Atlas 57014 40v 80v Dual Voltage Battery Teardown

As I could not find any information online about the internal workings of the Atlas 57014 40/80V Dual Voltage battery, I decided I'd tear my brand new one down and share my findings with you all.

First off, as far as I can tell, this battery pack is actually two independent 40V packs in the same housing. Seems that the tool you plug the battery into is what either wires the two packs into parallel for 40V 5.0Ah or into series for 80V 2.5Ah. I have specified in the pictures what I will refer to as cell group "A" and cell group "B".

Specs (Atlas 57014):

• (20) EVE INR 25p 18650 cells in total (10 per each 40V pack)
  • 2500mAh capacity
  • 20A continuous discharge rated, 30A with 75°C cut-off
  • 4.2v max charge voltage, 2.5v max discharge voltage
  • 0.5C standard charge rate (1.25A), 1.6C max charge rate (4.0A)
  • 0 to +50°C charge temperature range, -20 to +80°C discharge temperature range
• 10 cells in series configuration x 2
• Temp sensor connected to cell group A
• Battery level indicator seems to only read from cell group A
• (2) 50A fuse on-board, one for each cell group

I do not know what protections the battery management board offers beyond the 50A fuses and temperature protections. I also don't know what the temperature thresholds are for the battery refusing to charge or discharge.

Disassembly:
Use a T15 Torx bit to remove the 4 screws holding the two halves of the housing together. Pull up on the black portion to slide it out away from the orange section. In my case, I could not get the whole pack inside to move around. Turns out that the battery board was glued to the orange housing in a few places. I broke the adhesive using a nylon pry bar (often used to remove vehicle interior trim panels) pushing the whole pack backwards (not prying upwards) towards the spring-loaded latch (reference attached "Pry battery to break glue" picture). From there, the whole cell pack simply lifts up out of the housing.

Connections:
The main battery contacts consist of 6 different contacts. A-, B-, A+, B+, temperature sensor, and another with an unknown function (reference the attached pictures). The temperature sensor is physically touching one of the cells in group B but is electrically connected to cell group A. It is likely a 100Kohm NTC it reads around 92Kohms in a cool room (sensor to group A ground). As for the 6th connection with an unknown function, it reads open to the power, ground, and temperature sensor contacts (sometimes it will briefly show in the range of 50Kohms but within a second increases into Mohms before reading open). I can only speculate but this could be used as a protection active sensor pin or data communication line (it does have a connection back to one of the chips in the charger so it is used for something).

Charger:
The official charger simply connects the two cell groups in parallel for charging. As the contacts are physically bonded to each other in the charger (standard rate, P/N: 56993), even plugging in the battery while the charger is disconnected from the wall will still result in the two cell groups being connected to each other in parallel. This means that if one group is at a different voltage than the other, current will flow from one group to the other. If the voltage difference is too high, this can result in high currents that could result in charge currents exceeding max spec, heat buildup, damaged cells, and even catastrophic failure. So, don't connect the battery to the official charger if you know the two cell groups are not close in balance.

Manually charging:
PROCEED AT YOUR OWN RISK. I AM NOT RESPONSIBLE IF THINGS GO CATASTROPHICALLY WRONG, YOU DAMAGE PROPERTY, YOU CAUSE YOURSELF OR SOMEONE ELSE HARM, ETC.
If the official charger refuses to charge the battery, there is likely something wrong with the pack internally. You will need to do much more extensive investigation and repair work to get things back to working order if that is even possible. I won't get into that here.

If your battery pack seems to be in good working order and you just want full control over the charging process, then there are a few options depending on your equipment and preferences. Charge everything all at once, charge each group individually, split each group into two and charge individually, or charge each individual 18650 cell one-by-one.

Since the nominal full charge voltage of each group is 42 volts, your options change if your adjustable power supply can go over 42 volts or not. >42 volts means you can charge everything in one go or each group individually. <42 volts means you have to split each cell group into two and charge each portion like a 18/20V pack (reference the pictures and use A6 and B6 as the center tap for each group). Regardless of what voltage your adjustable power supply tops out at, charging each 18650 cell individually is very likely the safest and most consistent method to go with though it does take substantially more time. If in doubt, charge each cell one-by-one.

Repairs:
In the process of playing around with this battery pack, I managed to short out cell group B which did pop the 50A fuse for that group. Cell group A continued to work as expected while group B was essentially disconnected from the battery pack output contacts. When plugging the battery into the charger, the charger did not have any complaints but was only charging cell group A. Interestingly, the battery level indicator showed the same amount as before blowing the fuse which leads me to believe the indicator only reads from cell group A. Replacing that fuse brought cell group B back to a working state and did not change the battery level indication.

Now, none of the bus bars are setup to be fusible links as far as I can tell. When I accidentally shorted group B out while connected directly to the main positive and negative of the group (not through the board and protections therein), I did have one cell that died (yes, I killed brand new battery that hadn't even been fully charged yet). Measuring each cell individually showed that one of them was suddenly at around 500mV indicating that the CID (cell interrupt device) had done its job and the cell was now trash. None of the nickel plated copper bus bars showed any signs of damage.

To repair cell group B, I removed the dead cell and replaced it with another of similar specs temporarily until I can get another EVE 25p cell (they are out of stock everywhere at the moment). To remove the cell without disassembling the whole pack, I first used a small 5/64" drill bit to drill out the spot welds connecting the bus bars to the cell. I went very slow with not much pressure to ensure I did not drill through the cell itself. Once I could see some silver metal at the bottom of the hole, I stopped and used a small flathead screwdriver to pop the bus bar off of the cell (be very mindful of shorting out on the other bus bars or balance wires). From there, I used a 1/8" drill bit all around the small plastic lip on the negative end of the cell (don't do this on the positive end, it is far too easy to short the positive to the negative casing). This consisted of about 10 small holes all the way around until I could just see the green wrap of the cell at the bottom of each hole. I then used a razor blade and small flathead screwdriver to slowly chip away at the plastic until the cell could slide out freely. With the cell free, I slid in my temporary cell and connected the bus bars back up. Using an adjustable bench power supply, I charged that cell up to match the voltage of all of the others in the group. I proceeded to discharge group A just enough to match group B in voltage before putting the pack back together and finishing the charge on the official charger.

 

I have seen some reports of that sort of behavior but I have not seen it myself as of yet.

I ordered mine from Newark, but both DigiKey and Mouser have them too. The exact part number I got was: 2822HC50-RTR from Eaton Electronics though virtually any SMD fuse of a similar size (7.6mm x 6mm x 4.2mm) and 50A rating should work just as well.