Drgeoff, Yes, you are right, this is flash / EEPROM memory. Someone may still remember the old memory chips with ultraviolet erasure: but in order to reduce the cost, a version without a quartz window was released, which was considered once-programmable. But craftsmen still managed to wash it - with X-ray radiation. As in Russia, they are repairing Raspberries using a vacuum separator (designed to repair smartphone displays). I’ll explain my thought, because I expressed it poorly.
So, the PMIC chip was created, which created a number of problems. Moreover, developers and sellers did not even bother to say that even a smartphone repairman can replace this chip. It seems that some reasons were thought out that you can only buy 100 pieces of microcircuits, otherwise it’s impossible in any way, otherwise something will not work there.
And I brought it up with the “cheap version” of EEPROM - everything seems to work, only once and you can’t reprogram it. This, excuse me, is a delicate way of half-deception: you seem to get a full-fledged version of the chip, but in reality it is just a rejection.
It seems to me that here on the forum almost all technical specialists. And they understand that problems with PMIC, as we all recall, have arisen constantly. And none of the developers informed users of the causes of the problems. Just buyers suffered losses. The developers did not even bother to release any document with an explanation. Yes, to the credit of RPIs, bad computers often changed.
I really liked and liked it - I saw that PRI is a company with a human face, and not just a company doing business.
In Russia, electronics are divided according to their own rules: for example, microchips are sold with acceptance 1 (the so-called acceptance of quality control department - the technical control department, when the factory tests the chips), acceptance 5 (customer acceptance, in the case of the military - the military representative controls the tests) and acceptance 9 (when only the most qualified personnel are involved in the work - for space and nuclear power plants). For example, acceptance 5/9 does not mean that the chip is radiation-resistant - resistance to special. factors are indicated in the (non-public) documentation for the chip. But there are simply electronic components that do not check, but simply write about “tolerances”, or indicate that the components are from a “batch with marriage”. (high failure rate).
But there is also a problem microcircuit, not Russian, which in a small batch suddenly and suddenly dies by itself, for no reason (we all remember that), "dies" due to the fact that the GPIO pins are located incorrectly (who is this thought of placing power contacts close? Is it intentionally that there is a danger of short circuit?), if even the second version of PMIC periodically dies for no reason, then the manufacturer could at least release some documents, like Nvidia, Nikon, Samsung and others did major players. RPI could not solder PMIC with refractory solder (because of which the repair price doubled), they could put a microcircuit in the “cradle” so that it could be changed without soldering equipment. RPI could simply send this unfortunate PMIC to people who suffered from incorrect operation and were able to replace the smd component. Could write instructions on how to properly change this chip (as this Russian guy did an engineer, but the English developers did not) This is what I wanted to say. These are all inexpensive procedures on a general scale, but forcing respect for managers and engineers.
In the end, it was possible to clearly spell out the instruction that any breakdown associated with food would be compensated one way or another.
Honestly, I did not want to write all this, so I incorrectly hinted that sooner or later these issues would still be raised. Sooner rather than later. Because the links to the constantly "dying" Raspberries of the fourth version are too numerous.
We, the buyers, do not pay our money for showing statistical charts, they say “only a few PMICs” suddenly stop working. Well, when, about twenty years ago, Fuji pierced with a microcircuit (due to improper temperature conditions in the hard drives, the microcircuit “died”), or when IBM was unable to correctly solder a video chip in the T series, this at least served as an example of how not to act .
And Nvidia simply changed explosive batteries without question and sent a completely new device. Nikon tried to resist “shutter problems,” but eventually pleaded guilty.
I’m not trying to blame someone, I’m not saying that Raspberry is a bad computer, I just read a few articles today and was shocked that the problem was somehow hushed up. This is what managers do, but this is not the approach of an engineer or technical specialist.
Maybe I'm wrong, but I express my opinion.
PS As an example, I want to say that the users of Raspberry themselves figured out the increased voltage on the power supplies, that it was not only the thin power wire that was to blame, but the incorrectly created power system — instead of a normal power connector, they used a smartphone. There is no protection, not even a fuse. Well, this suits everyone, although it leads to the fact that a lot of perfectly functioning computers are in the trash - due to the fact that an inexperienced user constantly raises the voltage, because an incorrectly programmed lightning signal shows technically correct information, but leading to confusion .
In general, these are all my “philosophical reflections” - what to do in such cases, how to act and whether it is permissible to teach students to program on devices that are not entirely correctly made.
I also do not write well in English, so my words may seem rude. I would not want that to be so perceived.
I respect the ideas of Raspberry and I was delighted when the developer suggested creating a "computer for everyone" (or even a "computer for poor people"). This is a brilliant and wonderful idea, it is humanistic (which is the most important). maybe that's why the PMIC marriage problem upsets me like that.