WeatherDuino Forum

I have exactly the same issue. Data logger is working fine. Am overseas at the moment and was going to bring this issue up when I returned home. Have ordered additional Arduino unit to see if this fixes the problem.

(22-07-2015, 00:55)Werk_AG Wrote: [ -> ]vk2kzo, data from I2C_Scanner it's ok, but the output of the FlashMemory_Test shows that the system is not comunicating with the flash memory.

This lack of comunicattion with the flash memory may not explain the "freeze" problem, but should be solved. Recheck all the flash memory weldings.
Just to be sure, try to disable the data logger function in the user config options, and check if the freeze problem persists.

Do you have any close picture of the flash memory?

(21-07-2015, 15:34)Werk_AG Wrote: [ -> ]vk2kzo and qldbureau:
Please follow the uncle_bob advise and check if you are using a 0.1uF (100nF) capacitor where on the board is marked R3/C3. Install only a 0.1uF capacitor.

Use the I2C_Scanner and FlashMemory_Test utilities and please put here the results.

What ARCHIVE_PERIOD are you using?

I have just checked my board, I have the 0.1uF installed at R3/C3.

Here are my I2C_Scanner results. Bearing in mind from some of my last posts in my build thread, we found that it does not pick up the RTC EEPROM for some reason, however the flash memory seems to work.

I2C Scanner
Scanning...
I2C device found at address 0x27 !
I2C device found at address 0x68 !
I2C device found at address 0x77 !
done

Flash Memory Scanner test with no power supply on the 12 volt port:

Security register (byte 64 to 127) :
0xC, 0x1, 0x15, 0x15, 0x1C, 0x1D, 0x1F, 0x22, 0x0, 0x0, 0x5D, 0x2, 0xFF, 0xFF, 0xA2, 0xFF, 0x30,
0x30, 0x31, 0x54, 0x36, 0x30, 0x35, 0x33, 0x12, 0xB, 0x3F, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0x3B,
0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF,
0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF,
Status register :10001100
Manufacturer ID :
1F
Device ID (part 1) :
22
Device ID (part 2) :
0
Extended Device Information String Length :
0

Interestingly here is the same test with power also applied to the 12 volt port:

Security register (byte 64 to 127) :
0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0,
0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0,
0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0,
0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0,
Status register :0
Manufacturer ID :
0
Device ID (part 1) :
0
Device ID (part 2) :
0
Extended Device Information String Length :
0

So it appears to me that supplying the 12 volts stops the logger from working, which I was not previously aware of.

The part number of the memory chip I sourced is Mfr. Part No. AT45DB011D-SSH-B which I sourced from RS Components in Australia.

My archive period is set to 5 minutes, the default.

Do you guys have R5 an R6 installed by chance?
If so remove them.

(22-07-2015, 12:22)uncle_bob Wrote: [ -> ]Do you guys have R5 an R6 installed by chance?
If so remove them.

No never installed in mine.

I did have the resistors installed but removed them but made no difference. My logger chip was also purchased from RS but I believe it is working as it should be. When I am connected to my computer receiver is being updated (usually). When I start Cumulus the logger historic data updates fine but the receiver does no longer refresh as the display does not change until I disconnect from Cumulus. As the display does not change while I have Cumulus open I assume that the logger is not being updated. I cannot comment on what happens without 12volts being applied as I am away from home at the moment.

(22-07-2015, 12:48)qldbureau Wrote: [ -> ]

(22-07-2015, 12:22)uncle_bob Wrote: [ -> ]Do you guys have R5 an R6 installed by chance?
If so remove them.

No never installed in mine.

Just to clarify mine does not usually lock immediately. It has ranged from within half a minute to maybe a few hours before activity ceases. Since I have the LEDs on the front panel it only needs a glance for me to know.

I have tried different plug packs. One capable of 1 amp at 12v, the other up to 3 amp 12v, and of course I have changed USB cables.

The memory logger flash test recording all 0's when 12v was supplied was curve I did not expect. I am not sure the significance of this. It possibly may explain the day my PC auto restarted on a Windows update and stopped at the login screen, (Windows 10 Preview actually) whilst I was at work. I expected a large log download when I restarted it and there was nothing. Only I cannot remember if I had left the 12v supplied or not. Otherwise, if I leave only the USB connected and Cumulus is not running, there are archives downloaded when I start Cumulus again.

I will do more checking; as far as I can see Windows 10 Preview has been working perfectly with this. I will test separately again to be sure, and see what happens when only plugged into 12v power, as well as putting my Windows 7 laptop on for awhile. I probably will not get to that until the weekend.

The common point of the three reports are the "freeze", that seems to happen at different circumstances for each of you:

Quote:Disconnect usb lead, LCD display immediately shows updated screen.Same on two rx units .Dodgy serial to usb chips? Has, on occasion, worked fine on Cumulus, but only for a short while, then locks up as before...

vk2kzo - "freeze" occurs as soon as you connect the usb cable, but occasionally may work.

Quote:Just to clarify mine does not usually lock immediately. It has ranged from within half a minute to maybe a few hours before activity ceases. Since I have the LEDs on the front panel it only needs a glance for me to know.

qldbureau: seems similiar circunstances to vk2kzo, adding that the logger seems to stop working when the board is powered by external power.

Quote:When I am connected to my computer receiver is being updated (usually). When I start Cumulus the logger historic data updates fine but the receiver does no longer refresh as the display does not change until I disconnect from Cumulus.

dormani: a slightly different behaviour, but with the same results.

From all reports, I think I can conclude that the problem is somehow related with the USB communication. I never seen this behaviour, and I'm trying to reproduce the problem, one thing I'm sure, it's not software related.
Not all Arduino Nanos are equal, even those that seems to be, and use the same chipset. I did some research on google, and found some reports of problems with the USB communications with certain CH340 "chinese" Nanos, when we get power from the 3.3V pin, without decoupling it to ground via a 100nF capacitor.
Add this 100nF capacitor to the RX v4.03 board is very easy. I'm not sure if this will solve the problem, but if any of you want to try, here is a picture showing a easy place to solder the referred 100nF capacitor. If you prefer for esthetic reasons, it can be done in the under side of the board.

tried the 100nF decoupling, Werk, didn't seem to make much difference - annoying isn't it !! Also replaced the Nano, had to desolder (!), also didn't make much difference, although it was only another chinese Nano............ (used a socket this time...)

Well after all that. This morning I connected the 12 volt power supply again, and after no freeze for half an hour I just left it. It has now been running for 14 hours without freezing, with Cumulus running except for an hour or so tonight where I disconnected the USB cable and let it run on 12 volt power alone. No freezes. When I reconnected the USB cable and restarted Cumulus it downloaded 13 archives out of the logger and has continued running without a glitch.

Obviously this is great for me thus far. Nothing has changed as far as I know, unless the re-flashing last night so I could run the scanners has reset something.

(23-07-2015, 11:36)qldbureau Wrote: [ -> ]Nothing has changed as far as I know, unless the re-flashing last night so I could run the scanners has reset something.

Hmm, maybe you're on to something.