What is more fun than plugging in your phone and coming back to find your battery on empty? Stepping on a LEGO block with bare feet or arriving hungry at a restaurant after closing probably qualify. [Alex Sidorenko] won’t clean your floors or order you a pizza, but he can help you understand why cheap chargers won’t always power expensive devices. He also shows how to build an adapter to make them work despite themselves.
The cheapest smart device chargers take electricity from your home or car and convert it to five volts of direct current. That voltage sits on the power rails of a USB socket until you plug in a cable. If you’re fortunate, you might get a measly fuse.
Smart device manufacturers don’t make money when you buy an off-brand charger, and they can’t speak to the current protection of them, so they started to add features on their own chargers to protect their components and profit margins. In the case of dedicated chargers, a simple resistor across the data lines tells your phone it is acceptable power. Other devices are more finicky, but [Alex Sidorenko] shows how they work and provides Eagle files to build whatever flavor you want. Just be positive that your power supply is worthy of the reliability these boards promise to the device.
Now you know why connecting a homemade benchtop power supply to a USB cable seems good on paper but doesn’t always get the job done. Always be safe when you make your own power supplies.
This is part of the way there. How about a universal one that tests the source to see what it can provide and checks the device to see what it expects, then automatically configures itself to provide the maximum current the device can safely pull from the source?
That’s a good idea, you might just want to make sure that kind of adapter doesn’t cost as much as a decent charger.
I put a 3 amp USB jack in my truck but the molded on port doesn’t even have the data lines shorted and only the power lines run out of the molding around the connector. So a gizmo that plugs into it with a type A on the other end, with the capability to provide as much power as any tricky device wants would be very nice.
I don’t see such a thing costing the nearly $40 Samsung wants for a Galaxy Tab charger. I have several other name brand chargers able to output plenty of power for Samsung and Apple devices, but they’ll only see those chargers as an ‘open’ data type port from which they’ll only draw 500ma.
That’s exactly what USB-C Power Delivery does. It can also raise the voltage to deliver more power (power = voltage x current, but current causes heating, so…). Unfortunately the extra electronics to do the negotiation costs money, so USB-C chargers cost more. But ut’s a standard, so over time competition should lower the cist.
Hey that’s pretty cool, I didn’t know about the variable voltage. Looking forward to a wider adoption of that standard, and hoping it doesn’t go the way of betamax in that it’s technically superior but never takes off.
The only issue I can see is an improperly implemented system…
The charger accidentally sending too high a voltage for devices that don’t tolerate accidentals (Think of the hard 7V absolute max of the AXP288 found in many Intel tablet PCs) when the charger misplaces 19V onto a 5V (7V Absolute max) rail…
Or a virus on the phone/tablet-PC tricks the charging circuit to request 19V from a USB C PSU when the charger circuit can only handle 5 to 6V safely.
Such a hard-bricking virus deserves to be coined: WannaFRY
Holy shit, that virus idea is fucking genius. A piece of malware that will cook your computer beyond all redemption! You may not be the hero we wanted, but you’re the one we all deserved ;)
It would be even better if the virus had a way to spread through the usb protocol and had a delay timer. That way if you get the virus on your phone it will deliver it to the next 2-3 chargers it uses before destroying itself.
Extra points if the virus display the ginger from Futurama.
Start here, http://blog.deconinck.info/post/2017/08/09/Turning-a-Quick-Charge-3.0-charger-into-a-variable-voltage-power-supply — good overview of Quick Charge 2/3 plus GitHub code for making your own microcontroller controlled PSU :)
it’s not just USB-C. My 3-year old Droid Turbo came with Motorola’s “Turbo charger” that can put out up to 12V to fast charging. Works great. Stock battery lasted three years – just replaced it back in February.
It’s called USB Power Delivery.
Prior to that, the simpler standard was USB Battery Charging 1.2 – which nearly everyone except for Apple complied with. Signaling a dumb charger was easy – the source was to short D+ and D- together to indicate the presence of a dumb charger that could source >500 mA.
Signaling a high current “smart” source was much harder, which is why you almost never saw CDPs or devices that would charge at high current from them until USB PD + USB Type C came along.
So… you’ve got a voltage and current meter, and a thermal camera to monitor the PSU to check if it looks like it’s going to catch fire…
Not sure how this is cheaper than a decent charger?!
There seem to be plenty of charging device ICs out there, but few for the device being charged. Am I missing something there? It seems the charged side also needs to communicate with the charger properly.
The category includes plenty of devices that only handle older USB power(or that just expect a dumb 5v or 3.3v and are intended to drive some low voltage bit of logic in a larger system); but PMICs would seem to be the class of devices you are looking for: since integration is prized for size and cost reasons it’s quite common for the ones designed and sold for phones and similar devices to support USB behavior.
Supporting particularly high wattages(as well as accommodating all the various alt modes) are more niche.
Why doesn’t it work? Because he has the DM1 and DP1 pins swapped around. M=minus, P=plus.
while this is a good catch, this is actually valid, as per datasheet: http://www.ti.com/lit/ds/symlink/tps2514.pdf When signals are reversed, the IC imitates a 5W (1A) charger; when the signals are connected the right way, it imitates a 10W charger.
Also, this circuit actually works, read the article.
This article is not about bad USB power supplies. It is about phone manufacturers bending backwards twice to violate any industry de facto standard! Obviously signalling is necessary between the power supply and the charger (which is inside the phone? This is Hackaday, a charger in a phone is neither the USB cable or the switch mode psu you plug in the wall!!) when you want to exceed USB specs, i.e. push more than 5.5V through the USB cable to fast charge your phone, but there should be no backward compatibility issue with plugging in a standard USB power supply, if the phone manufacturer mentions the connector on the phone as “USB” in the specs. There are strings attatched to that name and they are at least falsely advertising their phone!
The article and those of you mentioning in the comments that the phone needs to “know” the current capability of the USB charger do not know how electricity works: If the charger (inside the phone) monitors supply voltage, which it does, it will know when it draws too much current as the 5V supply will sag when the power supply is asked for more than it can deliver. Laptop PSUs have worked like that for decades and been fully capable of determining if a 90W or a 120W brick was connected without anything else than the two power wires. This is about trade wars, not EE.
Absolutely. A decent switching power supply is unbelievably simple and cheap these days, and charging circuitry is plenty smart enough to figure them out. Phone manufacturers that pull this crap just want to mark up a DRM-infected wall wart 2000%.
I’ve tested a bunch of the laptop power-bricks and found some of them will try to deliver the same voltage until something overheats or the primary side current is too high and shuts down.
Some of the dell and HP PSUs do this…
The HP PSUs send an output of an error amplifier to the laptop so if the PSU starts to fail then the laptop can automatically re-manage resources, i.e. charge slower, slow the CPU, etc.
The dell ones have an ID chip in them so the system can automatically set up different charge and speed modes… but without pending-PSU-doom detection.
For the latitude E6400 ID:
less than 60W – Slow down the system and complain,
60W – standard charge or throttling if fast-charge enabled,
90W and up – No power throttling whatsoever (thermal throttling* still happens)
.
*To disable this on some dell laptops:
hold FN+SHIFT whilst typing: 15324
then hold FN+R, adjust settings in this SMBIOS thermal-diags page.
Crashes the Linux kernel (known on 3.x and on 4.10 to crash so far) requiring pre-boot adjustment (i.e. whilst in GRUB) and may leave artifacts on M$ Windows. Doesn’t work on core i-series systems apparently.
In fact, it is not… It is an standard: the USB Power Delivery.
http://www.usb.org/developers/powerdelivery/
Ooops… I should read the article first and comment after… Sorry.
There is also the ID pin USB micro connectors.
you’ve added an important comment. how well do phones charge other phones? OTG is wonderful, and manufacturers are wise to this. ‘Standard reverse charging’ and ‘Rapid reverse charging’ are options with the phone i’m typing on now.
Is this still true? To me it seems that most devices made in past few years happily charge according to USB Battery Charging specification, i.e. just short the data lines together and it will assume the source does current limiting.
True for everyone except Apple (they have never complied with the standard in any way, shape, or form) and Samsung (their tablets required a “little extra” in such a way that Samsung tablet chargers were detected just fine by devices that wanted BC1.2 supplies, but Samsung tablets would not charge from a basic BC1.2 supply).
Well first i would check that there isnt any debris in either the Device socket or the cable plug. From my experience (YMMV) all of my devices have charged when using any of the little bricks that have been provided with the phones. Ive used LG chargers with blackberries and samsung chargers with a nexus. They might not be able to activate the fast charging stage but they have all been able to drink up some juice. If they haven’t it was usually due to a worn out plug or socket or debris in either one.
Sometimes when I plug my SwampScum Galaxy S5 into a weak charger, the phone wakes up repeatedly to let me know that it is “charging”. The end result is the phone uses more power to wake up than the charger is supplying and the battery slowly depletes!
It’s not that simple anymore!
First, here’s another source for you. http://www.ladyada.net/make/mintyboost/icharge.html
That was where I first learned about 5V USB charged devices using resistors across the data pins to negotiate current.
But.. .that’s not how the new chargers work.
Here’s the problem, we have faster phones with more accessories now. And of course we want the phone to work while charging. Also, we want bigger batteries that hold more energy. And we want this all to charge quicker, not slower. So… the amount of power had to increase.
That’s why we needed these non-standard schemes in the first place. Our devices had to start drawing more current than a standard USB port allows for. But.. nobody wants to blow a fuse on their PC if they try to plug their phone in. So, devices had to be smart about how much current they will draw. That can be done with USB enumeration but that requires a more expensive chip. The USB standard did have something in it about allowing a certain amount of current to a charging device if the data pins are shorted but that wasn’t enough. So.. we had this mess of different brands and different schemes.
But that was years ago!
Power demands kept rising. The USB standard was 5V fixed. So the only way was to increase the current. After a point that gets hard. Especially when everyone wants it to work through the resistance of some cheap thin substandard USB cable bought off the usual online sources or some shady shelf in a gas station.
So.. the voltage had to increase. That’s where the “Quick Chargers” come in. They still default to 5V but when a compatible device is plugged into them they actually negotiate about the voltage. Some can go as high as 35V. Are these proprietary, non-standard and mutually incompatible? Not exactly. Actually the situation is improved from the old 5V and resistors across the data line days. Every manufacturer had their own resistor values or voltages on the data lines. With ‘quick charging’ at first everyone used the same chips from Qualcom. Sure, they are patented. You cannot legally make your own compatible device without buying their chip. But.. whatever brand name is on your charger, if it’s a Qualcom chip inside and a Qualcom chip in your phone or other device it will work. As QC came out with new versions they even made them backwards compatible. It wasn’t exactly open source but it’s better than what came before.
Now we have a second ‘quick charge’ (negotiable voltage) standard. Now we have USB-C. Yet another standard. Does one have to pay for a license to make something USB-C compatible? I don’t know. I bet people here do know though. Anyway.. I guess that’s better because now you might get quick charging by plugging into your desktop or laptop. That’s something qualcom was never going to do.
Anyway… what this means. That TPS2514 chip is great for automatically detecting the best combination of resistances and voltages on the data lines to get the most current into an older device. Does this still work with devices with Qualcom chips and/or USB-C? Will it work on future devices?
Even if it does work it will be limited to 5V which means ‘slow charging’. On the other hand, charging slower might be better for the battery. Also, something I used to want to build is a big 5V power supply with a bunch of USB ports connected to these chips. That way I could have one charging station charging all the devices in a room. That would be more energy efficient than loading every outlet up with a wallwart right?
“Stepping on a LEGO block with bare feet”
Better yet stepping in a upside down 8 Pin DIP IC with bare feet.
Seen these circuits before, never worked for me?
I was actually thinking of implementing this in reverse so to speak…to slow down a charge. Reading BatteryUniversity docs on LiIon batts lifespan which can have tremendously increased lifespan if charged and discharged slowly as in sub rated values. We cant affect the discharge side much as the phone or tablet consumes that but charging can be slowed down at least for our plug it in and go to sleep routines. No sense in fast charging a device only to have it fully charged for many hours unneeded better to trickle it in so it is fully charged just in time for your wake up and thus avoid many shonky battery replacements of dubious quality in its lifespan. The thing is no one makes such a feature in their chargers. The big multi port bricks are quite powerful but slow charging is not one of their modes unless the device asks for it.
!!! I like this guy.