Strange reset during the day

[atop8918]

No, I haven't made any released changes to fix the resets.

[boB]

What I would like to see done when possible is for those that are having reset issues to
unplug their Ethernet/internet from their Classic for a day or two and see if they still
have resets.

Temporarily plugging in the Ethernet to check to see if it has reset should be OK though
if it is unplugged as soon as the information has been gathered or downloaded.

This will narrow down if it is still part of the network or something else.

Thanks,
boB

[RossW]

What I would like to see done when possible is for those that are having reset issues to
unplug their Ethernet/internet from their Classic for a day or two and see if they still
have resets.

Bob, as I mentioned in IRC, I'm interstate for a couple of days.
I can't unplug the ethernet, but I *HAVE* firewalled it both directions from the internet.
It can't initiate a connection out to anywhere, and nowhere outside can reach it.

Will see if it makes any difference. (It will of course still see local network broadcasts, ARP requests etc)

[boB]

What I would like to see done when possible is for those that are having reset issues to
unplug their Ethernet/internet from their Classic for a day or two and see if they still
have resets.

Bob, as I mentioned in IRC, I'm interstate for a couple of days.
I can't unplug the ethernet, but I *HAVE* firewalled it both directions from the internet.
It can't initiate a connection out to anywhere, and nowhere outside can reach it.

Will see if it makes any difference. (It will of course still see local network broadcasts, ARP requests etc)

This will be a good Pre-test before you get back, Ross.

Thanks !

boB

[dogfox]

dogfox,
Was web access enabled or disabled when it reset ?

Enabled.  It was after the second reset that I started researching whether others were experiencing these issues and I read about web access possibly being a culprit.  I have since turned web access off.    I have no reliable way of seeing if its still resetting or not.

I have had web access turned off since then and (I might be tempting fate here) I have not logged any voltage spikes above what I have set on the Classic's Aux1 for triggering the dump load.  We had pretty high winds (40-50kph) here all day yesterday and all night long and, thankfully, no resets that I can see.

Please look into the code for creation of the csv log files from the local app.  There seem to be irregulatities in the data (missing time periods, date/times reverting back to previous periods, etc.)  I can send you my files to check out if you wish.

Cheers 

[Resthome]

Please look into the code for creation of the csv log files from the local app.  There seem to be irregularities in the data (missing time periods, date/times reverting back to previous periods, etc.)  I can send you my files to check out if you wish.

Cheers

Yeah there is a bug in the Date Stamp formula that gets replicated incorrectly. The data in the time stamp column is correct. If you re-replicate the first cell of the Date Time formula all the way down your first column in Excel the Date and Time will get corrected. In Daily dumps from the app I get two days worth of data and I get one line showing the next day and the following line shows the formula reverting back to using cell B5 so all the dates come out incorrectly as is should be using the next B cell in sequence.

[dgd]

BTW folks....

The newest firmware...

http://fusion.midnitesolar.com/MidNiteSolarSetup_6-6-2013_v3.72.exe

has the data abort information registers in them.

If your Classic has done a reset, check these registers if you have modbus reading capability.
Otherwise, PM or email me.

Registers   4340 through 4343   (addresses  4339  through 4342)

Those registers should contain the information we need to help track this problem
down due to the watch-dog resets.

Thanks,
boB

boB,

My pv classic just crashed a few minutes ago and rest the kw/hr to zero.   200BA94A  fw 1401  network 1194

Noticed it when running local app froze.  First crash for a couple of weeks.

Modbus registers contain data   
reg              decimal            hex
4339             0                     0000
4340            18                   0012
4341            58240             e380
4342            61440             f000
4343            57633             e121

dgd

[RossW]

Modbus registers contain data   
reg              decimal            hex
4339             0                     0000
4340            18                   0012
4341            58240             e380
4342            61440             f000
4343            57633             e121

This from three I'm watching at the moment (none have recently reset)
4131=0x3200   <-- Useful register, as it shows the WD Reset bit
4341=0x0012   <-- these 4 however are identical values to yours
4342=0xe380   <-- but are you showing registers or addresses?
4343=0xf000    <-- yours appear to be 1 lower, suggesting address not register?
4344=0xe121

[dgd]

Modbus registers contain data   
reg              decimal            hex
4339             0                     0000
4340            18                   0012
4341            58240             e380
4342            61440             f000
4343            57633             e121

This from three I'm watching at the moment (none have recently reset)
4131=0x3200   <-- Useful register, as it shows the WD Reset bit
4341=0x0012   <-- these 4 however are identical values to yours
4342=0xe380   <-- but are you showing registers or addresses?
4343=0xf000    <-- yours appear to be 1 lower, suggesting address not register?
4344=0xe121

registers..
04130  is  0x3280
Are your registers that are the same values as mine just showing what happened at last reset?
04130 ic cleared each day end.

dgd

[RossW]

registers..
04130  is  0x3280
Are your registers that are the same values as mine just showing what happened at last reset?
04130 ic cleared each day end.

Odd, the numbers you quote as "registers" being 1 short of what I observe AND what's in the midnite modbus register map/pdf.

I have not yet seen any of those registers contain anything different to what they are now. Ever.
After a reset, before a reset, in the morning, at noon or night. Ever.

[dgd]

I looked at my other 'live' classic and it has same modbus values.  Maybe boB has not implemented reset debugging yet 

dgd

[boB]

I looked at my other 'live' classic and it has same modbus values.  Maybe boB has not implemented reset debugging yet 

dgd

Hi guys...
These values,

4340            18                   0012
4341            58240             e380
4342            61440             f000
4343            57633             e121

....appear to NOT be a result of a watch dog reset but default values when
the Classic first powers up.

I think that if you power down the Classic and power it back up again normally
that you will see the same values.  In fact, maybe you could give that
a try just to make sure.

Remember that if, on, MyMidnite, your flags register shows an 0x00000080 bit
set that that is a Data Abort or, WDT reset.  It ~could~ be that your
classic is resetting for some entirely different reason ?  Possibly an internal
power supply brownout ?   That appears to be what Ross is experiencing ?

OK, off to the science museum in Munich for the day.  I'll probably be back
when you guys all wake up.

böB

[RossW]

Possibly an internal power supply brownout ?   That appears to be what Ross is experiencing ?

I've got around 1500AH of AGM cells in reasonable condition, the inverter will call for the generator long before the midnight should ever see a voltage within 5% of nominal. (And since the midnite will work on a 12V system, surely nothing a 48V system could give it would EVER come close to a problem??)

[boB]

Possibly an internal power supply brownout ?   That appears to be what Ross is experiencing ?

I've got around 1500AH of AGM cells in reasonable condition, the inverter will call for the generator long before the midnight should ever see a voltage within 5% of nominal. (And since the midnite will work on a 12V system, surely nothing a 48V system could give it would EVER come close to a problem??)

I'm just saying that something else must be making the processor reset besides WDT timeout.   I wish I knew
what it was and could see it here.  Internal auxiliary power supply dropout is one probably cause in
addition to some weird jump to memory zero bug.

boB

[dgd]

I reset both Classics a few minutes ago and on 200BA94A the hex flags changed from
10043280   to
100432C0   then
10043200
so it seems only a power off/on reset will clear the ..80 flag.  I though this was supposed to clear every day 11:59pm? The MyMN shows it as an Internal reset and really needs to be auto reset at least daily.
the modbus registers are:
04340       0x0012
04341       0xe380
04342       0xf000
04343       0xe121
so these are just the normal power on values. 

I have definitely noticed that since changing to 1401 firmware these random resets are much less frequent

dgd