know where to start. Looking for something that has adjustable high and low voltage cut off that allows me to stay away from the normal min max charge and dis charge voltages to prolong the life of the battery..like charge to 4V max and discharge to 3.5v min...
I guess I would also need a charger as well....been looking on ebay mostly but there are so many to choose from...
Any recommendations??? Any help would be greatly appreciated......
 
Forgot to mention 40a discharge and capable of balancing...this will be my first try at building my own battery...I don't need anything fancy...just something reliable without a screen programmable on my computer..no phone apps..
Was looking at this one...but I honestly don't have a clue if it is any good...please help...I can't afford to find out the hard way...
I figure if anyone would have some sort of advice it would be the people on this forum...
https://www.ebay.com/itm/14S-51V-60...-BMS-UART-RS485/113831902309?var=414053333234
 
The most Obvious answer would be to use the Tesla BMS :mrgreen:

10 years ago I could have told you how to communicate with the slave modules.
(I was under NDA forever - so - technically I could not - but I once held the knowledge and I have a bear trap of a mind)

At one point I wrote up the comms algorithms to exercise every interface on the BMS master - one of which - would be the slave modules.

I dont remember if it was CAN or isolated SPI
A quick look at the part numbers on a module would determine that
Likely CAN eh? I dont think ISOSPI even existed then.

The Protocol I remember
At least the nuts and bolts of it

You only need to go thru as many permutations as are found in any Programming language (to represent numbers) then multiply by constants until you hit a known stimulus match. There are 50 or so messages, or on the order of that. The constants would likely be weird numbers like 0.023671234

Bit
Byte
Int
4 byte int
8 byte int
Float
Double
Double Double
.....

I mean...
You can only represent a number so many ways...

and
There is only Little and Big Endian

and
I will give you one hint for free

"No Nibbles, no Bits"

I will give you a second hint

420

... from that you can solve it and I am amazed that few have.

...
...
...
of course you have to understand CRC's and Correction & Detection.

...
That is more than enough to solve it.
Yes - all the bits must be correct
Yes - it should accept a single bit error but I think that was taken out.

-methods
 
Standard CAN protocol

COB ID and all that
reading up on CAN OPEN will allow you to match the patterns you see

Look at the uControllers used and the year the part was introduced.
Look at the programming languages used then
Look at the capabilities of the chips (original data sheets) to get the valid variables

I do not believe there was any funny business outside of the built in functionality
(no hand calculations of over-sized variables)

Almost all of the messages are fixed length. IIRC there is only one that is variable.

...

It is retarded that a decade later people are still stripping out the most reliable BMS on the planet*

Lawyers :kff:

-methods
 
In other words OP, the functionality you seek seems to not be available in an OTS unit.

And if someone with the necessary skillz devoted the hundreds of hours required to build one, odds are the result would not be cheap.

Rather than looking for "a BMS", break down the functionality you need, and then look for the various components required to implement each.

Note just the appropriate contactors will likely cost more than $100.
 
He is on to it.

...

To try and provide a hack solution:
Since OP is only building a 14S pack he could grab any of the standard EBIKE BMS's.
If he needs to drive more current than the mosfets are rated for he could use the mosfets to control contactors.
Lots of hacky things that can be done.

Anything cheap will likely not have a programmable LVC
Anything cheap will likely have a shit balancing that only happens at HVC
Anything cheap will be just what it is

... You can always strap on some Hobby Beepers :mrgreen:

If you want to spend $200 or $300 there are plenty of programmable solutions out there. I am not aware of anything in the $100 range.

... But I have not looked in many years. I only commented because we SHOULD be able to leverage the Tesla BMS. Maybe I will stalk Elon and ask permission.

-methods
 
Question if one was to utilize a bms on a pack which would you recommend? Looking to build a pack for my Luna X1 ludi thats bigger than the stock.
 
methods said:
If he needs to drive more current than the mosfets are rated for he could use the mosfets to control contactors.
I would point out that many good DC breakers have remote trip options. Take a voltage-only BMS, connect the fault output to the breaker's remote trip, and you have your BMS. Breaker trips on overcurrent, and the BMS trips the breaker on overcharge or undercharge. You also need to reset it manually, which is a good thing IMO.
 
chargery bms16 is 110 dollars, using contactors, it'll handle whatever amps your contactor handles. It has lcd screen, fully programmable from the unit itself,1.2 amp balancing. You have to supply your own contactors(2 of them one for low voltage shutoff, and one for high voltage shutoff), that be an extra cost. I been using the chargery bms8 for the past year and it has been the best bms I've used so far.

Its not a small bms, its 2 units the control box that goes near the battery, and the lcd box where you program it and view battery status. The 2 units are connected by a data cable. Then you have a shunt that measures the amps flowing. For expensive lithium cells this is what I recommend. Easy to program with out a computer or Bluetooth, takes seconds.

 
Contactors
I got off on the DC Breaker kick for a while, but at the time, I think they could not do the Voltage I was after.
IIRC they are suitable for Ebike use.
I try to check in on those every few years as Solar is driving their use case up (I believe).

$40 new for a plenty good contactor from Gigavac rated for 1,200V and 50A continuous... but of course the best place to look for contactors is in the dumpster. :D Ebay, Rejects from the production line, ... Here I look for damaged cables. There are at least 3 reasons why a contactor ends up as scrap:

1) Damaged cable or something we do not care about (Production Line)
2) Blown Open under full load (As a part of some safety system that is regulated)
3) Miss-Match, so incorrect coil voltage, not built to spec, etc (so brand new and perfectly good)

Zero Motorcycles uses a miniature TE contactor. I dont know the coil voltage. 12V likely.
Any company building DC Fast Charge would use the full size Contactors, gas filled, likely in TE or Gigavac.

A mostfet pack can be used (as you see in most Ebike BMS) but special care is to be taken to protect the Gate to Source potential, and the Drain to Source should be protected as well. ...

A standard SSR will not work as they are based on Triac and require a Zero Crossing to open.

Of course you can just drive an OptoCoupler to crow-bar your Throttle to 0V and no doubt, just turn off the switch-power to your Controller. So... for that case... you need much smaller relay or mosfet that could probably fit into the controller body.

... For anything low power I think we all know that anything flies. When we get up to high power we are more strict about enforcing System Interfaces, so in that case, the battery really should have its integrated LVC protection that actually breaks the path of current.

BMS's
Since there is an Arduino module available to act as Master, and it can poll OEM slaves and ignore what it likes, there is now no reason (I suspect) that one could not re-purpose the OEM BMS's. On the Linear that means jumpering the unused channels. On others I suspect it will be the same as you do not want to leave them floating.

In that case - Scrap BMS + Scrap Arduino + Scrap pack of mosfets. Cost you thousands in time and labor but parts cost could be driven close to $0.

-methods
 
At the 3-4Vpc level, takes very little power in watts to get up to very serious amps.

A contactor breaking 500W at 1S needs to be just as robust as 12kW at 24S

Amirite?
 
john61ct said:
At the 3-4Vpc level, takes very little power in watts to get up to very serious amps.

I am unclear as to what you are saying.
3-4Vpc, is that 3 to 4 cells in series (so 3S to 4S) or are you saying 3-4Vpeak, as in HVC (high voltage cutoff)?

Answering both,
Maximizing 3S to 4S equates to 16.8V
Watts = Volts * Amps
Very Little Watts = 16.8V * Very little Amps, but... you have not defined "Very Little" or "Very Serious"

Very Little would typically be mW, say 50mW.
Anything close to 1W would create "serious heat" at the board or component level
Even half a watt (500mW) will easily smoke parts and burn your fingers
So we will call "Very Little Watts 50mW

Plug that in...

50mW = 16.8V * AmpsUnknown
0.050W/16.8V = 2.3mA

2.3mA is a TINY amount of current by the standards we are talking about (Relay Breaking), so, by the proof above, factoring in the assumptions made, I am stating that your statement is likely incorrect.

....
Generally speaking, most "relays' are "Automotive"
They utilize a 12V coil and it likely takes a watt or two to hold them closed
They are often rated for 30A
They regularly switch loads of 10A, 20A... hundreds of watts on a 12V system

For Ebike Use we consider those unacceptable.
This is not for the current flowing (tho it can be). . . but for the potential on the lines. 36V is much harder to break than 12V.

To understand this, maximize the values.
We assume 1 Amp
We assume two cases: 1V and 1,000,000V

When the relay opens, in case 1, there is very little voltage behind the current. Definitely not enough, even with flyback, to jump the gap and ark.

In case 2, at any current, the potential is so huge that it can just ark right across the relay contacts...

So that tells you (along with the relays being STRICTLY rated for max voltage) that the distance the relay opens drives how how of a voltage potential it can break.

Voltage is POTENTIAL
Current is FLOW

A contactor is breaking a flow of current. Now you can think of this as the potential getting angry... so... Potential + Flyback (additional potential created by breaking the flow of current thru an inductive load), which is outside of the scope of this question.

... relays are also rated (or bound by) their CURRENT CARRYING CAPACITY, and this is a matter of internal resistance at the contact points, size of the contact paddles, size of the wires or studs, thermal capacity of the heat sinking material it is made of. If you exceed the current rating of a relay for short periods of time, or at a low duty cycle, this is often OK. If you do it continuously the I2R (Current Squared Times Resistance) losses accumulate in the form of power, which is heat, and the relay melts.

john61ct said:
A contactor breaking 500W at 1S needs to be just as robust as 12kW at 24S

Amirite?

I dont know, here is the math

W = 500W
1S = 4.2V worst case
500W / 4.2V = 120A

W = 12,000W
24S = 100.8V
12KW/100.9V ~= 120A

"Robust" is far too general of a term to use for the "Rating" of a relay.
Relays are rated by the current they can carry and the voltage they can stand off.

So yes,
The 1S and 24S relays in your example would require the same current carrying capacity.

No, the 1S relay would not have to be nearly as robust as the 24S

Proof:
You could run 4pcs of Automotive relay on the 1S application and they would ... possibly survive.
You could run the same 4pcs with 100V and they would fail, very quickly, if blown open under load.
If switched only at 0V, they may work.

Generally speaking costs go up for Voltage Handling. You will find this is especially true once you jump from something like 12V to 2000V.

There are guidelines for minimum spacing, wire insulation values, material insulation values, plating materials on the contacts, ... many factors.

-methods
 
P.S.
And since I dont come around much,
And since I was accused of being a "know it all" - DISCLAIMER

"When I am here I post as much as I can as fast as I can. Most of it is factual. I double-check nothing. I reference nothing. I just write as fast as I can. There will be some number of errors in what I write. Errors occur in poor use of language, poor interpretation of language, a mistake on my part, etc. I always welcome a correction to any statement of fact I make. I prefer a correction come with a proof or some other rational"

Most Importantly,
And seen above,
In the statements I make I "Show my Work"

That way,
more confusion is eliminated... and it is easy for the jr Engineer to point out a flaw in one of my calculations instead of challenging me, my knowledge base, or my assertions.

So
Show your work

Attack work, not personalities.

-methods
 
methods said:
john61ct said:
At the 3-4Vpc level, takes very little power in watts to get up to very serious amps.
I am unclear as to what you are saying.
3-4Vpc, is that 3 to 4 cells in series (so 3S to 4S) or are you saying 3-4Vpeak, as in HVC (high voltage cutoff)?
I apologize for my lack of clarity wasting your time, what I meant was, the example case where the power circuit being open/closed by the contactor is at 1S voltage, i.e. 3 to 4 volt (per cell)

Ideally I would like lots (thousands) of watts, but at that low a voltage, or even nominal 12V compared to say 48V, the amps rating is very restrictive, power not so being so much the issue.

We are talking controlling propulsion battery C-rates here, charging / discharging, in that context "very little" power being hundreds of watts, not board component ratings.

An automotive relay rated for 30A, OK for say 25A, means only 75-100W at the low per-cell voltage, not much at all. While 1200W at 48V.

Responding to this specifically:
For anything low power I think we all know that anything flies. When we get up to high power we are more strict about enforcing System Interfaces, so in that case, the battery really should have its integrated LVC protection that actually breaks the path of current.

______
>> A contactor breaking 500W at 1S needs to be just as robust as 12kW at 24S

>> Amirite?

"Robust" is far too general of a term to use for the "Rating" of a relay.
Well, since I'm looking for devices designed for hundreds of amps, they all seem to be no problem with the voltage rating, therefore it is the Robust current-handling that is the differentiation factor, **in this example use case** voltage can be assumed OK.

Except for a relay rated for say "8-300V", would a 3-4V cutoff present a problem? I've been told no, but why is that lower limit there?


> relays are also rated (or bound by) their CURRENT CARRYING CAPACITY, and this is a matter of internal resistance at the contact points, size of the contact paddles, size of the wires or studs, thermal capacity of the heat sinking material it is made of. If you exceed the current rating of a relay for short periods of time, or at a low duty cycle, this is often OK. If you do it continuously the I2R (Current Squared Times Resistance) losses accumulate in the form of power, which is heat, and the relay melts.

Yes I won't do that. But the fact that I am up at say 80% of power rating every time the contact is opened or closed, means I need quality components (switch, FET, relay, SSR, solenoid, contactor) where I can trust that rating. Since it's a mission critical application maybe preventing fire, and thousands of duty cycles required.

methods said:
Attack work, not personalities
Amen




 
At times there are valid reasons for having a lower bound on a voltage range.

At other times, if the supplier is really good, they will only post the actual range that they Validated. If they did not validate performance below 8V, then they will bound its operation at that range.

That has to do with the passing forward of Qualification for things like Military, Aerospace, etc. where a strict pedigree is required on everything in the system. It is very costly to develop this pedigree, and work is often done twice or three times... so... it is of great value to use standards and have preferred vendors, that you can trust, so that you can use their Quality Data.

Another similar reason you see specs like that is "Requirements Matching". Someone, somewhere, wrote a document that stated that they MUST OPERATE from 8V to 300V... so... that is what they rated it for. You WOULD BE AMAZED by how folks in Procurement often times have no idea what they are even procuring. They are just asking questions, getting answers, and ordering parts. So, by that method, some suppliers make it crystal clear that their part is suitable for selection.

It sounds sort of crony, but if you understand the severity of the quality requirements, any suspicions wear off. There is nothing the government contractors want more than a new supplier that can supply a better product at a cheaper price - BUT - they have to be able to meet the stringent requirements and they have to pay for the testing themselves.

Testing is expensive.

-methods
 
But in practice for less stringent procurers,

if a contactor is rated for 200A from 6-200V

you could be pretty confident it will be OK at 180A at 3.5V right?

generally speaking, not holding you to it 8-D
 
A test is worth 1,000 opinions
Hook it up and see

If the relay is rated for AC, well then, by default, it is seeing 0V all the time right?

Post a datasheet and let me read it. I dont like recommending anything blind. The reason is that, sure... the relay will no doubt work at a lower voltage, BUT, your particular selection may suck for some other reason... leading to failure... which may be misconstrued.

Be very aware that Power Ratings are wildly un-useful.
Some assume a magic infinite heat sink
Some assume you are in cooled oil
Some assume you have wind blowing by
Some assume you understand all the plastic will melt
Some assume you are heat sinking into very large cables...

and btw... the reason why they put a minimum wire size and suggested attachment method is literally to heat sink the contactor out of the studs and into the cabling...

So in that case
You may find
That something like a minimum wire gauge makes all the difference

(so that is why I am superstitious when I see a specifically called out range, instead of a Maximum)

-- Pics or it did not happen

-methods
 
Regarding contactors, what effect does voltage have on arcing ? Some Chinese vendors only state coil voltage, but not the rating of the contacts.
 
methods said:
"When I am here I post as much as I can as fast as I can. Most of it is factual. I double-check nothing. I reference nothing. I just write as fast as I can. There will be some number of errors in what I write. Errors occur in poor use of language, poor interpretation of language, a mistake on my part, etc. I always welcome a correction to any statement of fact I make. I prefer a correction come with a proof or some other rational"

Stream of consciousness is somewhat acceptable when talking to individuals already on the same level of proficiency. Not a good educational approach though, plus it's painful to read. You've been doing it for a while, and likely got similar feedback in the past, so feel free to ignore me too :)
 
methods said:
A test is worth 1,000 opinions
Hook it up and see
Problem is I'm just window shopping, getting a feel for what devices will support different use cases.

Of course testing once off says little about how reliable it will be after thousands of cycles over the years.

At the high end, this is a ballpark I'd like, but of course different use cases I can make do with less

1,500A - Cranking: 10 sec.
775A - Cranking: 1 min
500A - Intermittent: 5 min
300A - Continuous

Gigavac has a great rep, Fujitsu, Tyco, REC BMS recommends one, Luna another, Blue Sea makes great stuff

but all too pricey if a dozen are needed.

What do you think of the specs on this pretty robust MOSFET?

https://www.digikey.com/product-detail/en/ixys/IXFN360N10T/IXFN360N10T-ND/2116933

 
cricketo said:
Regarding contactors, what effect does voltage have on arcing ?
same as it does on anything else, fuses, breakers, switches, etc.

if the distance between contacts when off, under teh conditins inside the contactor, does not exceed teh distance required to break the arc, the arc will ocntinue and either

melt the contacts away preventing any further contact,

or

burn them so that future contacts have high resistance and voltage drop across tehm,

or

fuse them together so contact can never be broken.

or

set fire to the materials around the contacts if they're flammalbe at plasma-arc temperatures,

sometimes nothing bad wil ever happen, but osmetimes one or more of the above things will happen repeatedly and damage things further over time until the worst case failure happens eventually.



Some Chinese vendors only state coil voltage, but not the rating of the contacts.
then i would never use anything by those vendors, as you cannnot know what they are actually usable for, without your own repeated destructive testing on sacrificial units. or ellse taking the risk of a fire, or system failure either not turning on or not turning off.
 
john61ct said:
What do you think of the specs on this pretty robust MOSFET?

https://www.digikey.com/product-detail/en/ixys/IXFN360N10T/IXFN360N10T-ND/2116933

Is that 0.85V drop at 120A ?
 
Back
Top