Do as gormanpe did; burn the SD Cards on a PC which has a micro-SD Card slot. That seems to have been the only solution for resolving the issue so far identified. YMMV.
I still use leaded solder, but I am not a manufacturer.renice123 wrote: ↑Thu Dec 12, 2019 10:12 amPS I remember the times when used lead solder. When environmentalists forced all manufacturers to abandon lead solder (yes, it is harmful), many electronic devices began to fail. Badly soldered parts fell off. Alas, this is so. I believe that we should fight for the environment, but we should not splash out along with dirty water from the basin of the child =)
A couple of posters have stated having a feint green LED rather than bright green LED. That's interesting because the LED is wired across 3V3 and 0V switched by a FET controlled by the ACT_CLK line.
Hi, renice123. I see now you are probably from another country and smart and not just too young or other problems. Maybe we can blame bad google translate. It's not required but showing approximately where we are from in the avatar can help avoid misinterpretations.
I am adding this info just for reference.renice123 wrote: ↑Wed Dec 11, 2019 9:33 amIt seems to me that if we want to find out what happens when the "green LED of death" appears, then we need to provide more technical information.
For example, the following questions immediately arise.
1. Which memory card was used (and even on which card reader the image is recorded).
SanDisk Ultra 32GB A1
2. Which case was used (closed, open, metal, plastic, and so on.
3. What kind of power supply. Whether there was a voltage drop or frequent malfunctions in the electrical home network.
Original power supply
4. Whether the temperature of Raspberry was controlled (for example, I have constant temperature control turned on and when I reach 60 degrees, Raspberry is turned off).
Cooler running non stop.
5. Are any devices connected to the GPIO
No devices were connected to GPIO. Only the cooler was connected to 5V.
6. Where is Raspberry physically located and can it be that the microcomputer is exposed to humid air or is in a stuffy shoe box =)
Office space 25C
7. In what mode was the computer used? Constantly worked or often turned on and off.
Headless with no X. K8S was running but without any pods.
What programs were most often used (server or web browsing through Chromium and so on)
Have any changes been made to config.txt
Indicate the firmware version and whether EEPROM was tried to flash (before or after the "green death LED").
Stock EEPROM, no changes in config.txt were made.
8. How the computer behaves after the appearance of the "green death LED". How much current does Raspberry consume in this case (measure the current with an ammeter). Can check some signals on the board with an oscilloscope? What happens if you lower the voltage from the power supply to Raspberry (for example, to 4.8 Volts or lower, to 4.5 Volts. Will the red light disappear?).
9. Perhaps it makes sense to post photographs of Raspberry and setup, the environment in which the microcomputer worked. That is, indicate the brand of keyboard, mouse, connected devices.
Only the ETH was connected, not WiFi or keyboard, mouse was used.
hippy wrote: ↑Thu Dec 12, 2019 10:33 amA couple of posters have stated having a feint green LED rather than bright green LED. That's interesting because the LED is wired across 3V3 and 0V switched by a FET controlled by the ACT_CLK line.
The most likely reasons for that LED to be feint rather than fully on is 3V3 isn't 3v3, the 3V3 is switching levels, or the ACT_CLK is switching, creating a PWM-like dimming of the LED.
One possibility where the LED is feint could be that the PMIC is starting up and then shutting down, repeatedly pulsing the 3V3 line. If you have a scope you could perhaps take a look at the 3V3 rail. A multi-meter would probably show some voltage but not full 3V3 if it is power-cycling 3V3.
Do you have a multi-meter you could use to carefully measure the 3V3 with respect to 0V ?
Possibly, possibly not. I actually managed to get myself a 'red plus green LED always on, no HDMI' situation a few days ago but I was hacking about with files within a 'pieeprom-YYYY-MM-DD.bin' image and updating the Boot Eeprom with that so, whilst a notable outcome, it wasn't entirely an earth shattering surprise.