New IoTaWatt, RRR after ~48 hours of being fine

So I have a brand new unit (2239 S06292), which was working fine for the first two days of operation.

  • Since 1am this morning it’s started with the RRR code.
  • removing the 9v reference voltage lets it boot and appear on wifi
  • adding this at any point immediately results in RRR
  • admin and user passwords are not working, getting constant prompts to login with 401 as the only response with either password.
  • not extracted the SD card yet but looks like It could be the culprit based on other recent posts?
    • superblock issue when reading SD on fedora FFF-FFF
  • known good 64GB SD card formatted as FAT(and FAT32), results in GRR (with github /SD/ contents copied to /
    • going to try with the card formatted smaller just incase, no dice.

side note, old SD card still goes dull green until the vref is added, but cannot be opened on either windows or linux (I have other stuff I can try there)

IoTaWatt doesn’t support EXFAT SDcards. Replacements should be 32GB or less and formatted FAT32. We strongly recommend formatting with the SD Association Formatter.

The recommended size is 8G SDHC formatted FAT32. Larger cards are more expensive and provide absolutely no benefit. Also, the IoTaWatt SDaccess is inherently slow so using super fast cards is likewise of no benefit.

Recommend reformatting the original card with the above formatter, reload with Github files and restart.

My only windows device is a work laptop and the SD Association Formatter requires admin access which is a non starter.

I have formatted the original card using gparted, and loaded the files contained in /SD/ to / on that card.
This leaves me with a dull red LED, no AP and no connection to the previously saved WiFi (which it was doing in the GRR state before formatting this card). I’ve performed a RTC reset by removing the battery but no change, no new files or logs have been created on inspection of the 8GB card.

Placing the 64GB card back in reverts to RRR.

I have 2x 32GB cards on order (current pricing on 8GB from reputable manufacturers are double the price of 32GB variants on Amazon UK)

update - opened serial console and after several resets wifi is working

The dull red is because you removed the RTC battery.

The biggest problem with this ongoing SDcard issue is that people are doing things that are unrelated to an SDcard issue and creating compound problems. This problem potentially affects only about 250 recently shipped units. THERE IS ABSOLUTELY NO REASON TO REMOVE THE RTC BATTERY.

If you cannot use the SD Association formatter, I guess you will need to wait for the new cards.

I get your frustration, and the RTC battery was removed AFTER testing everything else I literally wrote my last message and edited to add the RTC after attempting it.

Troubleshooting — IoTaWatt 02_03_20 documentation mentions failure of RTC to init, reseating the battery should force reinitilisation.
backed up by Always dull red light - #20 by Akos - mentions resolved with RTC reset.

Note the device is now working and has wifi again.


Help me out here, I don’t see where reseating the battery is mentioned in the troubleshooting docs. I’ll revisit that instruction to explain that initializing the RTC means the IoTaWatt getting the time from an internet time server and setting the current time.

I didn’t comment on that thread as it related to very old units and a hardware mod to the ESP, I will do so today. While the user believes the RTC battery had something to do with it, I believe it was coincidental to everything else that was happening. I’ve troubleshot hundreds of problems with prototypes and production units and never had an issue that was caused or remedied by removing the RTC battery.

I say all this not to be argumentative. Like the post referencing the prior battery removal, others pick up on it and try it along with every other proposed solution to any problem, again, causing compound problems or symptoms that make identifying the original issue difficult or impossible.

My inference of the word init was for the clock to start counting from 0, prior to any sync occurring as the IOTaWatt can be run without internet*.
Combined with the linked post re dull red led, my assumption was that by forcing a reset may clear a fault state and put the system in a state for resync (akin to clearing bios cmos).

*(I re-read the quick start and it does mention the need for an internet connection for first run for this purpose, but the blinders were already on earlier :stuck_out_tongue: )

Its all good, I try to be logical in my approach and should know better as I work with databases for a living (users urhgh). I also tinker with a lot of microcontroller stuff as an end user for several of my hobbies (drones, 3d printing, iot etc) so the Dunning–Kruger effect is in full force here and those loose associations made enough sense that it seemed reasonable to try.

1 Like

That works both ways, so we’re both culpable.

1 Like