Nixies work fine, pins correct, wiring ok, weird symptoms

Honestly I think I possibly cooked a controller.

Voltage is fine, have it running at 9V at VIN and the high voltage is within range
MOSFETS are getting a warm but nothing that's alarming
Nixies all show the right numbers and have no dead numbers (using 4x IN-12A's, 2x IN-17's, and 2x INS-1's)
The wifi controller is configured to NY-EST and that seems to beam down correctly, can connect perfectly fine.
Factory reset it twice and replaced my cable looms in case I mixed them up or had some sort of interference.

The seconds will count up fine, but it goes:
01, 02, 03, 04, 05, 06, 07, 08, 09, 00, 11
Then continues, 12, 13, 14, 15, 16, 17, 18, 19, 10, 21
Also at some point my minutes will display in the 60+ range and my hours is displaying 34

Does it sound like I smoked something? I have about 10 extra K155ID1's since I was planning on building my own until I found this driver, I'll give that a swap tomorrow. I also have an Arduino UNO laying around if reflashing the firmware sounds like a fix or replacing the ATMEGA chip...Thanks in advance.

Hi iamtuck3r,

Before you go swap happy on the K155ID1 and atmega, have a close look at the wiring on the last (1 second) IN-17. It sounds like you've got the wiring shifted to the next higher digits on that tube. This can be verified with a factory reset. If I'm right, the 1 second tube will display 1 digit higher than all the other tubes when they count up together.

I've made this mistake before and it sort of sucks when it's on a tube that has to be soldered instead of just unplugging and shifting the pins by one.

Let us know what you find,
Jim

Ok, just got home. Went ahead and double checked the IN-17's and the wiring from the bulbs to the PCB all looked fine. I created a circuit board Eagle PCB and had a board printed out so I'm confident the board and wiring is right. I also I did the factory reset and all digits light up with 9's. Just to make sure the wiring was right again I swapped the 10's seconds cathode with the 1's cathode and received the same thing. 10's seconds now showed it counting up 1-9 and the 1's didn't reflect correctly.

New development. I studied the tubes for a bit and noticed it's not just seconds. The ten's group in everything is +1 digit. It's currently 6:37:22pm and right now the clock is showing 19:47:32. I'm not sure if the wifi module is also communicating with the board correctly as it's not addressing daylight savings time (which is why the hours is not reflecting 18). Another weird thing is if I have the 'blank leading zero' set to off the 10's hour shuts off, but if I enable it then it shows a 1...so the +1 theory stands for all 10's positions on the clock.

I'll attach some pictures of my entire setup if that will give you a better idea of what I'm working with. I DID buy this clock kit quite a while ago. Maybe 1.5 to 2 years ago so I'm going to try flashing the most updated firmware next. It's the Rev 4.x version.

Picture 1) Actual schematic
Picture 2) Board layout
Picture 3) Actual PCB
Picture 4) Whole setup
Picture 5) Right after factory reset

OK. It does seem to be something to do with the tube wiring. First off, I'm assuming you're using some version of the Classic Clock kit. Do you know what version of the firmware it is running? In later versions, after a factory reset is done the clock counts continuously 0,1,2,3.....8,9,0,1,2..... until you press the button on the 8's. You've said it shows all 9's and the picture you've provided shows that. I don't recall that being normal, but that's why I asked you the version of the firmware.

I'm going to study your pictures, an earlier assembly manual, the spec sheets for the IN-12 and IN-17 pinouts and see what I can come up with before noon tomorrow.

If you see this before then I need:
1 - the model clock you are building including any revision numbers.
2- the version of firmware that it is running on.

I'm pretty sure we can get this figured out.
Jim

I remember placing the order, the emails show Feb 10th 2017. It's a Classic Rev4 with the Wifi module. Getting the FW number as we speak.

Edit: I ended up swapping the ATMEGA with the newest version of FW flashed from my Arduino UNO. Now when the clock boots up it cycles 0-9 on all digits perfectly. Only problem is now that once that's cycled a few times it's completely blank. It's a little after midnight here so I'll continue the troubleshooting tomorrow.

Edit 2: Saw your comment about pressing the button at 8. The sequence then stopped and pulled the time but still doing the same thing and now it seems even less responsive to the wifi module I'm just going to pull the tubes from the sockets and manually wire things up and see if I get the same result tomorrow.

Hi iamtuck3r,

I took a closer look at your board and the terminals. If it were me making the board, I would have stayed with the same terminal numbering scheme that Ian used just to avoid confusion. I'm speaking of the cathode connector. You have it marked 1-10, and Ian's schematic has them marked 0-9. From what I can tell, they should be connected:

Your board - Ian's board
1 - 0
2 - 9
3 - 8
4 - 7
5 - 6
6 - 5
7 - 4
8 - 3
9 - 2
10 - 1

I think your troubles may be in that connection. If you could wire up individual plugs for that connector, you could play around with the wiring to see if there is a problem there.

With the new firmware on the new Atmega, It is behaving as expected. Ian coded a time-out into the initial test sequence so that it would not burn up the MOSFET running uncalibrated. If you unplug it for a minute or so to let it cool down, you can run it again.

Also the firmware on the Atmega must match the firmware on the WiFi module for it to communicate properly. You must also flash the ESP-01 with the firmware included in the WiFi time provider folder. You can pick up a USB - ESP8266 board from eBay or Amazon for under ten bucks. You just need to attatch a push-button to the bottom of the board to put the ESP-01 into flash mode - google it. There is an add-on for ESP8266 boards that will work in the Arduino IDE to program it.

Let us know what happens,
Jim