The thing is, lithium ion batteries catching fire or exploding isn't that unusual. Li-ion batteries have major advantages over other battery chemistries, so they're literally found in everything. If it's a run of bad batteries, why aren't other company's phones catching fire? That implies that it's not really a battery problem, it's a system problem. Which is to say it's a Samsung problem.
The story is hard to pin down, so I've been bouncing back and forth between engineering sites like EE Times, business sites and those for the technically-inclined consumers, like C/NET. Samsung, like most companies in this situation, is trying to balance being 100% open and responsive to customers, while not saying things that would discourage new customers and all the while realizing there's nothing they can do to please some customers. At first, they thought the batteries were being pinched too hard in the slim case of the Note 7, as seen in the bottom video at the C/NET link, (originally posted in mid-September) but that cause seems to have lost some of the blamestorming it was subject to.
The company replaced the returned phones with units thought to be fixed in some way, but as of a couple of days ago, it has been reported that some of these supposedly known-safe units have already caught fire. Today, the company warned users not to use any phones, even these replacements, and said they were shipping fire resistant glass shipping containers to return the phones.
Unfortunately, there is no clear answer I can give you as to why these particular phones are an issue. It's being reported that the relevant Samsung engineers are virtually locked into the building until they solve this, and have been unable to reproduce the problem. Each phone that catches fire tends to gather news coverage, but there really haven't been many. C/NET says they can account for 112 phones catching fire. That's less than .01 % of the phones sold.
I've got to tell you this is a hard kind of problem to fix. I can also guarantee that the first thing that's going to be found is actually not the cause but a dead end. The problem is only going to happen once (after that, the phone is destroyed); they'll need to simultaneously test a lot of phones because it only happens to a small percentage of the product, and it doesn't happen immediately. It's not new phones on their first charging cycle, and the phones were only out a few days before the first fires were reported, so it's not old, worn out batteries. While one might expect the problem to show up during charging, there are several reports of the phone just catching fire while sitting on a table or desk or something. I would say it's going to take time to get this resolved. The best thing they could do would be to find a way to make an infrequent, rare, disastrous failure happen more often.
EE Times relayed an interesting idea:
Attributing an unnamed source who has spoken to Samsung chiefs, the Financial Times reported, “Problems with the phone appeared to have arisen from tweaks to the processor to speed up the rate at which the phone could be charged.”That makes me wonder if the way the code executes in those two processors isn't different somehow. Since Samsung decided to cancel all production of the Note 7; if not killing off the latest model phone then putting it into a medically induced coma, that seems to indicate Samsung is relatively sure the problem isn't simply defective batteries. It seems that the Korean giant realizes it must look elsewhere — other than batteries — to explain the alarming phenomenon of its smartphones bursting into flames.
Samsung's Note 7 comes in two versions, with one using Samsung's own Exynos 8893 processor and another based on Qualcomm’s Snapdragon 820 processor.