Teardown Tuesday. What's inside a Braun electric rechargeable toothbrush? And some basic measurements on wireless power transfer and charging. Chips Used: EM6682: www.emmicroelec... TSM7401: www.goodexcel.c...
Dave, the transformer circuit is a simple split coil dual phase rectifier. For the positive phase one coil and one diode rectifies, for the negative phase the other coil and diode do the rectifying. In this way the efficiency is better and the distortion to the primary is less.
The 800Hz pulse is so that the microcontroller can read the battery voltage without the charging voltage interfering. Learned about that little gem watching bigclive tear down some battery charging units.
The 2 little black thingies are pellets that neutralize chemical gasses that can be produced when a nimh cell is overcharged. I've heard of some cases of exploding devices that did not have these neutralizers, so most enclosures with nimh cells now contain one or two of these. See: www.generalchemical.com/assets/pdf/Potassium_Hydroxide_Pellets_ACS_PDS.pdf for more explaination :)
You can actually replace the battery inside, at least once or twice, until the sealing isn't good anymore. I actually "repaired" mine twice before I bought a new one, because the only part that actually dies is the battery, and it dies fast because of the constant charging. I also replaced it with a NiMH one that had nearly twice the capacity than the stock one. It's a bit of a fiddling, with the soldered battery contacts, and really a shame that they don't bother to make it more accessible, but that is probably just planned obsolescence so you buy a new one at least every two years or so.
I tore mine down when it failed, cleaned the guts up and put it back together :) The charger on mine seems happy enough running on 120 volts even though it's a 240 volt model.
The pulsed charging is used to detect if the battery is fully charged, and then the charging will stop. So the IC goes like I'll charge for 400 ms then stop 100 ms and measure the voltage, then repeat if lower than full charge. and btw, it's nice to see the younger you :) we've all changed over the years
I pulled one of these apart and all you have to do to get the whole lot out is to unscrew the bottom and pop it off like you did, and then invert it and push the shaft against a table top or the floor (whatever is hard and you won't damage the surface) and it will push the entire assembly out the rear end in one piece.
My guess is that pulsed charging is some sort of inteligent charging control, done by the micro, in order to extend the service life of the battery. Basicaly, every time charging stops, ADC can read the battery voltage and decide whether to continue or not.
It's called Teardown Tuesday for a reason. Each week is new tear down. Sorry for those who think it's maybe replacing other material. Obviously I can't do enough videos to please everyone all of the time.
Dave I replaced the battery in my unit a few years ago and the motor and the shaft do come out you just need to push harder. Perhaps the o ring on the shaft is jammed up with old toothpaste.
instructions for taking the toothbrush apart are on the back of the charging unit - visible on the video too! - keep up the great vblog! best regards from denmark!
One of the coils is for the positive, one for the ngative half wave. It´s basically one big center tapped coil. It´s providing full rectification while saving two diodes. More copper - less diodes, whatever is cheaper :)
Might be arbitrary delay, selected by software developer. As timers are very limited on this MCU, they might have just selected maximum possible delay, and it's still 700Hz... (freq/1024 or something). Datasheet was saying that there is 10-bit timers, hence the number.
Items like this are what raise the overall noise floor so much in the average home. Many are not filtered well at all and have fairly high frequency switching frequencies. It's interesting to have a scanning receiver, do a sweep and then turn off the main breaker in your home and sweep in again to see how much quieter it has gotten.
I rather think that the pulsing the charger goes through may be designed to help keep the battery in good shape as this has been found to restore batteries that will not take/hold a charge anymore. But, that's just a guess.
I just happen to know how these toothbrushes actually work. For example: That pulse just tells to the mcu that it is in the charger, so it wouldn't turn pass MOSFET off. You could ask me some details more.
Possibly that was just a convenient frequency to use in the uC program? Maybe the timers have a maximum period dependant on the clock speed of the chip...something like that
I had to try out my hand-held scope: with a single full ground wire loop, it showed perfect sine waves, 81kHz, Vpp was 408mV. The amplitude decreases as the brush is lowered on the charger pin. no glitches. My charger is one with the UV sanitizer. Cool teardown!
I think this is a situation of Dave not reading the manual, and by manual, I mean the diagram you describe on the bottom of the base. No need for percussive maintenance.
I really liked the trick of using the probe as an inductively coupled current loop. Who knew a toothbrush could be that interesting? Were you ever able to get the precise post-mortem verdict? Was the battery simply past its max charge cycle limit?
Dave, the two coils and two rectifiers connected at cathode (common cathode) is used in some SMPS topologies where the secondary is center tapped to provide full-wave rectification with only two diodes instead of four.
I always wondered what was in my brush, woo 4-bit! It sounds like you came to the conclusion (but didn't say it) - I rekon that the two coils are wound in opposite directions to perform the full wave rectification. Also, whilst charging, you need to monitor the battery voltage, so I would imagine that it is being pulsed so that a reading can be taken at the ~800Hz to prevent over-charging. You may also find that the 'active' pulse shrinks as the battery voltage rises until it's just a trickle!
Could the pulse charging be to monitor the charging process? Charge, turn off and measure the voltage, charge some more, turn off and measure, and so on.
The reason for 2 coils is that each coil supply the circuit with half cycle via one of the diodes, there is saving in the components cost since the alternative would be a full bridge rectifier, standard bridge rectifiers are not suitable for this frequency, additionally since the induced voltage is very small and since there is a 0.7 volt drop across each diode therefore saving a diode in this configuration in compare with the full bridge can also save on power losses across diodes, as the charger is permanently plugged-in it makes sense to reduce the consumption to as little as possible.
Wow, this is brilliant, love how your talking about all the individual components and are also testing it with the oscilloscope. As somebody who is relitivly new to electronics and who has just had an oscilloscope donated, this is an invalable teaching aid :o)
This post ist really old, but it'll get some clicks anymore. I think there are two coils to pick up both halfwafes and route each over one diode only. So the drop is only about 0.7V, not 1.4V it would be by using only one coil an a diode bridge.
Why do I get the feeling Dave just set in motion a world record spike in the number of toothbrush chargers hooked up to an oscilloscope? Madness. Anyway, so when I tried it with my Phillips Sonicare, I was getting 82khz.
Awesome video, @ 28:00 I presume that's the uC disconnecting the battery from the charger so it can test the battery level - and disable the charge once complete. I've loved wireless toothbrushes forever - and understanding their circuit is awesome.
Renewable Oilcan Rim Living closer to Germany, we are told to brush for 3 minutes _minimum_, though that number is based on manual brushing, not motorized brushing.
LiPo's don't have a very long life; they only have a couple hundred charge/discharge cycles. Also, they can have to tendency to explode. Wouldn't want that in your hand first thing in the morning :)
Another great video by Dave. A very educational video that illustrated some interesting concepts new to me, particular the CRO probe mod. Keep up the good work and thanks for re-sparking the electronics within me.
On my Phillips, the magnets create the vibration...the vibration has a sweet spot, calibrating the correct distance gives the correct vibration, apparently the spacer in between the magnet had been worn and is gone, I also think I found a flaw in the metallurgy connecting leading to the head, I suspect the metal may have cracked.
I have a completely different model (2 actually) that refuses to either run or charge. I managed to successfully disassemble it non-destructively this weekend. I thought I was at the end, but you've given me a few ideas on what to check next. Thanks Dave! ps. I don't have a non-phone camera to post a reply/reaction, but rest assured I'll do my best Dave/DaveCAD impressions while working on it!
Using basic office translucent tape (the milky white kind, not the clear transparent) allows to see ship IDs sometimes better. Just stick a piece on top, massage it a bit and it may show the text without reflections etc. that make it often very hard to read.
Hi Dave, I know this is a rather old teardown but I have only just found it. I was researching to see if there was anything on the Braun toothbrush and whether mine was repairable. On mine the LED stopped flashing during charging but as you observed the problem could have been in the toothbrush itself or the charging base. I was fortunate in that I had an older model charging base from an earlier toothbrush which I knew was working at the time the toothbrush failed. When I fitted my current toothbrush to the old charging base, hey presto, the LED started to flash. After several hours charging the toothbrush definitely had more oomph in the motor so the fault was definitely in the charger base. I used a modellers razor saw to cut around the join line on the charger base and managed to lever off the base plate. As you surmised the innards definitely are potted. Nothwithstanding, I decided to pick away the encapsulant (which was a semi flexible white opaque type) and eventually I uncovered several components on the pcb. There were 4 large smd diodes and a single MELF diode plus an NPN smd transistor. All diodes and the transistor checked out ok but the MELF unit was probably connected across the B-E of the transistor since it gave me 0.6V in both directions. Unfortunately, I suspect that trying to remove the pcb from the encapsulant will end up doing a large amount of damage but it would be my guess that the charging coil has a broken connection and is open circuit. I may still go ahead and try removing the board to see what is on the other side but I was a little surprised that the charging base circuitry was a little more complex than I first imagined.
I think both of you got it wrong. AFAIK the disassembly is done by removing a metal ring from the top of the case - it kind of fixes the water seal to the shaft. If done properly the inner part slides out very easily. BTW: You missed a strange feature - the motor continously receives short pulses in off-mode (~1A for several µs at ~10Hz). Desoldering it stops the pulsing, but it won't turn on when put back together. Charging the device restarts the clicking and enables the toothbrush again.
So, I mean that maybe it is a protection against wasting energy in case of improper thing laying on the base module or broken short-circuited toothbrush-module. And interrupting of charge current is basically for single-way communication purpose.
I know it's an old video, but right next to that recycling diagram, there was another one showing you how to take it apart by pushing on the toothbrush end.
they wanted to show off? prevent interference from other sources? be high enough to not emit an audible whine? not enough interrupt timers left for the uC to do it on a more reasonable basis?
perhaps those ferrites near the base were cheap magnets, allowing the base cheaply beeing turned off via a reed contact. (i own a braun3757 which keeps getting considerably warm even with now brush hooked on. So there will be some 2-3 watts of standby waste...)
My experience with braun toothbrushes is that they sometimes get wet inside and electronics get corroded. Maybe that's what explains the two coils, some sort of failover mechanism...
Another fantastic episode. It would have been interesting to see how much power the charger was drawing from the mains while it was charging the battery, just to see how (in)efficient the coupling is.
emcgon Maybe that's why my brand new Oral-B has a wonky pushbutton under a redesigned rubber cover, while the unreadable battery compartment labeling hasn't been improved since the old design.
Ferrites and catalyst pellets are made using very similar processes (the only real difference being the material), so it would make sense that they appear similar.
are you sure those "ferrite beads" aren't actually a chemical pellet that is used to absorb hydrogen, in case the batteries vent a little? lots of aluminum flashlights (maglites, etc) have these pellets in case the batteries vent; prevents the gas building up, at least for awhile anyways.
wonder why they did not add a resonant capacitor to the coil, from my little experience with building and testing a simple ZVS oscillator (the one usually used to make flyback drivers) and making a wireless thing, I was able to greatly improve the output by adding the correct value one based on the inductance of the coil. In fact, it went from simply tickling a 10W LED with a few mA's, to completely destroying it once I added the MKP 0.1uF cap! (shoulda had some proper regulation. Oops!)
I bet the two coils are the same and here's why. You said it is full wave rectified and it is not. They have two half wave rectifiers and are feeding a sine wave in one and a cosine wave in the other to produce a highly inefficient full wave equivalent. You said some mysterious stuff about why they had two coils. They were just avoiding the $ of 2 diodes at the cost of efficiency. Right?
Jeremy Adair Almost. It's sine and minus sine, 180° difference, not 90° . In addition to saving the cost of 2 diodes, it also saves one diode drop and the associated power loss. Economically, the extra coil might be at least as expensive as two extra diodes, but saving one diode drop is significant in such a low voltage circuit.
What you can do is brush some talcum powder on these ic and then wipe it with your finger that will wipe off the powder from the surface leaving the particals stuck in the indentation and then youll be able to read it easily
You half the voltage drop. Its a classical center point circuit. You need only 2 diodes and the current flows always through only one diode. Cheap and efficient! U_center_point_circuit = SQR(2*U_eff) - U_diode U_full_bridge_circuit = SQR(2*U_eff) - 2*U_diode
I don't know whats inside baseunit, but my guess is that it's a way mobile-unit uses to tell to baseunit "hey, I am proper current consumer, and I am charging". Base unit monitors RMS current profile via primary coil and if it doesn't match predefined profile (e.g. RMS current is constant and doesn't going low 800 times per sec), then base unit decides: oups, something is wrong inside mobile-module, or may be it's even not a module at all, but someone just put metal ring upon basemodule inductor
also I was wondering if they made a voltage doubler with a capacitor and some diodes to run the micro when it was running off the 1.2V battery. 1.2V is not enough to power the LED (if it lights up when you're using it) or the FET's gate that runs the motor. As to charging, the resistance of the coils is probably plenty to limit charging current (they seemed to be really fine wire).
They wanted a full wave rectifier with only half the diode drop of a rectifier bridge. The ferrites add some more metal to the flux around the coils to increase flux linkage and thus efficiency,
Most likely some kind of compromise of Cpu power use and timers and A/D conversion. The system can't handle much more information with just 32 khz CPU (4bit) doing comparisions. And running it any higher than mandatory would be waste of power..
I read my comment and it really wasn't clear as to what I meant. At around 7m56s you cover its effects, but from the progression of the video I infer the testing was done by the probe RF loop. I want to see, and probably others too, is the effects of it when probing the circuit charging/no charging. Maybe current waveforms? Thanks.
Dave, I don't think you covered the portion on how the ferrite beads affect the input waveform. Maybe you could do it? Would be nice to see if the ferrite beads was just engineering bluff or the real deal.
Dave..ur the only I have ever come across who can turn an oscilloscope probe into a transformer to pick up high frequency signals..wow..download all of the experiences u have come across thru ur time in the industry, onto a computer..put it up for sale and watch ur self make millions (with me being ur first customer :P)...just fabulous! keep it up !!!
Braun was part of the Gillette company until 2005. Braun is now a wholly owned subsidiary of Procter & Gamble. I'm living near the german headquarters and know two engineers worked there. Better call it "engineered in germany" cause it's definitely build elsewhere. I think not a single part of it is made in germany anymore.
I did this exact same experiment with the same toothbrush(the toothbrush was a working model) and when I put the brush on the charger the amplitude reduced unlike yours
It's just a coil and a magnet -- a transducer, or 'speaker' more than a motor. It relies on the resonant frequency of the armature with brush head so that it can just input a little pulse once per cycle to keep the oscillation going. On the phillips sonicare, it's tuned to (I think) middle C or 'A 440' (I forget which), so that it's aesthetically pleasing and won't sound awful or 'sick' compared to most music playing in the background.