DONE - Set the 80V battery and tool Kobalt/Greenworks finally free

[Zucca]

Hi,

Just got a 80V Kobalt mower with two knock off 80V chinesium batteries (long story).
BMSs on the two batteries are okay, cells are balanced and happy. The batteries are also charging no problem on a Kobalt charger.

Issue is when I use the battery in the mower after 3 sec, I got the beeping and the tool shut off...

It already happened in the past...
https://www.eevblog.com/forum/reviews/greenworks-60v-battery-ohm-terminal/

I am assuming that the bidirectional omega line is making the tool upset and cuts the power off.
This pisses me off immensely because clearly it a proprietary restriction.

I am trying a new approach...

I bought the cheapest 80V Kobalt batteries I could find on ebay and then here what I have in my mind...
1) separate the BMS board from the cells
2) try to simulate the battery to make that BMS happy
3) install everything INSIDE the tool, so that the tool needs only 80V (Bat+ and Bat-) from outside to run... no more omega line...

In other words instead to go crazy and simulate with a µC the pulses on the Omega line, I will use a disconnected Kobalt BMS/Battery board to create it for me...

some baby steps tonight...

Pulling 4A from chinesium battery... with nothing connected on the omega line...

works....

Board inside the mower, potted to protect it from water/rain.... makes sense...

1 - Cabling to the brushless motor... 3 phases? Why 3 thick wires?
2 - Negative and omega line (blue cable)
3 - Positive cable

Wish me luck...

[Zucca]

All right, ebay battery is here.
20S1P, last 5 cells are at 2.5V no wonder it will not work.
All the others are about 3.75V-3.79V.

Some observations:

1) BMS is monitoring all the 20 cells voltages with 21 wires.
2) there is a T sensor.
3) there are 2 board: the first is the BMS that also generate the COM/Omega signal, the other is just an battery connector adapter with some big mosfets on it.
4) the board is tortured with some soft white rubber to keep everything in place and prevent intermittent contacts.

Next steps:
1) Balance the cells and test the resurrected battery out
2) take apart the battery... and simulate a good battery cells with 20 1K series resistor between batt- and batt+
3) smoke it out.

 

[Zucca]

No good news today.

The battery is not recognized by the Kobalt charger... the charger does not react at all... not even a light... it seems the battery is made of air.

I tried to power the mower by the battery but it shuts off after few seconds with beeps.

Here the cells voltages

3.74292   3.74292
7.4934   3.75048
11.2368   3.7434
14.985   3.7482
18.7269   3.7419
22.5105   3.7836
26.2942   3.7837
30.0777   3.7835
33.8692   3.7915
37.6555   3.7863
41.4397   3.7842
45.234   3.7943
49.0139   3.7799
52.8054   3.7915
56.591   3.7856
60.3705   3.7795
64.1497   3.7792
67.9274   3.7777
71.6995   3.7721
75.4647   3.7652


Mean   3.773235
Standard Error   0.00397944795827662
Mode   3.7915
Median   3.7797
First Quartile   3.76151999999999
Third Quartile   3.78455
Variance   0.000316720121052639
Standard Deviation   0.0177966322952585
Kurtosis   -0.808156214833955
Skewness   -0.85095591128901
Range   0.0523999999999987
Minimum   3.7419
Maximum   3.7943
Sum   75.4647
Count   20

will now balance the cells to 4.2V full charge and see... meanwhile I will buy for testing a brand new 80V Kobalt/Greenworks....

[Zucca]

Some progress, with the new battery everything works.
It could be that Kobalt/Greeworks realized they were cloning the battery omega/com protocol and reacted by introducing a new second one.
So good old battery is a no go with new tools, with new one everything fine.

Now that I have a good working battery, I will try to strip off the BMW board and move it inside the mower...

Wish me luck! 

[Zucca]

I successfully took apart the BMS without any sparks or shorts.

Attached some pictures of the journey.

Does anybody knows what is the best method to disconnect/break the spot weld tabs so I can recycle the cells?

Next step is to built a 20x1KOhm resistor ladder to mimic the cells and make and see if the BMS is happy with it.

 

[Zucca]

So I installed 20 1.5KOhm resistors in series to simulate the 20 cells in series.
Connected everything but did not work.

I am now trying to rebuild the battery I took apart hoping it will work again.

[Zucca]

Battery rebuilded. Works like new, pfeeewwww 
I think there us a resistor on the board for the battery terminals.. not on the BMS.
It could be the BMS is sensing the battery terminals board and it was not happy without it.

Anyway next move is to install a second battery in parallel with a diode connected to each battery to prevent a charge sharing.
I will use a SBR20A100CT, which is overkill but it is in one of mine bins.

For sure the BMS is not sensing the battery current, there is just a fuse... so let's see.

[Zucca]

Done.

I also added a $10 chinesium battery monitor.
At the end it was easy just two batteries in parallel with a diode in series on each one to prevent charge sharing.
So goal achived: my mower now just need 80VDC to work.

To all Greenworks engineers/management who designed/requested that proprietary one wire bus, you can't win again the rest of the world.
A remarkable suggestion: do not design system to piss customers off, but to make them happy.