Syonyk

When you're trying to hit something moving 20kt, with something moving 30-35kt, from a few thousand yards, it doesn't take much error in estimating speed, heading, or distance, to make them miss. It's honestly more remarkable that they hit at all in those conditions, even with a "spread" (shooting several along slightly different headings hoping one or two will hit).

So, no, you've not worked with batteries, battery management systems, or anything of the sort.

Determining battery health is hard.

I'm guessing you've not dealt with power electronics and batteries terribly much.

Depending on what limit has been hit, it's quite likely there is no way to log the cause of the error. Low voltage protection circuitry on most batteries doesn't have a status line. It's never supposed to trigger except in exceptional cases, and it just cuts power. All you know is that the power disappeared suddenly, and you've rebooted. Telling the difference between that and assorted other hardware faults, especially if you never designed the hardware to look for it, is really difficult.

You can certainly design a system that will latch the cause of the shutdown in the battery management IC - but you can't really add this in after the fact.

I've suggested to a range of people that if their only complaint is runtime, and the phone is a few years old, getting someone to replace the battery is far cheaper than a new phone. It's a novel concept, and I'm quite unsure if people just don't know if that's a thing, or if that's the socially accepted excuse to spend a lot of money on a new phone.

I'm pretty sure that a lot of people are very happy with water resistance in exchange for having to do a bit more work to replace a battery (that they don't actually replace).

The number of people using their phones in the hot tub, or in the sauna, astounds me on a regular basis. I can't imagine doing that. But, with modern devices being genuinely "drop them in the pool" grade waterproof, neither does it seem likely to be a problem.

I'll agree on thinness, though. The number of phones in massive, chunky cases says "A lot of people don't care about thin."

Imagine the tradeoffs, though.

A "user-serviceable battery," by requirements, is going to be a hard shell plastic sort of thing - which means a decent fraction of the "total battery space" is a protective layer, not active cell components - so some significantly reduced capacity compared to having a "non-replaceable" battery ("slightly more difficult to replace"). You also end up having to devote space to whatever mechanisms keep the rear shell in place, and may have a harder time waterproofing it as a result (which seems to be standard anymore - the number of people I see at the gym using their phones in the hot tub or sauna is boggling).

Batteries, under light use of phones not kept in pockets, last a very long time - 3-5 years isn't unreasonable, and many will last longer. Batteries, under heavy use of a phone kept in a pocket and run hard, will still typically last 1.5-2 years. So in exchange for "slightly more inconvenience less than annually," you get a good bit more capacity and runtime.

Apple, in general, hasn't made their batteries nonsensically hard to replace. They've used the "pull tab sticky" sort of thing for some while, which is far nicer than "glue the whole thing down," and their newer devices are using some sort of electrically released magic (apply 9V to the adhesive, battery pops out).

There's no such thing as a free lunch.

It was more to prevent unexpected shutdowns. Which, I'll add, were a problem with Android devices at the time, and the Nexus 5, in particular, had three battery OEMs, one of which would only last a year before being unable to run the device in high demand situations (say, "taking a picture with the flash").

As lithium batteries age, their internal resistance goes up - you can model a battery as a voltage source and a series resistor accurately enough. Over time, that resistance goes up, which means, for a given current, you end up with less voltage "at the output." Most power supplies will compensate by pulling more current to provide the needed power, which will drop the voltage more until you slam into the low voltage protection circuitry that cuts power.

The Nexus 5s are the ones I'm most familiar with, and they absolutely had this problem with one of the battery OEMs (the only way to tell which OEM you had was to pull the battery out, they were labeled on the back). The typical symptom was, "The phone shuts down when you try to take a picture," because camera modules are power hungry, the CPU was spinning hard to keep up with rendering the view from the camera (and possibly doing some pre/post frame capture to find the best frame, I don't recall when that showed up), and the flash pulls a LOT of current, very briefly. So everything would simply shut down when you hit the button to take the picture.

Apple decided to attempt to limit this problem, and they locked out the highest tiers of CPU performance (which are the most power hungry), if the device was having brownout issues. It's a reasonable enough strategy. Where they failed (IMO) was in not alerting the users that it was happening, or that it was a battery health issue. The later iterations of it, where it tracks battery health, and will tell you if your battery is going bad and needs replacement, are what they should have rolled out, and didn't. My guess is that they didn't think it was going to be a major issue for many devices, so it was just a CYA sort of thing that would prevent shutdowns. Unfortunately, that also happened right around the same time that US carriers started dropping the "New phone every 2 years on contract!" thing, and so the iPhones of that era started being used rather substantially longer than the previously-expected 2 years, and, Apple, so drama for clicks.

Had they just gone about telling users, "Hey, it looks like your battery is getting weak, would you like to schedule a replacement? Otherwise, we've limited performance slightly to prevent shutdowns." - I think it would have been fine. And they did settle on that eventually. It just took a few iterations.

The book from a year or two ago, "Means of Control," by Tau, goes into some pretty good detail on the data collection and sales from just the adtech firms - where the entire ecosystem seems to be, "You can't use our data for anything but advertising... wink wink", and everyone knows exactly who is bidding on ads, and never winning any, just to slurp up location data and sell it. Or the "companies that don't sell the government." Also, they don't vet any clients beyond "The credit card is good."

> And because giants like Meta, Google, and Apple must collect as much of your personal data as possible, there’s little they can do to protect your privacy.

I quite disagree with the "must" there. They choose to collect as much data as possible, because that's their business model.

And the good news is, it's fairly easy to opt out of quite a lot of that.

Turn location services off, turn your phone off when moving about, and pay cash without "personal tracking cards" associated with you. Just about everywhere has [local area code] 867-5309 registered, if you care.

They're an absolute pain in the rear to deal with, because they're self igniting, and it propagates between cells.

Your typical lithium 18650 - vape cell, old laptop cell, whatever you know it as (18mm diameter, 65mm length, cylindrical), has a high end capacity of around 3500mAh - so 3.5Ah (amp-hours - so will take an hour to drain at 3.5 amps, 3.5 hours to drain at 1 amp, handwave goes here). At 3.7V nominal, that's around 13 Wh (watt-hours, a measure of energy capacity).

As a first order handwave, when a cell runs away and burns off all the materials in it (electrolyte, plastic separators, etc), you'll get about twice the energy out of the cell as the electrical capacity - so, ballpark, 25Wh for a fully charged 18650 running away. Except, it doesn't run away in an hour. It runs away in about 30 seconds, so doing the math on that, you end up with about 3000 watts for those 30 seconds. That, meanwhile, can heat nearby cells up enough to cause them to enter thermal runaway, and the whole pack will just go, until cooled sufficiently.

"Dumping a lot of water on the pack" will, generally, cool it down enough to stop this. Assuming you can get the water where it needs to be, and in something like a shipping container battery, that's far from given.

At this point, you've got a damaged battery, in unknown condition, with none of the existing current paths able to be relied on, and probably new current paths that may or may not exist yet (water, metal, corrosion, and those paths are often high resistance and slow to form, which creates a lot of heat). It's not really safe to disassemble it or work on it until things have been discharged, because if the pack has energy left in it, it's prone to do exactly what this article talks about - reignite, later, inconveniently.

As far as disassembling it, would you go work in a few megawatt-hours of energy, in unknown configuration, with the state of the safety systems unknown, in a charred environment of unknown toxins (what you get out of a runaway is far from predictable, beyond "generally unfriendly to humans")?

It sounds silly, but if the pack is confined and the fire isn't going to spread to other packs nearby (which is why they tend to be quite spread out), the safest thing to do really is to let it burn to completion. At that point, if it's actually burned out, there's no energy left in the cells to do anything terribly nasty, and you've burned off most of the electrolyte and such.

Anyway, the right answer is lithium iron phosphate for grid scale energy storage, but even those can catch fire if water gets in the wrong places, and they will, with enough prodding, burn.

> I treat it now more like advice from a friend. Great information that isn't necessarily right and often wrong without having any idea it is wrong.

"Drunken uncle at a bar, known for spinning tales, and a master BSer who hustled his way through college in assorted pool halls" is my personal model of it. Often right, or nearly so. Frequently wrong. Sometimes has made things up on the spot. Absolutely zero ability to tell which it is, from the conversation.

I don't believe it was capable of it, which is why it was so massive. The SR-71, which required inflight refueling repeatedly, only held 80k pounds of fuel (about 12k gallons). I don't have any good sense of fuel burn vs speed either, but in general, jets like to run high and fast. The old Lear 23s burned about as much fuel (pounds per hour) idling on the ground as they did at cruise, and I think the SR-71 (which mostly used the turbojets to keep the afterburners lit, at cruise...) fuel economy up high was quite good. Apparently the major problem with performance was keeping it from overspeeding - left to their own devices, the engine (... entire engine assembly, however long it was) was running so efficiently that they just wanted to go.

Per Wikipedia, the XB70 carried: 300,000 pounds (140,000 kg) / 46,745 US gal (38,923 imp gal; 176,950 L), on a maximum takeoff weight of 542,000 lb - so about 55% of takeoff weight was fuel.

A 747-8I carries up to 63,034 gallons, or about 400k pounds, on a max takeoff weight of 987,000 pounds, or about 42% of takeoff weight.

Interestingly, the ranges are about the same. The XB70's combat radius (there and back) is 3,725 nm, for a straight line range of 7450 nm, the 747-8I's range is 7730 nm.

High altitude supersonic flight is actually fairly efficient... if you can handle it.

> ...especially when someone is particularly passionate about them.

The engineer-type brain is very much prone to "... in order to prove we can," as opposed to "Because we should. Or because this is useful. Or because this even does the job claimed."

Across a range of fields. A/B testing "engagement hacks" falls into this category, as far as I'm concerned. It was certainly successful at the stated goals.

That era of aviation was nuts. I wish I was around for it. Men with slide rules working out the limits of material science, aerodynamics, and everything else, all at once. Because it wasn't enough to just push one limit, you had to push half a dozen others to get things to that first limit. And the rate of advance was just staggering.

The XB70 flew in late 1964. Concorde was doing revenue flights in 1976, cruising at Mach 2, with passengers being served luxury food.

> The Air Force learned that pushing the technological envelope resulted in plane that was difficult to build, difficult to maintain, difficult to fly, and perhaps even more importantly, was incredibly expensive; the program cost nearly 1.5 billion dollars, or around 11 million dollars per flight.

And nothing has changed. Pushing the limits is expensive. Always has been, always will be.