Invalid date in history after upgrade

[SolarMusher]

Have upgraded to 1181 FW two days ago (MNGP1042FW), and have invalid date in data history in day 1 (05/05/2003), day 2 (05/05/2003)  . Day 3 is right with 12/12/2012. I re-updated MNGP FW this morning to be sure I had the good FW with the same result, did a VMM, nada  .
Maybe I've to wait a few days to see if it will take the right dates for the next days?
Erik

[boB]

First, is your MNGP set to the correct time and date ?  For fairly new firmware, the Classic will not be updated with
the time and date from the MNGP unless the MNGP's year is at least 2012.

Every minute or so, the MNGP tells the Classic what time and date it is if the year is 2012 or greater.

When you go to the TIME/DATE setting menu, you can also look at the time and date that is in the Classic itself
by holding down the LEFT-ARROW key and tapping the SOFT-RIGHT key.  It will show you Classic Time and date
as well as the seconds counting up.

It may be that when you powered the Classic down and up to update the firmware, the battery in the MNGP is not making
good contact (or it's dead ?) and if not making good contact, remove the coin cell battery and poke the two prongs
down with something and then re-insert the coin cell battery.

That battery should be somewhere between 2 and 3 volts.  Actually if it is only 1 or 2 volts you might want to replace it.
It should not be called upon to be used though unless the MNGP has been sitting without power for quite a while like
many months at least.

If the battery isn't keeping the time and date up while you update the firmware, the time/date in the MNGP will
be lost and the Classic won't get set correctly.

boB

[SolarMusher]

Allo boB,
Thanks for your answer on saturday!
I've checked both Classic and MNGP time and date this evening and they are right, so I don't think this problem came from the MNGP battery. My Classic has always had rock and roll datas as soon as I play with firmware, settings, EA, time or VMX  but after a few days without change, datas were accurate.
I'm gonna wait and see, I will report in a few days.
If I need to change the battey, will the MNGP still work without for the time I'll change it?
Thanks,
Erik

[boB]

Allo boB,
Thanks for your answer on saturday!
I've checked both Classic and MNGP time and date this evening and they are right, so I don't think this problem came from the MNGP battery. My Classic has always had rock and roll datas as soon as I play with firmware, settings, EA, time or VMX  but after a few days without change, datas were accurate.
I'm gonna wait and see, I will report in a few days.
If I need to change the battey, will the MNGP still work without for the time I'll change it?
Thanks,
Erik

Everything will still work without the battery.  It just won't remember the time and date when you turn it off and back on again without the battery.You can always set the time and date every time you power it up of course.

When powered up, the Classic sets its year to 2003 before the MNGP tells it what time and date it is.
After that, if something else happens and the date gets reset, the earliest year that the Classic will
reset to is 2000.

You know that if you want to test things out you can force the Classic to a new day without it actually changing
the time and date in the MNGP by going to the LOGS--DAILY menu, holding down the LEFT-ARROW key and
then pressing the ENTER key.  The screen will display  "NEW DAY"  and it will make the Classic think
that it is 23:59 which will increment the daily logs, reset the daily kW-hours on the main screen,
accumulate that into the LifeTime kW-Hours, etc, and force a new BULK/ABSORB cycle in anticipation
of the new day.

Doing this quickly one after the other will, of course, show 0 kW-Hours for the next NEW DAY because
there is no time to accumulate any kW-Hours of energy unless you wait until you get like, 0.1 kW-hours
at least.  But all the other stuff will show up like the date, battery voltage, temperature, etc.  Things that
are always there.

boB

[SolarMusher]

Hi boB,
Checked it this morning and day 1 is correct with 12/15/2012, Day 2 is still poping up invalid with 05/05/2003. So I'm just thinking that it was because I've tortured it with new FW and some news settings  . I'll check it next mornings to see if it'll keep the right date, I'm sure it will. Just one unusual thing, datas displayed are very looow, maybe because it's a little cold this morning in the lodge, we'll see later.
Oh! a great idea for Christmas: a Canadian version of the MNGP which could display datas at very low temps  .
Dogs are running here!
Thanks for your help and detailed explanations, have a great sunday!
Erik

[Resthome]

Okay for the last two versions of Classic Firmware Data Logging has me perplex. The first issue showed up with the previous production firmware 1070 – 8/17/2012.  I had used the latest Local App 0.3.10 to export the Offline data on 11/27/2012 and it contained 250 days of data. On 12/8/2012 I returned to visit the Classic and exported the Offline data using the Local App as I had done previously. This time the data seem to be corrupt. I have index 1-4 with what looks like correct data then 5-9 look like incorrect data kW showing 0 and the Hi Bat only showing 12.7v. Index 10 looks to have correct data then 11-15 the same kW showing 0 and Hi Bat only showing 12.7v. This is repeated numerous times throughout the log. Even the data that previously showed the correct data for 250 days (i.e. No 0 kW data. When I looked at the classic directly those index showed Invalid Date of 2003. Date is correctly set on Classic and clock battery is good since it retains correct date on power down.

I decided to install the 12/01/2012 -1181 Production firmware. Since the logs still showed the same corrupt entries. I decided to start fresh and erase all the log data using Putty terminal emulator with the USB connect and instructions provided by Bob in a forum thread. This worked correctly and all the entries were erased. So I waited for the next day to see if the correct date and data would be written to the log. What I discovered was the data was written to index 1 with the time of 23:59 and to index 2 with the time of 16:23. The 16:23 would have been about the time that I finished the firmware update and did a VMM. So I suspect that may have been written during the VMM or firmware upgrade. I also discovered that index 381 showed the previous days data with time of 23:59. This was directly from the Classic. When looking at Offline Data with the Local App the data showed up in index 1, 380, and 381. With index two days behind and not the previous day data. Index 380 had the previous days data. As a side note the index in the local app goes up to 384. The Offline Data also showed index 292-320 with kW of 6553.5.

While scrolling through the Classic logs I did however notice that the data for the index was not always refreshed even though the index number had changed. This happens most often when fast scrolling backwards from high index number that show invalid date because nothing has ever been written to those locations yet.

I'll have to wait until after the holidays when I return to the site to see if any of the data is corrupt again.

[boB]

The Offline Data also showed index 292-320 with kW of 6553.5.

While scrolling through the Classic logs I did however notice that the data for the index was not always refreshed even though the index number had changed. This happens most often when fast scrolling backwards from high index number that show invalid date because nothing has ever been written to those locations yet.

6553.5  is data with all 1s  (0xFFFF) and obviously did not get cleared when you tried to clear it.  This may be one off the problems.
Another thing that just now occurred to me is that the index for the data within the memory area it is saving to inside the Classic
may itself need to be re-initialized after clearing the data.  I will look into this.

As for the index not exactly following the 0 to 380 (384) points, I know about this issue.  I opted to leave that index
in there even though it is sometimes wrong because it will be very close and when you are somewhere in the middle of
the data, that can be good information.  That will be fixed eventually.

boB

[SolarMusher]

Allo boB,
My Classic is back on the track!
Day 1 date and day 2 date are right, so invalid datas 05-05-2003 have only occured with new FW + VMM and the Classic cell was not dead.
But still have trouble with datas accuracy in local app as it even shows Float time 36 minutes with low Hibatt voltage at 50.4V and several index with 0kw  . I think that each time that I just click on the Loc App config tab, it wants to save something into the Classic and corrupts data. Also, if day 1 is yesterday for the Classic, it seems to be two days ago for local app.
Erik