Hacker News Reader

Comment by rishabhsagar | Hacker News Reader

The explanation of “This is to protect batteries from sub par quality chargers” is not particularly convincing to me. As a user, wouldn’t that be my call? Shouldn’t providing a safe charging be a charger manufacturer’s responsibility? Why is Pixel deciding to not accept charging at 10w from ‘unknown’ charging device? At best I can see the argument for not supporting charging from insecurely designed charger on the account of warranty, but even then I think it should be consumers call as to what device they are using to charge their phones and phone should at most warn me about unsafe chargers, rather than outright refusing to charge at full potential speed.

View on HN

Shouldn’t providing a safe charging be a charger manufacturer’s responsibility?

Have you seen the manufacturing quality of some of the chargers out there? I wouldn't be surprised if this wasn't an effort by Google to prevent a long series of "Google phones explode into flames!" clickbait headlines when ultra-cheap knock-off charging pads start showing up at flea markets.

What makes you think Google chargers are any better ?

I would be surprised if there wasn't a single factory providing the same chargers with different stickers for at least a dozen of brands, including Google.

Ken Shirriff's charger teardowns show that there can be vast differences in quality and safety between different brands: http://www.righto.com/2012/10/a-dozen-usb-chargers-in-lab-ap...

Also check out Lygte Info[1], the site of a Dane who checks chargers and batteries usually bought from the Chinese sites like Banggood and AliExpress. Spoiler: Not all of them are bad.

[1] https://lygte-info.dk/info/indexUSB%20UK.html

This is pure FUD. I don't know how you think wireless charging works, but it doesn't send current directly into the battery. The charge is received by an antenna in the phone, and the same protections that protect it from overcharging when you plug it into an outlet apply to wireless charging too. You just replaced a cord with the antenna.

If you can damage the phone that way, then the phone is defective.

If I were being generous, I'd say that there are parts of the charging process (variations in charge, temperature, etc) that the hardware can't monitor, but have to be kept in certain ranges to maintain good lifetime/safety for the battery. So, instead of adding the sensors, they're restricting the charging devices. I can imagine that wireless charging is a bit complicated (see apple's problems) and it's easy for other vendors to get it quite wrong.

If that were the case, I think that they should include the stand with the phone.

Otherwise I think it's a bitter-tasting money grab.

>Otherwise I think it's a bitter-tasting money grab.

It's engineering laziness. Designing a phone robust against bad chargers is harder than designing a phone to work with one charger and putting DRM on it. It's the kind of thing markets can't do very much to solve which requires regulation, unfortunately, because there are too many bad actors.

i get that this is theoretically inconvenient for you, but how big of a problem is this really? afaik, most people don't wirelessly charge their phones, even if they have capable devices, and the pixel line tends to be pretty low volume to begin with.

given the small number of people that actually care and the possibly large cost of supporting full speed charging on a wider range of chargers, I'm not sure this is a problem at all, let alone one that merits new regulation.

If you assume faulty charger, you have to figure out a way of dropping at least 5 extra watts of heat somewhere in your phone. That seems practically impossible without significantly compromising on mechanical properties of your device.

All Qi chargers already have a temperature sensor in them and any modern phone has many temperature sensors (MEMS sensors, SMPS, lipo battery charger). The Qi receiver already communicates back to the transmitter and can tell it to increase or decrease power. Having the Pixel 3 decrease charging rate if it noticed heating up would be trivial.

You're forgetting the problem of the charging stand being implemented incorrectly & catching fire. Phone sensors won't help you there & while you may think all Qi chargers would have a temperature sensor in them, the fact that something simpler like knock-off USB charging blocks cause fire hazards leads me to not be so bullish on the Qi front.

And hey, look.

https://www.youtube.com/watch?v=CEGlmQS692w https://www.youtube.com/watch?v=IxstYrJQkk8

I'm sure you can find more examples. Just because we'd like the world to work a certain way doesn't mean it actually does in practice & the job of engineers sometimes is to deal with the reality of the world.

The issues you raise aren't solved by limiting transmit power - they are only solved by actually using Qi standard chargers. If you are using a faulty transmitter it can catch fire at 5W or 10W.

The Qi standard as of 1.1 has the receiver communicate back to the transmitter the amount of power it has received. The transmitter then calculates transmit efficiency based on the amount of power it is outputting. If the efficiency is too low the transmitter will fault with the assumption that it could be heating up a foreign object.

When the flea market charging pad that the Google phone is laying on goes up in flames, Jenny Cheapskate isn't going to know if it was the pad that burned down her home, or the phone.

Your logic allows me to lock up all hardware in an Applesqe way and get praised for my altruism as I pocket hundreds of millions in extra peripheral sales.

The world has gotten this far with standards that allow equipment from different manufacturers to function together in a system. What a shame if marketing people and greed destroys that ecosystem.

Hundreds of millions from the sale of Pixel 3 Qi chargers seems wildly optimistic.

...now. How about in 5 years? Will we still be plugging the wires to our portable devices?

yes.

I certainly hope so.

if you want fast charging, yes.

That shit could still happen. They aren't preventing the chargers from working, just not working as well.

That's not how standards work. If a cable or charging block is out-of-spec it can easily pose HW damage risks or fire hazards. It's impossible to consider every possible way this stuff can be broken.

Take for example https://gizmodo.com/a-google-engineer-is-publicly-shaming-cr...

OnePlus USB-C cables work fine with OnePlus devices but can damage any other USB-C device.

Another example where a USB-C cable destroyed his USB PD analyzer & his chromebook ports: https://www.amazon.com/review/R2XDBFUD9CTN2R

This was because the cable was completely miswired.

There's also numerous reports out there of how crappy charging blocks are fire hazards if driven at the full level they advertise.

https://globalnews.ca/news/3365247/electronic-charging-devic...

It's not totally unreasonable for a responsible manufacturer to either outright ban charging from unknown blocks/cables or to fallback to a trickle charge that disabled the highest-speed charging.

Since there's so many the only way to do this is to maintain a whitelist of good chargers at the expense of fast charge not working with an arbitrary number of chargers. You could think "well, just warn the user & give them a choice." However, the majority of users would just learn that most of the time it's OK & just hit "OK" blindly even when connecting to new chargers they don't know about. Also the news reporting would still be the same & wouldn't capture the nuance of using a third-party charger since news cycles are more instant & don't allow the necessary amount of time for engineers to receive the unit & perform diagnostics to figure out what happened.

> Another example where a USB-C cable destroyed his USB PD analyzer & his chromebook ports

That sounds to me more like a shitty design on both of those.

Short of sending hundreds of volts and causing arcing, this should be preventable, should it not?

Non-shitty devices will behave well, but shitty ones won't. That's the problem.

That cable supplied voltage on the ground line. Quite some WTF factor on that, but also difficult to protect against.

You can protect against reverse voltage, ie, voltage from the ground line.

A simple diode is the easiest way, using various power MOSFETs can reduce the voltage drop induced and you can even build a circuit that allows charging the battery and supplying power on the same connectors while also protecting against reverse voltage.

It's neither hard nor expensive.

TIL.

Are you saying that it's possible to protect against all kinds of shenanigans on the ground line even if the device has no other ground, or only against negative voltage?

There is only reverse voltage and over voltage.

Ground is basically your reference, if your ground has potential above your voltage input, you're in reverse voltage, if your ground is too far below your voltage input, then you're in over voltage.

Both can be detected and prevented from causing damage via various means on both sides of the usb plug.

Overvoltage can usually be solved by having a Zener diode within acceptable overvoltage short the thing to ground and blow out a fuse, this will cut the circuit and prevent further damage.

I recommend this video from GreatScott: https://www.youtube.com/watch?v=7Tk5ghH_U2s

>https://gizmodo.com/a-google-engineer-is-publicly-shaming-cr...

That's exactly what I thought of when reading the GP. There is some absolute garbage out there being sold.

There are also manufacturers using the existence of such garbage as a scapegoat.

For instance, the fitbit ionic battery plague is caused by shorted MLCCs -- but they still try to blame it on 3rd party chargers every time they get an RMA.

Some devices will break when you draw advertised levels of power. That's just a fact of life, and isn't something you should try to fix with DRM. It's not a very common thing.

It also doesn't damage the device itself.

The backwards cable is worth noting but not relevant to a system that's fundamentally AC.

Outside of some kind of massive EMP, nothing a wireless charger does should be able to hurt your battery.

That’s why circuit designers almost always create charging and power circuits with over voltage and current protections into fragile electronics. Certainly on a system as fragile and complex as a cell phone. This isn’t an excuse, this is bucking trends of circuit design that has existed for decades. And it’s even easier to prevent in wireless transmissions than physical ones because wireless requires careful tuning of resonance to work efficiently. The circuit to prevent this would be simpler than this handshaking one to detect “ok” chargers.

Actually in some cases you can. E.g. Gear S3 uses proprietary charger that is rated at 0.7A, 5V at the input. And several people reported watch raising overheat alarm when charging with select 3rd party chargers (not all). Apparently Samsung made it so the watch charges will really weak power, slowly and using powerful 3rd party chargers may push it out of the rated parameters. No permanent damage reported afaik.

From the Qi Wikipedia article:

> Rather than down-regulate the charging voltage in the device, Qi chargers meeting the A2 reference use a PID (proportional-integral-derivative) controller to modulate the delivered power according to the primary cell voltage.

So... a defective charger could plausibly damage the phone unless the phone contained a suitable protection circuit. Such a circuit could be quite simple - a transistor set up such that switching it would take the resonant receiver out of resonance would do the trick. The designer would need a bit of care to ensure that switching the transistor under wouldn’t fry it.

But IMO this design is a bit nutty. Phones already have voltage converters. Why not just use them?

Even old Nokia phones were able to test the maximum current drawable from the charger connected to their charging port, and adjust charging current accordingly. A cutting edge smartphone which cannot manage its battery charging process sounds a bit strange.

AFAIK Qi is a standard which can communicate too, so things look stranger...

For the uninitiated, Nokia phones had a feature called net-monitor which acted as a global debug console and had 60-something pages. One of these pages were battery and charging status and data.

If you can see bad quality, then it's up to you to decide if it's worth to risk.

Not that I entirely find it convincing, but a reasonable counter-argument is the Samsung-airplane fire situation.

If a model/brand becomes known for catching on fire, that entire model/brand can be easily restricted from things like planes.

So bad chargers and replacement batteries are a potential threat to the reputation of a brand... that said, the problem with Samsung batteries catching on fire was all Samsung's fault.

The Samsung Airplane Fire was most certainly not related to a charger, that was due to the faulty battery. The Note 7 deserved that reputation.

A more reason comparison would be the Nintendo Switch and 3rd party docks. It turns out the Nintendo Switch does not implement USB-PD to spec, and a 3rd party dock fried the IC for the USB charger, bricking the device.

From what I've been told on here, Nintendo did implement USB-PD correctly, and correctly identified itself as a low-current device, but third-party devices ignore that and attempt to draw full amperage from the charger, which causes it to shut off.

Source: https://twitter.com/marcan42/status/845368239622307840

Nintendo most assuredly did not implement USB-PD correctly.

At a minimum, the Switch 1) enters the proprietary AltMode before even querying the attached charger/dock to see if it supports it 2) tries DR_SWAP even after the dock says that it doesn't have dual role capability 3) always requests 0.5A before requesting full amperage, causing problems with to-spec SRC_CAP readvertisements 4) does not properly use the CAP_MISMATCH flag required by the USB-PD spec, 5) the dock violates Power Rules compliance by not passing through the entire USB-PD advertisements from the attached charger, 6) the switch and dock both leave excess capacitance on Vbus, but the switch is a dual role power device, causing problems with safe and compliant hardware that correctly checks for 0V before swapping power, 7) far from being oversupplied by poor chargers, the switch hogs the entire 3.0A from the adapter when its maximum draw should be 2.6A 8) as noted in your twitter link, the switch PSU and dock hard-crash and require a power cycle to start working again, whereas correct, compliant behavior would be to negotiate a 5V/0.1A error signal, etc etc etc.

Source: https://plus.google.com/102612254593917101378/posts/2CUPZ5yV...

It's not the charger that has all the issues.

That was entirely Samsung's fault, made worse by the fact that Samsung didn't allow customers to change the batteries in its phones.

The reality is that bad chargers exist. It is wholly possible to build a phone that is resilient to a wide range of wrong input. If your phone breaks because of a substandard charger it's your phone's fault.

I agree. I find it a bit disturbing that people are sort of quick to defend this. I don't see this as being any different, other than scale, from what John Deere is doing.

> I think it should be consumers call as to what device they are using

If the sketchy charger bricks the phone but the customer is dishonest about using the sketchy charger at time of RMA, do you expect Google to engage in he-said-phone-said arguments with customers? "Ah but your big-brother device says it was plugged into a toaster oven last week."

If a sketchy charger can brick the phone, I'd consider the phone broken by design.

So did Google? Not trying to be trite here, but that's essentially exactly the point. They didn't want to let a sketchy charger brick the phone, so they disallowed sketchy chargers.

You could say that the phone should try to accept any charge and never break, but that's a nontrivial amount of electrical-engineering and expense involved in accepting basically any wattage/voltage that a sketchy charger may produce.

Not to mention that the sketchy charger can brick the phone now. The phone can just regulate its own power supply and charge however it likes, all the charger does is supply current to the coil.

Not to mention the potential liability if a crappy charger overcharges the battery and causes a fire in someone's pocket.

all of the battery charging control is in the battery, not the charger. The charger just generates the EM field (in this case) or a supply on traditional cables.

In a traditional wired charger, the transformer coils are in the charger. In case of wireless charging, one half of the coil is inside your phone. I am not an electrical engineer, but whatever current is induced there, you have to put it somewhere. So the argument makes some sense. Still not a nice move.

A transformer is a two part system of two inductive coils. If you open the circuit on one end, the closed end won’t be able to induce a current into the second coil. It’s as simple as a single MOSFET controlled by some power management IC that measures current to prevent this from happening. It’s easier to prevent damage from wireless charging than it is from a physical connection.

> whatever current is induced there, you have to put it somewhere.

No you don't. What is induced is an electromotive force, e.m.f., that is voltage. If the coil is open circuit then no current will flow. It's slightly more complicated than that because wireless charging uses tuned circuits not simple iron cored transformer but the general principal is the same and the control circuit in the phone can simply disconnect the coil completely in order to not accept any power.

> As a user, wouldn’t that be my call?

Maybe they're trying to keep their warranty repair costs in check. Maybe they would have allow it after your warranty expires and you're on the hook for the new battery instead of them.

>As a user, wouldn’t that be my call?

Sure. But it's the manufacturer's prerogative to build a device they want to build. Also this design decision is not to enforce copyright protection or is based on some other anti-consumer design, but rather is done for quality reasons - and that's a fair choice for them to make.

>Shouldn’t providing a safe charging be a charger manufacturer’s responsibility?

'safe' is ambiguous in this context. The charger may be 'safe' in that it won't catch fire and kill anybody, but may still prematurely cut your batteries life.

If your phone is under warranty, you'd potentially also need to opt in to the phone tracking how often you've used an unsafe charger. And for how long. And it would, reasonably, void the warranty.

A valid response is for consumers to say they don't want warranties any more. But if you do want a warranty, you need to use your device with approved devices...

Should be your call but laws open them up to liability so they don't allow you to just use anything you want. When something goes wrong, it'd be the phone that explodes and the HTC/Google part of the two companies has deeper pockets.

> Shouldn’t providing a safe charging be a charger manufacturer’s responsibility?

There is no such responsibility if they are overseas and shipping directly to the consumer.

> it should be consumers call

You should run that line of reasoning by your local fire marshal, and then also your insurance agent.

Maybe a reasonable workaround would be for Google to do the same thing Apple does for some of their products and sell licenses to third party manufacturers of repute.

Reply on news.ycombinator.com