IotaWatt Not Updating

Hardware IoTaWatt
1

Hi Bob (@overeasy)
I’m trying to get my IotaWatt to update but it seems it doesn’t want to, have you any ideas from the log?

No clock yet: SD initialized.
1/3/18 21:37:16 Real Time Clock is running. Unix time: 1515015436
1/3/18 21:37:16 Version: 02_02_20
1/3/18 21:37:16 Reset reason: Software/System restart
1/3/18 21:37:16 Trace: 15,16,17,18,11,90,91,92,12,13,14,15,16,17,18,11,90,91,92,12,13,14,15,30,31,32,33,37,16,17,102,103
1/3/18 21:37:16 ESP8266 ChipID:456452
1/3/18 21:37:16 device name: IotaWatt, version: 3
1/3/18 21:37:16 Local time zone: 0
1/3/18 21:37:19 Connecting with WiFiManager.
1/3/18 21:37:24 Updater: Update from 02_02_20 to 02_02_27
1/3/18 21:37:24 MDNS responder started
1/3/18 21:37:24 You can now connect to http://IotaWatt.local
1/3/18 21:37:24 HTTP server started
1/3/18 21:37:24 dataLog: service started.
1/3/18 21:37:25 dataLog: Last log entry:1515015435
1/3/18 21:37:25 statService: started.
1/3/18 21:37:25 timeSync: service started.
1/3/18 21:37:25 WiFi connected, SSID: WiFi
1/3/18 21:37:25 Updater: Update from 02_02_20 to 02_02_27
1/3/18 21:37:29 EmonService: started.url: 192.168.10.34/emoncms, node: IotaWatt, post interval: 5, unsecure GET
1/3/18 21:37:29 EmonService: Start posting from 1515015435

Regards
Dave

2

Hi Dave,
I just got back tonight. I’ll look into this tomorrow and see if a 02_02_20 system will update for me.

3

@Dave,
Hi Dave. I regressed an IoTaWatt to 02_02_20 and it updated fine to 02_02_27, so no easy answer there. The auto-update has been rock-solid, so this is going to be something off the beaten path.

First, can you check your SDcard (file manager, left margin lists files and directories). See if there is a library 02_02_27 (or any other similarly named directory). If so, delete it (right click, delete). Then restart and see if update completes.

If that’s not the case, or it is and update still doesn’t complete, could you connect a monitor to the USB plug (microUSB to USB) and start up a serial monitor (115200 baud). restart the IoTaWatt and post the entire restart sequence. There are additional diagnostics possible via the serial and it may indicate why the new release is not being successfully downloaded).

Thanks,
Bob

4

Hi Bob @overeasy
I cant explain it but I took the SD card out and put it back in and rebooted, the unit downloaded the firmware!
I have noticed that it fails to get the time from the NTP server, I presume that the server address is hard coded in?

** Restart **

No clock yet: SD initialized.
No clock yet: Real Time Clock not initialized.
No clock yet: Version: 02_02_20
No clock yet: Reset reason: Software/System restart
No clock yet: Trace: 92,12,13,14,17,18,11,90,91,92,12,13,14,17,18,11,90,91,92,12,13,14,17,18,11,90,91,92,12,13,102,103
No clock yet: ESP8266 ChipID:456452
No clock yet: device name: IotaWatt, version: 3
No clock yet: Local time zone: 0
No clock yet: Updater: Update from 02_02_20 to 02_02_27
No clock yet: Updater: Update downloaded and signature verified
No clock yet: Updater: firmware upgraded to version 02_02_27
No clock yet: Firmware updated, restarting

** Restart **

No clock yet: SD initialized.
No clock yet: Real Time Clock not initialized.
No clock yet: Version: 02_02_27
No clock yet: Updater: Installing update files for version 02_02_27
No clock yet: Updater: Installing GRAPH.JS
No clock yet: Updater: Installing INDEX.HTM
No clock yet: Updater: Installing TABLES.TXT
No clock yet: Updater: Installing CNFSTYLE.CSS
No clock yet: Updater: Installing EDIT.HTM
No clock yet: Updater: Installing GRAPH.HTM
No clock yet: Updater: Installation complete.
No clock yet: Reset reason: Software/System restart
No clock yet: Trace: 130,131,132,133,134,135,136,137,138,139,18,11,90,91,92,12,13,14,17,18,11,90,91,92,12,13,102,103,92,12,13,14
No clock yet: ESP8266 ChipID:456452
No clock yet: device name: IotaWatt, version: 3
No clock yet: Local time zone: 0
No clock yet: MDNS responder started
No clock yet: You can now connect to http://IotaWatt.local
No clock yet: HTTP server started
No clock yet: dataLog: service started.
No clock yet: dataLog: Last log entry:1515096790
No clock yet: statService: started.
No clock yet: timeSync: service started.
No clock yet: WiFi connected. SSID: xxxxxxxx
No clock yet: historyLog: service started.
No clock yet: historyLog: first entry:1514833320
No clock yet: EmonService: started.url: xxxxxxxxxxx:80/emoncms, node: IotaWatt, post interval: 5, unsecure GET
No clock yet: EmonService: Start posting at 1515096790
No clock yet: Server: Sent less data than expected!
No clock yet: Server: Sent less data than expected!
No clock yet: Server: Sent less data than expected!
No clock yet: Server: Sent less data than expected!
No clock yet: timeSync: Failed to get NTPtime.

Regards
Dave

5

That’s a problem. The unit is not logging power when it doesn’t know what time it is (or what day it is for that matter). There is some issue preventing the UDP NTP transaction. The time server is time.nist.gov, which is the global address for all servers.

You should reboot that system until it gets the time. NTP is a pretty reliable and straightforward internet service using the basic UDP protocol.

Glad you got updated, but that was weird too.

BTW/ How did you lose the time? It was set in the logs posted yesterday. Did you remove the battery?

6

Hi Bob @overeasy CC @glyn.hudson
I think the reason why its lost the time is due to the horrific soldering of the components including the battery, I’ve taken some photos but its hard to photo SMT components due to there size but I think I’ve captured the important bits.
If I’m honest I only noticed it because one of the jack sockets fell of the PCB while putting it back together but it does highlight the requirement for quality control on the company that you’ve got putting these units together.

Regards
Dave

I’m pretty sure the brown resistor? could be a dead short?

Zoom 1
Zoom 12003×751 170 KB

You can see that the legs haven’t been heated when it was put of the PCB

Zoom 2
Zoom 21987×769 177 KB

I’m not too sure is that’s a gap under that contact?

Zoom 3
Zoom 3404×905 40.7 KB

IMG_5384
IMG_53842591×1727 2.04 MB

IMG_5386
IMG_53862591×1727 2.07 MB

IMG_5379
IMG_53792591×1727 2.03 MB

IMG_5373
IMG_53732591×1727 2.17 MB

IMG_5387
IMG_53872591×1727 2.04 MB

IMG_5385
IMG_53852591×1727 2.31 MB

7

Hi Dave,
That is a crappy board. Not only that, but it’s a 4.1 version board that has a problem with the battery backup of the real-time-clock. Here’s what a real IoTaWatt looks like:

IMG_0892
IMG_08922592×1936 1.11 MB

If you zoom on that you’ll see the soldering is great, but also that there is an RC circuit on the power pin of the RTC so that the voltage drops slow enough for the battery to kick in.

You have a prototype board, built with low-temp solder in my toaster oven. Great for development, but not for sale. Sorry you ended up with that. I wasn’t aware any of those early evaluation devices had been sold. You will get a production board.

There is a run in manufacturing now that will be available in three to four weeks. If you are in the USA, I have one production board from the first run I can send you and I can reconcile this with OEM.

Again, apologies for that.
Bob

8

Hi Dave,

Could you message the shop at [email protected] or PM me with your order number - we will arrange for a replacement to be sent asap.

9

Hi Bob @overeasy
Yeh that board is loads better! I’ve sent Gwilym a message as I’m based in the UK and I presume that the shop has stock? thou I am happy to wait as I haven’t got around to installing it and I had been eagerly watching your project develop.

Kind Regards
Dave

10

Hi @Dave,

Very sorry to hear about the faulty unit. You received one of the first iotawatts that has been hand assembled. Hence the discount. The units where tested to be working but we obviously missed the RTC issue. We will organise a replacement with a new production unit. Very sorry for the hassle.

There might be a few weeks wait while we are expecting delivery of more units.

11

Thanks Glyn no worries