Horrible quality of SmartBMS from bmsbattery.com

P

powersupply

100 W
Joined
Apr 14, 2013
Messages
235
Location
Santa Monica
I got 2 SmartBMS (Smart BMS 5~13 Cells in Series) from bmsbattery.com through an ES member.
http://www.bmsbattery.com/smart/330-lifepo4lithium-ion-smart-bms-for-513-cells-in-series.html
Most parts looked hand soldered, so I inspected them thoroughly.

I noticed several shorts and there were part discrepancies between the 2 boards.

Only one board turned on though it had one resistor shorted.

The other had some shorts, too, and missequipped parts which I need to replace.

Another member here had a toal of 3 SmartBMS from bmsbattery.com fail/never work.

What blew me was the greediness of bmsbattery.com.
I read that they omitted the SMD LEDs at some point, but this???

Though they advertise this SmartBMS as "Smart BMS 5~13 Cells in Series", they did ommit the 13s pin and even cable for a version ordered as 12s.

I just can urge everybody to really inspect their SmartBMS from bmsbattery.com.

And when you order, order them as 13s, so you get all the parts.

Click on the pictures to see the in high res.

Good luck with your SmartBMS from bmsbattery.com !


CIMG1911s.jpgCIMG1917s.jpgCIMG1921s.jpgCIMG1926s.jpgView attachment 2
 

Attachments

  • CIMG1927s.jpg
    CIMG1927s.jpg
    273.2 KB · Views: 4,836
BMS battery does not manufacture the stuff they sell. they buy it from the cheapest source. for a good BMS at a reasonable price you have to use the guys that agnius found.

http://www.bestekpower.com/

they have all kinds of BMSs and they are a manufacturer but i have been able to talk them into selling individual units. some of us split shipping on an order recently.
 
I would like to know which ebike battery company uses their bms. I see so many threads about bms problems. Nice to know which supplier to trust.
 
Wow. That is worse than my soldering. What the heck?
 
I think they are ecitypower. You might also find the last cell in your string (12 in this case) is not hooked up the cell 13 to tie too vcc through r1 as per the chips spec sheet. The boards may also require a reset before they turn on, which is a user definable setting if you have PC comms. Not very obvious when they arrive with no documentation. Documentation that describes the connection order, as just plugging them together in any old order may kill them.

Knowing they are rubbish I ordered two. Both were just chucked in the box with my motor that damaged one in transit(mashed a resistor from the board). Both came for 12s with no tie between 12 and 13. Only one worked(after replacing the resistor)and I blew that up as the charge wires are not protected against shorts as I had imagined. Thus I took unnecessary risks. I then ordered a 3rd from a european supplier that also came with the 12/13 connection missing.

This is not a consumer market. However they are $20 which is nothing. The competitors cost more and are not so fully featured. Missing off things like pack temperature sensing which would kill the power at raised temperatures, before thermal runaway took hold. Such protection was on my first ever ni-cad powered 12v drill. I kind of expect it now.
 
Friendly1uk, I read about your (and many others) bad experiences with bmsbattery.com, sorry about that!
But I must say that these are not really $20, as the shipping charge is a major revenue now.
The member I bought my 2 units off, had paid $30 shipping for these tiny boards, something that costs $3 max from China.

Even if BMSbattery does not manufacture the stuff they sell, they still should have *some* sord of quality control. I mean the botched soldering job was clearly visible.

I looked at bestekpower.com, too, but as you mentioned, they only do large (and much much more expensive) sales.

Anyhow, I just wanted to warn about what to expect from bmsbattery.com.

It could be a good deal, but only if you are willing to check the items you receive thoroughly.

And you must have the skills and equipment and parts to correct their errors of course.
 
That's the worst SMD soldering job I've ever seen. Looks like they may have reworked them. Not the kind of job you want to give to a kid making his first attempt at SMD soldering--with nothing to work with but a conventional iron with a big chisel tip. One of those resistors has a big chunk missing. Unpardonably bad for a "new" product at any price.
 
Found a post from methods describing that the main cap is not connected.
Indeed, it is not!
Haven't found his original post about that yet, and will add pcitures etc later.
 
I bough two of these from bmsbattery. One failed right away due to "bad connection sequence" (not documented anywhere). Other has a corrupted config, unable to recover (with their expensive programmer).
I contacted bmsbattery and explained things. They offered a free replacements. I had to pay $150 for shipping, of course. The offer did not seem very attractive, so I did no play along. I reminded them that I still have a few days left for paypal dispute. After that it was like "OK, I understand, replacements boards will be with you shortly". And I have received them within one week. But no win, as these are just as bad products. A lot of time and money wasted.
 
Here it is, C4 is not grounded, the track stops short of the greound via for a few hundret um.

Found it via this thread, http://endless-sphere.com/forums/viewtopic.php?f=14&t=48166&start=400#p808151
where methods talks about it, but couldn't find his actual posting.

Please methods, you mentioned you reverse engineered the board a bit, please share those findings, even if it is just a scribbled note somewhere.
 
very strange...
in most cases smartBMS perfect quaility
I dont know what can be better for LiPo than smartBMS
 
Hehe, I have checked my smartBMSes... All of them have this issue with capacitor. It is not because the PCB is badly manufactured, it is the design itself. Looks like someone has rushed out for manufacturing before checking it properly.

My guess is that bmsbattery is selling out a faulty batch of items...

iperov said:
very strange...
in most cases smartBMS perfect quaility
I dont know what can be better for LiPo than smartBMS
Click to expand...
I somehow would not be surprised if this post same from China.
 
Who needs more proof that these products from bmsbattery.com are junk!
I mean there are pictures and a simple search unveils that many had non functioning units from the getgo.

And given the lack of quality control, all other stuff from bmsbattery.com could very well be junk, too!

I dislike their shipping cost scam. They should be honest enough and put the real price of the units they sell and also charge the real amount for shipping. $30 shipping from China for these 5x12cm boards is just wrong.

Circuit, methods mentioned this capacitor issue during his exchange with you.

Do you know where his original posting about these SmartBMS is?

Thanks!
 
iperov said:
very strange...
in most cases smartBMS perfect quaility
I dont know what can be better for LiPo than smartBMS
Click to expand...

I agree, I only had some shorts at the main FETs and had to short b12 and 13 myself. Ordered 10plcs total now, all work flawlessly. Always test them before I use them:

-calibrate the ADCs
-I short the charge leads
-set the HVC to 4V and charge them to 4.2V, see if it disconnects
-draw 2500W till it says overcurrent
-set overcurrent to 160A and then discharge with 2500W till it reaches overtemp, temp probe is glued to the FETs

All of my boards did pass all tests yet. The original programming is useless, the boards dont even survive a short circuit on the discharge leads as friendlyuk already mentioned. BUT with the right settings it works just fine.

However the main cap was never connected. Thanks for posting this, I have to catch this up on all 10 boards :roll:

Edit: I have to add, some boards lost I2C connection to Arduino during heavy discharge. Maybe the main capacitor is the reason. Or maybe it is just a fail in my I2C connection.
 
methods said:
The first link - the balancing scheme is so poor that it basically does not balance batteries. If you investigate the method it uses you will find that for anything other than new, perfect, batteries the BMS will terminate high current charging nearly every time before the pack is balanced.
Click to expand...

maybe he forgot to turn on idle balancing?
 
Crossbreak, you give me hope with these SmartBMS I have.
I thought they are just utter junk, looking at all the shorts and bad layout.

I stumbled about the main cap not being connected while reading one of these interesting threads/exchanges.
Was trying to understand the issues with the SmartBMS.

Must say, amazing engineering from OZ.

How their chip would work without this cap (they also mention a 100nF cap, that is missing totally).
Impressive!

How do you draw 2500W from these, and won't the FETs get fried?
 
There is a bug in chip itself, that makes proper balancing impossible. Not so great engineering.
So either crossbreak has some units with different chip revision without this bug (which I doubt) or he simply okay with the way it is.
 
why does the lack of continuity of that cap keep the BMS from working? it is the cap on the high voltage input that powers the microprocessor so it is not critical. i don't think i have seen anyone actually prove it is not connected through a via by testing continuity with a voltmeter. until someone does it is not established imo.

the biggest problem i had with these BMSs that use the O2 micro microprocessor was the hysteresis of the HVC cutoff. it cut out at 3.9V and did not turn back on until it got to 3.65V after draining off the voltage on those shunt resistors so it went terribly slow. this might be why some think it is not working too. they may not see it turn back on.
 
dnmun said:
why does the lack of continuity of that cap keep the BMS from working? it is the cap on the high voltage input that powers the microprocessor so it is not critical. i don't think i have seen anyone actually prove it is not connected through a via by testing continuity with a voltmeter. until someone does it is not established imo.
Click to expand...
Yes, please measure. And attach a valid meter's calibration certificate, otherwise it does not count!

dnmun said:
the biggest problem i had with these BMSs that use the O2 micro microprocessor was the hysteresis of the HVC cutoff. it cut out at 3.9V and did not turn back on until it got to 3.65V after draining off the voltage on those shunt resistors so it went terribly slow. this might be why some think it is not working too. they may not see it turn back on.
Click to expand...
It is so slow because it is not balancing in HVC state. Hence the bug.
 
thanks for clearing this, circuit. I will take some BMS to unisversity and put it on a scope (should have a certificate, I'm, not sure) ), we need to do this anyway to get a reliable I2C connection.

At least in software mode, you can turn on balancing even in HVC state.

the biggest problem i had with these BMSs that use the O2 micro microprocessor was the hysteresis of the HVC cutoff.
Click to expand...

hysteresis can be setup. I use idle balancing, most times the cells are very close together when hitting HVC, can't really tell how this behaves if baldy balanced. I will test this. Normally it takes a while, then it charges again for some seconds.

circuit said:
So either crossbreak has some units with different chip revision without this bug (which I doubt)
Click to expand...
I dont believe I have a different chip, since I ordered 3 times within almost 2 years. Quality dropped each time. The first 2 boards I got worked perfectly out of the box.

or he simply okay with the way it is.
Click to expand...
No I'm not pleased about some functions, that's why I want the software mode.

powersupply said:
How do you draw 2500W from these, and won't the FETs get fried?
Click to expand...
There are 3pcs 4410Fets on the board. each one has max. 10mOhms, all togehter around 3mOhms that is a loss of around 12Watts in the FETs at 60amps, which is still with OK propper heat sinc.
 
circuit said:
There is a bug in chip itself, that makes proper balancing impossible.
Click to expand...

Hello again Circuit. I'm still trying to see this fault you talk of. My pack balances fine. I don't see an ov fault unless my charger is set to high, which is an ov fault. With the correct charger, charge currant has dropped off to below balancing currant so no cell can get ahead of the rest.
If this board had a low balance currant or you used an incorrect charger then we could have an issue, but all seems well.

This post is mostly theorising. Trying to find why it works and still accept there is a fault that stops some people from balancing properly. Although that too seems theoretical. Only seen in packs so out of balance that it should not of been in use. So far out that pack voltage was low enough to keep charge currant higher than balance current. Thus a single cell could reach the ov setting.

Mine works, so the fault lies outside what I'm doing with it. I think I'm getting closer to understanding the issue though. I think the bug is people asking to much of it. A common theme. I don't denigh it could be better though.
 
The bug shows itself when your cells get disbalanced due to some reason. You can do a test: charge your pack fully, then discharge it by, say, 2% of total Ah and then charge one cell fully. And then re-connect your charger to the pack and observe what happens.

It's all good until you cells need actual balancing.
 
dnmun, I *did* measure it, that cap is indeed open. I can affidavit to that, too if you want.
However, I must admit that I used an ohmmeter for testing continuity, not a voltmeter per you suggestion.

But seriously, having no cap whatsoever at the input of a high voltage linear regulator and still function as if?
That IS good engineering in my book.

Also, we all know that bmsbattery.com is probably not producing these, but as a reseller far away SOME optical inspection is the least they should do.
Especially given the pattern with the bad history they have here.

friendly1uk, what charging current would you recommend for a 20Ah pack?

Hope crossbreak can shed some light into this balancing / no balancing issue with his (hopefully certified) scope.
 
i meant using the ohmeter. one end of the cap should be on ground and the other end is on the top pack voltage. so check one end to ground, and the other end to the power.

for power there usually is a separate wire to the IC from the top of the pack and the top sense wire is separated from it. in my experience anyway. and they also have a TVS in there sometimes too. this is to protect the IC from reversing the battery voltage to the IC.
 
dont have that scope here... have to upgrade my test stand to take it to university, the error only occurs if the BLDC controller is in PWM region. Sadly my dyno is too weak to keep the motor in current limit region. Maybe I should just use less throttle... but the error occurrs only on heavy loads.

What I can do is just set HVC to 4V, take an unbalanced pack, charge it with EOS1420 and post the graph. Just tried, but I have trouble with my EOS, just wont connect to the software even if Windows recognizes it :x today's just not my day
 
You must log in or register to reply here.

Similar threads

K
12S Bluetooth Smart BMS from LithiumBatteryPcb has probably drained my Battery
Replies
9
Views
1,013
john61ct
J
LewTwo
Battery Autopsy
Replies
18
Views
1,361
LewTwo
LewTwo
E
trouble sourcing 12s bms/balancer
Replies
10
Views
1,138
ejonesss
E
A
18s build with 2x ECityPower bms (6s and 12s) in series
Replies
0
Views
543
altermontrealiste
A
LewTwo
RD6006 0-60 Volt Battery Charger
Replies
4
Views
2,090
Pizzabroodje
P
Back
Top