[0:00] Who wants to see a magic trick?
[0:02] I have a run-of-the-mill standard Raspberry Pi Zero W2, and I have a TFT display with
[0:07] a touch controller. It’s pretty easy to connect the two, just plug them together. You will
[0:12] however note that I have no battery and I’ve not connected it to any power source. But
[0:17] what I do have is my nice Pi branded notepad.
[0:21] Well what do you know, it’s powered itself up. And if we give it a few seconds, it even
[0:26] boots into X-Windows. Now of course the big question is, does it run Doom?
[0:37] And there we go, it runs Doom pretty well.
[0:40] So what’s going on? What is this magic? Well fans of the channel will recognise these
[0:44] wireless LEDs from previous videos. We can even power one of my wireless Christmas trees.
[0:49] And it works with my wife’s phone. This was a slightly nervous moment, blowing up your
[0:54] wife’s phone is not a good thing. By now you will have guessed it. We are playing around
[0:58] with some more wireless charging technology. I’ve been looking at a few wireless power
[1:02] options over the past year or so, but I’ve always found them to be slightly disappointing.
[1:07] They work well for simple stuff like LEDs, but I’ve struggled to use them for anything
[1:11] practical. So I thought why not try a proper version.
[1:14] As you’ve seen, this one even works with your phone. Let’s have a look at what it’s
[1:18] actually capable of. I’ve hooked up the transmitter PCB to my USB power meter, so we can monitor
[1:24] one to the power going in. And I’ve got the receiver hooked up to my electronic load.
[1:28] With the USB supply in 5 volts, we can get up to almost one amp on the output before
[1:33] it collapses. The transmitter and receiver are both compatible
[1:36] with the iPhone wireless charging standard, so we can give it a go with my phone charger.
[1:41] Interestingly this is using 9 volts for the transmitter side of things.
[1:45] With this we can get up to about 1.2 amps, that’s pretty impressive stuff.
[1:50] From the docs, my transmitter board should be able to go up to 12 volts on its input.
[1:54] But when I tried this, the voltage on the receiver seemed to become quite unstable, and it would
[1:59] shut off at a relatively low load. I also noticed that the receiver PCB started
[2:03] to get quite warm. We’re getting up to around 80 degrees centigrade on the receiver board
[2:08] when we run the transmitter at 12 volts. If we run the transmitter at 5 or 9 volts,
[2:12] it only goes up to a much more comfortable 50 degrees.
[2:16] Interestingly, the transmitter board doesn’t seem too bothered by the supply voltage at
[2:20] all. There’s only a temperature increase of around 5 or 6 degrees when running at 12
[2:24] volts compared to 5 or 9 volts. Let’s see if we can work out how efficient
[2:28] it is. I’ve rerun the tests and noted down the power input and the power output.
[2:33] We’ve got some interesting graphs. With the transmitter running at 5 volts,
[2:36] we seem to hit peak efficiency of around 73% at a load of around 0.5 amps, and we can get
[2:42] up to around 0.9 amps before we lose regulation. When the transmitter is running at 9 volts,
[2:47] we get peak efficiency of just under 70% at a load of 1.1 amps, and we can pull up to
[2:52] 1.2 amps from the system. The boards are supposed to be able to transmit
[2:56] 1.5 amps, and I think this is actually probably doable under the right conditions. You would
[3:01] need to get good alignment of the coils and the perfect distance between them.
[3:05] I’m also using a rectangular receiver coil with a round transmitter. Now I’m not sure
[3:09] how much of an effect this has, but I’m sure it’s probably not ideal. There are a lot
[3:13] of alternatives on AliExpress, some of them with multiple coils, which should give much
[3:18] better results. I’ve also tried to measure the transmitter
[3:20] range. I’ve set up the load to draw 0.5 amps, and with a 5 volt input, we get around 1 centimeter
[3:26] of range. With the 9 volt input, this goes up to about
[3:29] 1.5 to 2 centimeters. That’s pretty good, and you would normally
[3:33] arrange things so that there’s not much between the transmitter and the receiver coils. The
[3:37] PCBs are remarkably simple. The receiver board has just one IC on it, along with some passive
[3:43] components. The CP2101, which should not be confused with the very common USB to UART
[3:48] bridge chip that we find on a lot of our dev boards, is doing all the heavy lifting. It’s
[3:52] pretty amazing how much is packed into this one chip.
[3:56] On the transmitter side, we have a couple of ICs. The JW7951C is the transmitter chip.
[4:02] This is responsible for driving the coil and provides information on what the coil can
[4:06] see. The other chip is a bit of a mystery. It’s probably some kind of microcontroller
[4:10] and power delivery control chip, but I can’t find any reference to it anywhere. If you
[4:14] know what it is, then let me know in the comments.
[4:17] Now while we’re on the subject of PCBs, I should probably tell you about PCBway who
[4:21] are sponsoring this video. Why not have a go at designing your own PCB and send it off
[4:25] to them for manufacture? They’ll even do SMD assembly for you.
[4:29] But back to the video. Now obviously, you’ve already seen this working and powering the
[4:33] Pi, but just how much power does the Pi need to boot up and play Doom?
[4:37] Well, let’s have a look. I’ve hooked it up to the USB power meter and it uses a surprising
[4:41] amount of power, getting up to almost 0.6 amps when running the Doom demo. If you’ve
[4:46] watched the previous video, then you remember that under heavy load the Pi Zero can draw
[4:50] around 400mA, so our screen is adding quite a bit of load to the system, probably mostly
[4:55] with the backlight. However, it is well within the capabilities of the wireless power system
[5:00] given what we’ve measured. But I found it surprisingly hard to get it to run consistently,
[5:04] and it would often brown out during booting, launching X windows or running Doom. It took
[5:08] quite a few tries to get a good end to end take.
[5:11] The main issue is that when you can’t see the coil, it’s quite hard to get good alignment.
[5:15] This does explain why a lot of wireless chargers have locating magnets to snap your phone into
[5:20] position. Without them, you get a pretty ineffective charging system. Something to think about for
[5:24] the future projects.
[5:26] Now what can we do with this? Obviously, we can’t just have our Raspberry Pi only working
[5:30] when it’s sat on top of the transmitting coil. That would be a bit rubbish.
[5:34] So let’s make a complete wireless charging system with a battery. I’ve got a bunch of
[5:37] these basic boards that are based around the 4056 charge controller I see. We could hook
[5:42] one of these up with the battery and get it to charge. But I’ve also got these slightly
[5:46] nicer boards. These have the 4056 charge controller, and they have an adjustable boost converter
[5:52] which can output from 4.5 volts to 24 volts. In theory, this means that we should be able
[5:57] to run our Pi even when the battery voltage is quite low.
[6:01] So I’ve hooked it up to a battery, it’s measuring just over 4 volts. But we have a
[6:05] bit of a problem. The output voltage is too high. It’s 5.66 volts, and I can’t seem
[6:11] to adjust it any lower. We should be able to go down to 4.5 volts, but the trim pot is
[6:16] already at the minimum. The Pi 0 might be just about ok with this slightly higher voltage.
[6:22] Looking at the datasheet of the power management IC, it will take up to 5.5 volts, and we’re
[6:27] pretty close to that. But it would be nice to get exactly 5 volts. What’s going wrong
[6:31] with our PCB? If we look at the datasheet for the boost converter on our board, we can see
[6:36] that the voltage output is controlled by this resistor divider. The top of this divider
[6:40] is this resistor. This has the code 104, which should be 100 kOhms. Let’s disconnect the
[6:46] battery and measure what we’ve actually got in circuit.
[6:49] Well, for the top resistor, we’re measuring 57 kOhms. And for the trim pot, when it’s
[6:54] at its minimum, we’ve got 10 k. It’s not surprising that our output voltage is completely
[6:59] wrong. We need to either decrease the value of the bottom resistor, or increase the value
[7:03] of the top resistor. Fortunately, we do have these through holes where we can add some
[7:07] additional resistors. I’m going to put a 100 k resistor in parallel with the top resistor.
[7:12] This should give us a parallel resistance of around 78.5 k. So with that done, we just
[7:17] need to reconnect the battery. And after a bit of fiddling with the trim pot, we can get
[7:22] almost exactly 5 volts on the output. I have no idea what’s up with these resistors, but
[7:26] I’ve measured them on a bunch of the boards I’ve got, and they’re all completely wrong.
[7:30] Some of them are closer to 100 k, maybe within the 10% tolerance, but it’s pretty rubbish.
[7:35] Well, let’s see how much current we can get out of the step up converter. I’ve connected
[7:39] it up to the electronic load, and we can easily draw 1 amp without the voltage dropping. So
[7:44] if that’s not bad at all, this will easily power our Pi 0 even with the screen and under
[7:49] heavy load. So I’ve hooked the board up to the wireless power receiver, and it’s charging
[7:54] nicely. I can see the voltage on the battery slowly going up, so that’s working pretty
[7:58] well. And eventually, our battery is fully charged. Now if we hook the board up to our
[8:03] Raspberry Pi with a nice switch so we can turn it off and on, everything works as normal.
[8:07] We’ve got a fully wireless Raspberry Pi. We’ll give it a few seconds, and it boots
[8:12] into X-Windows, and we can even run DOOM without any problems. So I’ve been running
[8:17] it for a while, and I wanted to check just how hot the boost board was getting when running
[8:21] the Pi at full pelt. And it’s actually not too bad. The diode is getting a bit warm,
[8:26] but it’s definitely not hot, so that’s going to be pretty good for our future project.
[8:30] I did find some more boost boards in my stock of bits. These will boost up to 5 volts, and
[8:34] I’ve been thinking a bit about what I’ve built so far. It may be better to separate
[8:38] the boost circuit from the charge circuit. This would let us put a switch before the
[8:42] boost converter. The advantage of this is it will stop the boost circuit draining our
[8:46] battery even when we’re not using the Pi, but we’ll do that in the next video.
[8:49] Since we’ve been looking at Raspberry Pi’s, then check out this video. It’s using a Pi
[8:53] Zero for radar motion detection, face detection, and it’s got face recognition. It’s definitely
[8:58] worth a watch. And if you’re a fan of wireless power, then there’s a whole bunch of videos
[9:01] in the playlist which should be up on your screen right now.
[9:04] Next video, we’ll add sound and a controller to our Pi, and we’ll have a fully portable
[9:08] DOOM machine to play with.