Signature Solar EG4-WP 48V teardown

Johan

The law of conservation of misery.
Joined
Oct 2, 2019
Messages
71
Location
Sacramento, CA, USA
This post shows how I opened the case of the EG4-WP 48V 100Ah (5.12kWh) battery without damaging any cells or other electrical components. This voids your warranty. First try to contact Signature Solar to see if a less invasive solution exists. In my experience they do their best to help you.

Read this post if you want to:
  • reset the BMS from being stuck in "protect" mode back to "stand-by" mode*, or
  • replace the BMS by a different BMS**, or
  • see what's inside :)

Procedure:
  1. First remove both plastic handle ropes (can be done fully non-destructively);
  2. cut the case all around using an oscillating saw (the famous Will-Prowse tear down method),
    • Carefully saw right above (!) the seam (in my case this worked without touching cells - in your case it may not - it's your risk!
    • The case wall thickness is 5/8in (8mm) all around (see photo).
    • Apply multiple saw passes, gradually increasing the kerf depth.
  3. carefully pry the lid off with chisel (re-solidified debris may "weld" the lid a bit to the case's body locally).

PXL_20220506_182823829.jpg

Clearances (as measured):
  • from cutting plane to cells: 1/4in (6mm); 😅 OMG
  • between case long inner wall and cells: zero :oops:
  • between BMS circuit board green top surface and lid bottom surface: 3/4in (20mm)
There is a brown small microswitch on the BMS that resets the BMS. After briefly pressing, the on-board LEDS of each cell will briefly turn on and off and a few seconds later, the BMS has reset and should now be in stand-by mode. Pulling out wires is not needed.

If you only want to reset the BMS:
Instead of sawing the entire lid off, You could also make a small hole in the lid of the case. I recommend not using a drill (unless in a very controlled manner) because the clearance between BMS and lid is small. Instead, use a soldering stick or a router if you have one. Where to make the hole... this may very well vary 0.25in (6mm) per battery!:
  • 7 5/8in from the handle outer edge (see photo),
  • 4 3/4in from the case long outer edge on the negative terminal side (see photo)
Does anybody recognize the manufacturer of this BMS and if so, if there a way to access it and read/change settings? The BMS board contains:
  • Panasonic AN49503A Battery Monitoring IC (link)
  • Microchip 18F45K22 (link)

@Will Prowse @electric @Trkarl @RichardfromEG4

Notes
*This happens, for example, after connecting to an inverter without pre-charging. In my case, the BMS retried connecting roughly five times (each followed by a 4-min protect-mode-time-out) with a Cotek SD-3500-148. I tried reviving the BMS by charging is 24h using a an EG4 charger: unsuccessful. SS sent me a replacement.
** the Overkill BMS may have the same cell tap connectors, but the JK BMS may be a good alternative too.PXL_20220506_183226902.MP.jpg
This is what it looks like inside.

PXL_20220506_183307261.jpg
What BMS is this? How to access?

PXL_20220506_190706119.jpg
Distance from reset switch to handle outer edge (the ruler's zero is right above the switch center!).

PXL_20220506_190947806.jpg
Distance from reset switch to case long outer edge.
 
Last edited:

Trkarl

Solar Enthusiast
Joined
Mar 11, 2022
Messages
31
Location
Cape Coral, Fl
Have you decided to replace the bms? Are you going to try to top balance the cells like they should have been before the pack was assembled in the first place? I may get adventurous too because two of the three I ordered still have about a 500mv differential if I charge them full. To me that is unacceptable.
 

Johan

The law of conservation of misery.
Joined
Oct 2, 2019
Messages
71
Location
Sacramento, CA, USA
Have you decided to replace the bms? Are you going to try to top balance the cells like they should have been before the pack was assembled in the first place? I may get adventurous too because two of the three I ordered still have about a 500mv differential if I charge them full. To me that is unacceptable.
I am trying to balance first. See below the LEDs that may indicate balancing cells, this was yesterday, 7 of the 16 are on. This morning only 3 were on. The PCB is 10-20F warmer around those LEDs where the balancing resistors likely are. I manually keep a cell-level voltage log in Excel to monitor changes through time. I CC-CV charge using a bench top power supply to 3.45V per cell until the charging current is near zero. But right now it is too early for me to say if the balancing function really works properly. I have to apply a few more charging cycles I guess.

PXL_20220509_232207030.jpg
 

Johan

The law of conservation of misery.
Joined
Oct 2, 2019
Messages
71
Location
Sacramento, CA, USA
What is that goop on the battery connectors?
It feels like a soft rubbery caulk (so not hard like cured general purpose epoxy), like white silicone caulk. Maybe it is intended to prevent or delay the onset of loosening from vibration as discussed here:
 

Johan

The law of conservation of misery.
Joined
Oct 2, 2019
Messages
71
Location
Sacramento, CA, USA
Are you able to determine at what cell voltage the balancer turns on
So far I can only tell you this (quick and dirty):

The balancing LEDs sometimes turn on when I do CC-CV-charging to 54.26V (roughly 3.4V per cell on average). This battery voltage was read from the app, which matches my power supply @ 0.1V difference. However, sometimes they did not turn on at that battery voltage or they just blink a bit. So far I always observed the balance LEDs to be on consistently when CC-CV charging to 55.2V (3.45Vpc), so for now I balance at that voltage. But even then it still goes into protect mode sometimes because one cell hits 3.9V :(.
 

Trkarl

Solar Enthusiast
Joined
Mar 11, 2022
Messages
31
Location
Cape Coral, Fl
If you can get at the terminals you can manually discharge that cell a bit with a pair of alligator clips wired to a 12 v light bulb like a turn signal bulb or something. Just watch the voltage of that cell in the app so you don’t over do it. You could also use a resistor but I’m not sure what size. Hopefully someone else could suggest a resistor to use.
 

Johan

The law of conservation of misery.
Joined
Oct 2, 2019
Messages
71
Location
Sacramento, CA, USA
Update: The BMS increased* the imbalance (wtf) instead of decreasing it, so I decided to take the cells out of the case and manually balance them individually using two parallel 3 Ohm (actually measured less 1.8 Ohm) resistors. Took me a whole bunch of manual actions spread out over a couple of hours. Better had a BMS do this for me :confused:. But I wanted to experience this myself once. Result: 10mV unbalance at ~3.65Vpc. I kept the BMS attached all the time. It was easy to get the cells out, the case is undamaged.

*Cycling from 100% SoC (3.45Vpc) to 95% and back - imbalance was compared after charging at roughly same battery settling voltage.

PXL_20220514_013308730.jpg
Photo: Cell number 1 discharging.
 

Johan

The law of conservation of misery.
Joined
Oct 2, 2019
Messages
71
Location
Sacramento, CA, USA
I hope the BMS doesn't just undo the balancing you did. If it was making it worse before it might just do it again.
Unfortunately I do not have time right now to swap BMSs. I sold the EG4 charger and bought a Wanptek KPS6020D (max 60V, max 20A) on Alibaba for charging to 3.40Vpc. *So far* no cell exceeds 3.65Vpc, the SoC "only" reaches 99%, which is fine for me now.
 
Top