The combination "Found I2C slave (Nixie clock) (default) 105" and "Communicating with: Clock not found!" means that the WiFi module knows that there is a clock there, but does know know how to communicate with it. The status string also confirms this upper case letters means that the thing we are referring to is "on":
"W" - WiFi is connected
"N" - NTP time is valid
"S" - The SPIFFS file system was mounted
"U" - OTA Update is possible
"A" - Authentication for the web configuration is turned on
"d" - Debug mode is off
The "Display time" means that the NTP for some reason got an answer back that it could not interpret. I'm looking at this now, seeing if I can figure out why it got an answer back that looks valid "2021:02:21 17:37:27", but after processing it internally came back with an invalid answer "2000:01:09 12:29:06".
So: Next steps: Can you power off the the clock and power it on again, and let me know the version number that is shown for 1 second at start up?
In the meantime, I'll try to figure out if there is something funky with that time string that I don't know about.
The good news is that we can fix this remotely, because the "OTA" option is enabled. I can send you a new package to install via browser if we need that...
Three clocks assembled. Burned new at mega chips for each, burned new esp-8266 for each.
Clock #1 Rev 3.01 (define set at FWV2)
Clock #2 Rev 3. (Define set at REV3)
Clock #3 Rev 3 ( Define set at REV3)
All three clocks now show exactly the same behavior. All LED’s working. Hours and minutes Nixie’s appear normal. 10’s seconds is brighter Nixie than others. One seconds is blank. Swapping out Nixie tubes does not change behavior. Swapping out K155 chip does not change behavior. Swapping opto-couplers does not change behavior. ESP-8266 page shows clock not found! As before.
I completed Modular clock #4 today. It also displays nearly the same behavior. All LED's OK. Hours, Minutes digits OK. 10's of seconds extra bright, and 1's of seconds blank.
This one used the factory supplied ESP8266. It does communicate with the clock, but displays a different interface page than the units I flashed with the software off the repository.
WLAN IP 192.168.68.167
WLAN MAC BC:DD:C2:24:A7:D9
WLAN SSID GibbsCabin245
Time server URL
Time according to server 2021,2,28,16,41,20
Uptime 0 days, 0 hours, 1 mins, 57 secs
Time last update 48974
Serial Number 000513
Found I2C slave at 87
Found I2C slave at 104
Found I2C slave at (default) 105
Sketch size 387184
Free sketch size 573440
Free heap 27440
Boot version 6
CPU Freqency (MHz) 160
SDK version 3.0.0-dev(c0f7b44)
Chip ID 2402265
Flash Chip ID 146085
Flash size 1048576
So two questions today:
What is the best way get short video clips to you?
How do we do an OTA software update?
I'm sure once the main firmware and ESP firmware are properly updated, I will have 4 working clocks.
BTW, my order of pre-programmed chips has not arrived, and the tracking number does not give any information. Can you check on that?
Only Ian can answer your question about your order. Order processing is above my pay grade
Best way to get a video to us is to upload it to YouTube and post a link here. I'd prefer you post a simple URL link instead of using the video embedding feature of the forum since embedding is tempermental and IMO not to be relied on.
OTA updating applies only to the ESP's firmware. At this time it's not possible to flash the Atmega on the clock board OTA. Maybe some day. Since you are already flashing without OTA I'd recommend you stick with what seems to be working for you. I think your best bet is to make sure you have the latest versions of WiFi and Clock firmware from the repo, make sure you are compiling the clock firmware with the correct REV3 vs VER2 setting for each board, and flash away. I don't fully understand what is going on with your "weird digit" issues but it would be nice to confirm or eliminate firmware as the cause once and for all.
Look into it later when the dust is clearing off the crater.