I suppose it’s no secret that aftermarket clone batteries are the life-blood of people who need their devices to last all day. After becoming a Nikon D3200 owner, one of my main complaints was the short battery life of the standard EN-EL14 battery provided.
In an effort to prevent the use of clone batteries, under the guise of safety, compatibility and warranty, these batteries were specially encoded with some form of “DRM” so that plain regular clone batteries couldn’t be used. Instead, the camera would display an error message and refuse to operate.
“Fully Decoded” batteries then became available which emulate the original EN-EL14 battery, such that it passes the genuine check. These were slightly more expensive, but they were still significantly cheaper than the original replacement.
I purchased two, and they both worked initially, with no trouble. The outer shell of the battery seems to be a complete clone of the original battery, with spaces for the hologram sticker and a port cover. It seems as if they had prepared the moulds for making “fake” genuine batteries as well.
The battery functioned for about ten cycles, and then started degrading significantly in run time, before then being refused by the charger soon after placing it on the charger. Instead of the slow regular charge-blink, it would change over to the “error” not-charging fast-blink. I can confirm that the battery did not end up charging much, and would take about five photos before pooping out. It was cheap, but it was also disappointing.
It would be a good thing to take apart and see if I can learn anything about the pin-out, how the battery failed and how the DRM works.
Tearing it Apart
The battery was encased in a thin plastic shell which was tightly glued together at the seams. By applying pressure, the adhesive cracked away enough to get a flat-head screwdriver into the gap and run it around to separate its halves (not without damaging the bottom part of the shell though).
With the shell halves separated, the inside reveals a double sided adhesive spacer on the top of the lid. The battery was in the other half, the type being prismatic lithium-ion.
In between the two cells, there is another double-sided adhesive spacer.
The control and protection PCB rests on top, with an isolating pad and some foam adhesive spacers.
A small amount of insulating tape is used to insulate some of the connections.
The funny thing is that there’s a double sided adhesive tape on the rear (in white, foreground of image) where the protective tape wasn’t even removed and thus adhesive ineffective.
The PCB is where most of the magic happens. Lets take a quick look at the connector – VBat+ is on the leftmost contact (1), and the next one along (2) is connected to V+ through a resistor by the looks of it. The rightmost contact (5) is Ground, whereas the next left contact (4) seems to be connected to Ground through a resistor. The middle contact appears to be connected to a voltage-divider – I am assuming this is the data pin, where the data comes from the IC and lowered to the right level for the camera. Pins 2 and 4 might have originally be used for a thermistor, to monitor battery temperature, but the clone battery may have neglected to implement it.
Aside from the passives, there are several chips on this, most of which are unidentifiable. The main IC is marked LP1K36A A2391R60 MALAYSIA. This is possibly a microcontroller of some sort? There is a 6-pin IC labelled MOR8098Y, another labelled 3W23. There’s a four and a five legged IC with difficult to read markings. The only thing I could identify was the AO8810 Dual N-Channel MOSFET.
The PCB itself isn’t very high quality. You can tell from the fuzzy lousy silkscreening and the poor alignment of the drilling to the traces themselves. This was definitely made to a price.
Lifting the PCB up reveals the insulation on the underside, and a single non-resettable fuse marked LTV175. It’s nice to see at least this one form of protection aside from the software on the IC.
The underside of the PCB shows several test pads (there are some on the top too). The PCB is marked RHD033-V4 and is dated 6th October 2012. It claims to be RoHS compliant.
The connection of this battery is interesting. It is connected with the centre tap – this is presumably for cell balancing, which is a feature used to ensure the cells are balanced in their charge level to prevent damage to cells from being overdischarged. I didn’t expect a cheap aftermarket replacement to feature such care … but unfortunately, it wasn’t enough.
Why did it die?
From the outside, none of these cells look particularly traumatized. There’s no venting, there’s no swelling. The first clue to their trouble comes from the multimeter.
Under open circuit, with my Keithley Model 5110 DMM, Cell 1 measured 4.1389v and Cell 2 measured 3.6406v. That’s a big discrepancy which shouldn’t exist. Cell 1 is almost fully charged, whereas Cell 2 is in trouble.
Why might this be the case? A hint was given upon closer examination of the cells.
Unfortunately, there was no information about the cells, which are coded LS553448. An equivalent datasheet from Samsung (real, quality cells) suggests that it is (at most) 1030mAh per cell. Is it any wonder the genuine EN-EL14 claims 1030mAh? It’s definitely not the 1400mAh claimed on the outside.
Low quality cells is definitely a probable cause – one of these cells may be “leaky” due to internal low-resistance paths, or has lost capacity drastically (the higher voltage one at full charge while the other hasn’t achieved full charge yet). Maybe it’s a manufacturing defect, contamination, etc. In any case, I can’t be sure it’s not dangerous or won’t develop into a bigger problem in the future, so I’m swiftly disposing of them.
But what is even more damning is that the font on both cells is different. This suggests they are batteries from different batches, which means they may have been mismatched in capacity from the factory. This (if not handled carefully) will likely cause premature failure or limitations in capacity.
Now we’ve seen the inside of the battery that failed, and we have some idea why it failed. Lets just hope the other fully decoded clones are made with more care.