DALY BMS Teardown & Failure Review (16S 40A)

Note that this function is part of the Cellwise control chip, which is by no means exclusive to DALY, and it's highly likely a function in other control chips in many other BMS designs.
methods said:
I just got ripped so hard. They let me draw like half my schematic then grabbed my design to hold it for ransom. Now I remember why I do not use those shitty online schematic drawing tools. F those guys man. I never simulate anything.
I quit once something shady happens. I just started preparing free lesson and some online shitheads stole my design. Lost my taste for it, good day to you sirs.

After mucking about in Circuitlab for way too many hours, I found this site:


Have not seen anything I don't like so far!
You guys are asking good questions and doing good work. I have to go to the coal mine, but I will check back in the evening(s) and respond.

I am not allowed to respond in the AM otherwise 1min turns into 1hr turns into 1 day.

The Good News?
Coal Mine Work is friends with ebike work.

I truly do not know what scares me more. The idea of someone hand tweezing those from a random pot...

* Or the idea that someone bought a reel of resistors that look like that?!!!

If those came off a reel, they would be as gray market and recycled as you could possibly imagine. If they were hand placed... It's the sloppiest process you ever saw.



For folks that don't understand what we are looking at, a robotic arm places all these parts from a reel (basically). All the parts in the reel are identical and aligned. What you see above is only possible in several scenarios.

* Heavily reworked
* Reflowed twice...

They just simply isn't an outcome that is good. Lead free makes it all worse because all the temperatures are higher.

This is why we have ISO 9001 certified process.

I guess in some insane reality they could have two reels going... Lol... Two reals populating the same part in random order (not).


Based on how much solder I see

* Hand applied solder paste

There's way too much solder on those SOT23-3 parts.

Unfortunately these days even highly skilled people are cheaper than machines (plus maintenance) to do a job, not just in China.
I'm lagging on this.
Been working on a pilot production for weeks... But soon it will be the weekend.

That does look hand soldered. Yikes.
I wouldn’t want that as a job.

But all the joints look pretty decent.
This is a BMS that has successfully passed Short Circuit testing. It turns off in less than 1mS against a 58V 12AWG short.


From a sample of over 200pcs I had 1 failure. It landed DOA and presented very obviously. Strong smell, excessive heat, failed to RUN. You can fail a BMS to STOP or GO. There are arguments either way. STOP is safer for battery but far less safe for vehicle etc. GO is dangerous news for battery, but more reliable.

So this one -
It has its issues, but it has features like

* External Temperature Probes
* Internal Temperature Probes
* Short Circuit Protection
* Obviously balance and power cut
* ...

You can see from the name it is 16S rated for 60A. I found a sample on the BahBah, but I have not bought anything off that place in 10 years

That is it tho
* Bottom layer is all balancing
* Top layer is all power handling
* Two sets of 100mil jumpers tying them together

Has a bunch of fasteners all over that I am not pleased with. They have about 3 turns of thread and no locTite. Just asking for it.



  • 001.png
    472.9 KB · Views: 190
  • InsideUnderCover.png
    636.1 KB · Views: 190
  • ShortFastenNoLocTite.png
    177.2 KB · Views: 190
I have never purchased those, I have just tested them "along the way" as they were part of a different system. I have other BMS of this style. I guess the potting is just a bunch of bling that I dont need, because I could at least repair this one if I wanted to.

I have a blown one here and we could check the fets, but -> No point! They can change on every board run (and do!).

Similar Datasheet
Wow... 30mOhms is nothing to be proud of LOL
Be super easy to get that down to 3mOhms, then 1/3mOhm

Similar but programmed for 3.7V instead of 3.3V
That is definitely the same kit

I dont see the temp sensors external, but I see the internal one tucked in. The internal temp sensor is on a wire.**

$70 buckolaz
Those are like $250 if you buy it from a retailer.

I can confirm good consumption current, not noticeable
Here is a nice link

Unfathomable $37 single piece sample price. I will buy 10 right now at that price.

Oh, I see
$25-35 but you have to buy 200pcs

Ohhhhh Kay
Not sure I am ready to dip my junk into that blender quite so fast. I IMAGINE I could ship them a few tubes of fets to build with, but would that not be a nightmare.

Sample order does me no good. I need like 40pcs

Oh God... roll the spam dice




  • request.png
    33.4 KB · Views: 193
Forked. :kff:

40mOhm @ 60A would be something like 144W of heat dissipation. You could heat up your lunch with that.

Hopefully it’s actually a lot less than that.

Even 14.4W is kind of lot but certainly manageable.
Super Good Deal
* I vouch for the 12 volt 100 amp hour
* I hammer them in my golf cart four strung in series


I paid like $420 each
Worth every penny


P.S. yep - the 20mOhms in the data sheet is probably garbage
There is another video not posted but still available where I make the argument that you have to have short circuit protection.

I now judge bms's by whether or not their short circuit protection works**

If it doesn't work, then the first time you have a big snap, you might have shorted your fets

Data point
For the last few years I've been seeing almost everything fail to Short Circuit as opposed to open circuit. Those who've been around for 15 years will chuckle.

Ok, memory held 1 week


We are on to what was QA002 or QC002 marked FA002


Red on red so it's easy to miss. It's been hooked to the battery all week and as it sits I read the following

With balance tap attached and valid battery voltage
* Diode check reads short in both directions


With balance tap removed and unit powered down
* Read diode 500mV with anode toward the pack

That would indicate the other fet is shorted. As the capacitance in the circuit bleeds away (or fills up from the meter) the reading becomes harder to grab.

* Very clear and strong reading right after power down
* Mushy reading after a minute

Yeah something is charging up from the meter


I DID NOT "steal that cup"!
They stole my cup!!! It's exactly the same but it doesn't say guest. I put it through the dishwasher and then it disappeared. Some gaffler on site.

That was the gnarliest job ever...

Okay I need to cut this one out of the pack. I think it's time to start connecterizing. I use wago 221 now

* Big one
* Little one

The big one will go all the way up to 10 AWG. The little one will go up to 12 AWG. I have them in every size and shape.