Author: Ed

Back when I got a Philips Sonicare (on the recommendation of my dental hygenist, after a particularly nasty bout of plaque removal), the battery gave nearly two weeks of service between charges. As shown in that graph, the runtime gradually faded away to two days, at which point I decided it was time to tear the thing apart and see about replacing the batteries.

The instruction manual tells how to dismantle the case and extract the NiCd battery for recycling:

Please note that this process is NOT reversible.

Well, there’s a challenge if I ever read one, but Wouldn’t It Be Nice If you could take something apart, unplug its defunct battery, install a new one, and button it up again? Then you wouldn’t be forced to buy a new $70 toothbrush, which probably explains everything… and I suppose the replacement battery would cost $40, even if it were a pair of AA cells.

For reference, the instructions (clicky for more dots):

As predicted, suasion applied through a small screwdriver popped the top end of the case apart, but the remainder required concerted prying and muttering. The case halves mate with a tongue-and-groove joint that’s either sonic welded or adhesive bonded to form a watertight seal all the way around, to the extent that they suggested cleaning the thing in a dishwasher.

Eventually, though, it came apart:

The “motor” (actually, a solenoid that couples to the magnet on the brush stem) is firmly potted in place (on the right), as are the NiCd cells and the charging power pickup coil at the base on the left. The potting compound seems to be a clear epoxy, rather than a compliant rubber, and it doesn’t bond to the case at all. It is, however, a perfect fit and doesn’t pop loose without a struggle; their instructions will definitely break the PCB.

Seen from the other direction, six connections join the PCB to those immovable objects. The four pins (on the far left) go to the solenoid and the pair (just to their right) to the battery:

A few dabs of desoldering wick suffice to free the pins and release the PCB. Mercifully, the potting compound surrounding the charging coil slid out easily, as they (inexplicably) omitted a mechanical lock molded into the case:

Removing the NiCd cells required considerable prying, as described in the instructions, that en passant damaged their cases. I think if you weren’t paying attention, you could easily rupture a cell case with the screwdriver and spatter the area with potassium hydroxide, perhaps shorting the cell in the process and producing rather more excitement than most folks expect.

A closeup of one cell; the other bears similar damage:

I snipped off the cell tabs and applied them to the new NiMH cells. A bit of closed-cell foam between the cells and the PCB cushions the assembly:

Stacking more foam snippets under the cells filled the space left by the potting compound, then soldering the solenoid pins held everything together:

A wrap of clear adhesive (rather than the obligatory Kapton) makes for a tidy joint that probably won’t last very long, but it looks much the way it did before the operation. The case is no longer waterproof and won’t withstand the dishwasher. In fact, I must now store it with the brush end downward to keep the last few drops out of the handle.

There’s an interesting solder jumper on the PCB that I didn’t bridge, but the next time it’s opened up I’ll apply a dab:

The alert reader will notice that I’ve replaced 2000 mA·h AA NiCd cells with 600 mA·h 2/3 AANiMH cells, without changing the charger. The power transfer through the inductive coupling drives a trickle charger at about one hour of recharge per brushing, so there’s not much danger of overcharging the cells.

Now, to discover what runtime fresh cells deliver. This calls for another slip of geek scratch paper in the bathroom.