QI Wireless Charger Charging Pad for Samsung Galaxy S7 S6 edge Note 5 Nokia HTC 8X Google Nexus 5 6 7 Micro USB QI Phone Charger (WHL #32)
Here’s a random thing – a wireless Qi standard charger for phones!
I don’t even have a phone that is Qi compatible, but at work we’re looking into wireless power distribution AND data transfer at the same time. So I figured I can afford 1,74€ at AliExpress (mid-October) and get such an unit for myself. You know, dissect it, learn how it works, modify it to change characteristics. I previously had a look into the Royer converter or Royer oscillator for an entirely different thing – and here’s where ends meet.
But first things first – here’s the unit:
(protective strips already peeled away). Aside from the badly glued silicone ring that is excellent at catching fluff – nothing much to see here. Size is 69mm x 9mm.
Same on the bottom, except the power ratings: 1500mA at 5V via Micro USB:
Wanna have a look at the inside? Just lift the lid, there’s three plastic fingers and one key pin that hold it in place:
Not much, is it? Well, that’s what you can expect for less than two dollars… 
As you can see, the LED placement is a precision job that only the very best workers at the Wan Hung Lo factory are able to do. On a side note: I’ve only ever seen this LED light up in red (constant, not blinking). It’s probably a status LED that changes when a Qi compatible device is detected.
The PCB is held down with a single screw and and oriented with another plastic pin. It just lifts up and then folds out of the case.
Aaaand the IC used is: Unknown. No label. Not scratched off, just…not labelled at all. The sidekick on the right is a good old LM358 dual op-amp, but that’s not doing the Qi protocol stuff. Guess there aren’t tons of ICs that can do this, and then one could work out the pin functions and compare it to the datasheets and their application examples…but why bother. At 2€, just buy one of the ready-made units.
So here’s the interesting part: As I said, I have no compatible phone. But I tried something else…
For recording the output, I’ve hooked up these 9 turns of wire to my oscilloscope:
Waveforms shown below. Unfortunately, for these the DS1054Z didn’t show frequencies. Reading the first image, that should be around 166 kHz (6µs per cycle), at 600mVpp. The signal comes in bursts of around 170ms in length, but I didn’t spot much data-like structure on the beginning. Maybe the receiving device needs to suck some power from the field in order to initiate the pairing process, who knows.
Now what if I told you that THIS contraption is working:
That’s a friggen 2008 vintage Braun Triumph 9000 toothbrush. Qi came out in 2008 as well, but back then, nobody cared. Especially Braun didn’t care, because they change charger infrastructure across their models and even for product refreshes – mine happens to be discontinued after this model, meaning people nowadays pay up to 25€ for a working charger, when the brush was 65€ delivered back in 2008. Well, toothbrush prices skyrocketed since then, but that’s another story.
Thing is: I always wanted to build an USB compatible charger, as one Christmas a few years ago, I went to visit family for two weeks, and I forgot the charger. And three different charger models in the family did not work for my toothbrush. Back then, I connected fresh batteries to the opened unit and charged it in that crude way – but the charging status is a total mess, and once the device dies from real undervoltage, it doesn’t come back with a full battery, but it needs to sense the charging station in order to reset. That’s a pretty shitty design if you ask me.
Anyway, the display is really hard to capture, as it not only has the usual glass substrate in front of it, but also the curved thick plastic from the waterproof outer shell of the device. Both show a huge amount of glare when the direction of the observer is the same as the direction of the light source – which happens to be the case when you use flash photography. Humans haven’t evolved to having glowing eyes yet, so readability is really good – but I cannot get a decent shot. Here’s the best try…
For comparison, with the original charger:
(“Aufladen” means “charging”, and it started from 1/4 blocks when I first set it onto the charger).
While I was sceptical if that really does work or no power is sent to the device, I seated it for one evening, and it really did charge. Not as fast as it would in the 2.2W charging cradle, but it did. The 2.2W charger was discontinued for some EU efficiency directive – it simply failed the standards because it works wirelessly, just like Qi doesn’t reach much over 50% P_in to P_out. Braun’s solution wasn’t to make the device better – they simply made it a slower charger. Below some threshold wattage, the efficiency requirement was no more, and now toothbrushes charge slower and most likely even more inefficient because of that move. But hey, that’s how the legal system works.
So, two more shots from the alternating power inrush, as already seen on the scope:
Peaks are up to 1.6W, so that’s not too shabby. The standby consumption between those bursts are as low as 0.03W, which did show up in a lot of photos.
The important question is now: WHY does it work?
Hooking it up the same way to the scope:
This yields the following wave on the display:
Well…that’s not the same. For starters, this is an continuous (shitty) sine wave instead of bursts. Then, it operates at 75 kHz instead of 166 kHz. But that’s about half of the Qi, and maybe the real coil and charging electronics luckily accept this as well. The overall signal strength is much lower at around 230mVpp, but that’s heavily affected by the distance of the receiving coil. And as my toothbrush has the coil high above the bottom plate (see the charging cradle that it needs to sit in instead of on), that shouldn’t be an issue.
So…it works 
For comparison, I’ve got another charger that does not work as already described. It has a very pleasant color, which is typically chosen by young girls nowadays. The associated toothbrush (a 30€ jobby) died because of faulty seals, which caused some water to enter the device, slowly corroding away the battery contacts and power related IC pins. Test setup is once again the same:
(notice the bolt in the middle that houses the tiny coil in this charger, which fits inside the coil of the brush, rather then the large coil from my white charging cradle that surrounds the coil of the brush)
Waveform:
22kHz, 160mVpp, continuous waveform, different again. I wonder why they lowered the frequency when having a smaller coil – but they might have had their reasons. And if they didn’t, they at least made it incompatible to other models, so that they annoy people for no reason.
So I wonder if I should keep this one as an emergency charger, or if I should modify the unit to produce a continuous waveform? Maybe the IC is hackable once identified, but spinning some homebrew with…I don’t know, a 555 timer, can also do the trick. And for 1.74€ – that’s a nice base unit to fiddle around. Maybe I should order another one before China shopping gets more expensive next year. And maybe someday I will own a phone that does support Qi charging
Oh, I almost forgot: Reichelt had a similar unit in their newsletter recently. “GOO 44169”, with three-coil (yet only 1A) design so that your fat tablet fits the plate. Looks the same to me – now discounted from 32.70€ to just 14.99€. Limited time offer – buy now! (or…buy in China, whatever floats your boat. Goobay does nothing more than import China stuff, pay VAT and add a few-hundred-percent profit margin)
