The other day, my Dad came to me with his Colgate electric toothbrush, complaining that it was weak and didn’t want to stay running. He suspected it was the battery and wondered if it would be possible to replace it. As an engineer, with some pride, I decided I would not be defeated by a toothbrush!
Taking It Apart
Getting it apart was a bit of a puzzle, however, I didn’t anticipate blogging about it so I didn’t bother photographing the process. It begun with digging at the bottom where a black plastic panel was removed, underneath was a pink silicone plug which revealed a single Philips head screw. Removing the screw allowed for the bottom section of the toothbrush to be slowly prised out – it has a good silicone seal all around.
The “core” of the toothbrush assembly was, however, held in place by clips on either side which had to be simultaneously depressed while pulling on the core. A scrap of stiff business card was useful to hold the clips away from their slots.
The toothbrush is a simple rotating off-centre mass, located in the middle of the head stem. Mechanically simple, the PCB appears to be rather interesting with what appears to be a switching converter visible in the right. The battery is located in the centre.
The other side contains a PCB marked HT-B45X-A Main NEW. It has positions for four switches, just one in use in this case. Likewise, there are locations for another few LEDs, perhaps for higher-end models. The unit is controlled by a single microcontroller which I did not identify. The lower portion of the unit appears to be an air-coiled induction coil for charging.
While initially, I thought it was a simple AAA cell, this is actually a battery made of two half AAA cells. That’s not a common size that can be obtained, so I’m out of luck in terms of returning it to the original state. But that’s not going to stop me.
Making the Modifications
I decided to think outside the box, quite literally. Step one was to extend the battery leads which were quite short. Marking the side of the unit, I drilled a sizeable 5mm hole, enough to allow the unit to slide back into place and for a set of tweezers to fish-out the wires.
Sealing is courtesy of hot glue, leaving two flying leads. I hooked it up to a bench top supply at 2.4V and the unit definitely powered up and ran, so I didn’t destroy it in the process of reassembly.
I decided to use a pair of 2013 AAA Sanyo Eneloops with plenty of capacity. Filing the terminals allows the cells to be easily soldered, even though you should not solder directly to battery terminals. I find that if I do it quickly enough, the batteries survive, but there’s always a risk of battery damage or explosion, so do so at your own risk!
To secure the batteries, copious amounts of pink Nitto tape was used to wrap it up, with care to clear the base where it sits in the inductive charging cradle. It’s now bulkier and more awkward to use, but not bulky enough to be unusable.
Handing it back to my Dad, he was delighted at the hack – it’s better than throwing it out!
Testing the Battery
The battery was topped up on the R&S NGM202 Power Supply and then discharged right down to zero.
The resulting capacity was about 42% at C/5 to 2V, or 62% to 1.6V. However, the toothbrush does consume high current, so increases in internal resistance may have further shortened the run-time beyond what this capacity figure might suggest. The battery has lost a significant amount of its capacity, which explains why the unit failed to run for any significant length of time.
Conclusion
For most people, buying a new electric toothbrush is probably the best option when their unit starts to fail, but that doesn’t have to be the only outcome. When challenged by my Dad to revive this unit, I had to make an inelegant hack, but it definitely works and could probably continue to work for a while longer. Better than going to landfill, perhaps, but I wouldn’t be surprised if the cost of those old Eneloops exceed that of a new toothbrush bundle, since they usually come with a few heads to sweeten the deal.
Bonus: Another Fix
As it turns out, this wasn’t the only thing I attempted to fix this week. I also tried to repair the non-functioning fourth channel on a Rohde & Schwarz HMP4040.04, but unfortunately, even after replacing a pair of MOSFETs, a pair of Schottky diodes and a pair of fuses, the channel did not come back to life. It was worth a shot, given the cost of the components, but it’s likely to indicate further collateral damage that I didn’t spot (e.g. to the switching controller or other MOSFET/driver ICs).
At least I didn’t end up making things worse – the three functional channels remain functioning. Sometimes repair is knowing when to give up – trying to push your luck too far could well end up costing you more. Read the full journey on my element14 blog here.