Sorry, me again.
day before yesterday, I was surprised at the Classic again, the SOC% had shot up unbelievably much.
But I felt I had reported often enough, and decided to not mention it and keep my eye on it.
However, yesterday late evening (already dark), SOC% was at 74%. This morning it is at 99%.
Take my word for it, there was no bright sun overnight :P
Also, it is at 23.9V with only 6Amp. On a 460Ah battery also a hint the 74% is more realistic.
I did activate auto-eq (using the local app) but did not notice it going to 100% as result of that. Also timing is wrong for that, if that had caused it, it would have counted down all night and be lower then 99%. The 1% loss hints it may have reset to 100% only an hour or 2 ago. Alas I am not on board, so I cannot check if the 100% may be blinking on the classic itself.
Something is up with the mechanism that resets the SOC%. And as I use it to activate load, it is very important to my use case.
And if I look back, I think my other weird experiences also are probably related to this.
Edit: I went and checked: the 99% was *not* blinking. So it was not some accidental "turn off classic" event.
Mine does that when the charger kicks in overnight if the battery voltage dips too much. My whizbang jr will count up to almost 100% because I cut the absorb time short from the charger. Not sure if that is the same thing you are seeing or not.
No, most definitely the charger did not kick in. Not only is the charger off (I checked) but I would have notices the higher battery voltage during charge in mynidnite.com (checked that too). And as mentioned a higher battery voltage post charge.
It resets to 100% at (for me) unpredictable and undesirable moments. That I am now sure of.
Bummer. My is working flawlessly since the 1779 firmware update.
cybermaus
So you don't feel alone. When I first installed 1779 I had some very strange resets of SOC back to 100% during the night. I was only there for two night so was not able to verify or see if it repeated or if occurred around the 23:59 data save or the Auto Restart that I had it set to. Unfortunately to save on the battery bank I do not run the LA at night and I have no Internet connection for MyMidnite. I think I disable the Auto Reset before I left but not sure. Also using Skip Bulk so I don't overcharge when gone for weeks at a time. I was also trying to figure out the Net-Amp reset as it did not match my Link 10. After the thread on SOC Questions boB confirm the two were not tied to each other and used separate counting methods. Also using the LA 0.3.52
I'll be headed back up tomorrow for 4 days, but I also have a control board to change in the Classic to see if it resolves my Float Times being zeroed out 256 days in the past. Will not be able to verify if that is fixed for 256 days so it's just a try this kind of thing.
I also have one of the very early WBjr and it has a long lead going to the Classic from the Shut location. I understand a component was changed in latter WB JRs, believe it was a resister. So there a lot of variables.
I'll post back my findings with SOC next week,
Quote from: cybermaus on March 12, 2014, 02:11:30 AM
Sorry, me again.
day before yesterday, I was surprised at the Classic again, the SOC% had shot up unbelievably much.
But I felt I had reported often enough, and decided to not mention it and keep my eye on it.
However, yesterday late evening (already dark), SOC% was at 74%. This morning it is at 99%.
I had suspected some anomaly with SOC for awhile. A couple times I have seen quick, unexplained 100% values from lower values just too quickly.
This A.M. while under a couple hundreds watts of Solar charging after a night of extreme deep discharge with the SOC at about 60% (don't ask). Suddenly it was at 100% in a few minutes and less than a KWH of incoming recorded into a 450 AH 24 volt bank that is quite tired.
Just because I saw it happen and it is obviously wrong in my case. I am using FW 1758. I thought I was current?
Tom
I'm having the exact same issue as Tom, with the exception I'm only down around 82% when it went back to 100% instantly.
In bulk, a few hundred watts in from panels, firmware 1779. I took a screen shot if it would be of any use.
Later
W
Could it be that any of these cases coincide with the Classic rebooting?
Quote from: Halfcrazy on March 13, 2014, 06:27:03 AM
Could it be that any of these cases coincide with the Classic rebooting?
Not in my case. My local logging would have shown the KWH value reset and it didn't.
Tom
Quote from: Halfcrazy on March 13, 2014, 06:27:03 AM
Could it be that any of these cases coincide with the Classic rebooting?
I was wondering that. Is there any place I can see the runtime?
For now, I am assuming it was not a reboot, because on a reboot the SOW% goes to blinking. I already seen it for certain twice that it shot to 100% too soon and was not blinking, and I have at least two more suspected events. (if I reread my previous reports)
But if You can tell me where I can see the current runtime since last boot, I can make sure to check that next time.
Hmm, I may have seen a link with EQ.
Not sure if it is always linked, but it was linked this one time.
Late this morning it reached FLOAT, and this afternoon I put it in manual EQ. But maybe the sun was already to low, I do not think classic was able to finish it. Anyway, several hours later (when the 1 hour EQ should have been long finished?) when it was already dark outside, it was still reading EQ MPPT. Also, as power use was apparently high this afternoon, the SOC% had dropped back to 96%
So I pressed the "stop" EQ button myself. Local app showed went to RESTING, and the SOC% shot to 100% again!
Quote from: cybermaus on March 13, 2014, 01:57:04 PM
Hmm, I may have seen a link with EQ.
.......
So I pressed the "stop" EQ button myself. Local app showed went to RESTING, and the SOC% shot to 100% again!
Yes, that makes sense. Whenever the Classic goes from EQ or Absorb to Float, it will reset to 100% automatically.
You forced a Float by stopping the EQ. It could have either gone to Absorb if it was in Absorb when you started the EQ or it would have gone to Float if it was in Float before you started the EQ and then forced the EQ to STOP.
I suppose that if one "forces" a float by either pressing force float or stop EQ that it ~could~ hold off resetting to 100% SOC but that is the way it works right now.
So there are only 2 ways the SOC will jump to 100%...
One is that it goes to Float for whatever reason... It is assumed that the Classic goes to Float because the battery is Full.
The second reason to jump to 100% (and then count down from there) is when the Classic resets.
THAT we are working on still. But we are definitely working on it.
BTW, there will be some cases where the newest code up on fusion.midnitesolar.com will actually FIX some of those resetting issues ! It fixes the problem where the Classic resets when you are reading offline data logs either with the Local App or MNGP (graph display mode) while the Classic saves the data logs every 5 minutes AND also if you are sitting in the wind power curve graph while the Classic saved the minutely data logs every 5 (or 10) minutes.
I'm sure that at least some of the resetting has to do with the two data reading and writing to/from logs coinciding at the same time.
That code, if you want to download it, is http://fusion.midnitesolar.com/MidNiteSolarUpdate_1794_3-12-2014_v4.28.exe
You will want to do a VMM after installing it though because there is a new feature that will stop charging (reason for resting = 36 and 136) if the battery temperature goes above a new modbus register value and there is not guarranty that that value is going to be in the correct range at first until initialized.
More updated coming if you want to wait.
boB
boB
Quote from: boB on March 13, 2014, 06:43:13 PM
So there are only 2 ways the SOC will jump to 100%...
One is that it goes to Float for whatever reason... It is assumed that the Classic goes to Float because the battery is Full.
Bob, in days between bulk, what happens there. While in float (non float mppt) and WBJr EA is met might be another case for 100% ?
Quote from: zoneblue on March 14, 2014, 12:34:10 AM
Quote from: boB on March 13, 2014, 06:43:13 PM
So there are only 2 ways the SOC will jump to 100%...
One is that it goes to Float for whatever reason... It is assumed that the Classic goes to Float because the battery is Full.
Bob, in days between bulk, what happens there. While in float (non float mppt) and WBJr EA is met might be another case for 100% ?
Ending Amps should not have any effect on SOC% while in Float or EQ. Only Absorb.
"should not" of course is the phrase here... If you do find some interaction, I would be
very surprised but I have been surprised before.
I have a feeling the unit has reset for some reason or gone TO float.
boB
Quote from: boB on March 13, 2014, 06:43:13 PMThe second reason to jump to 100% (and then count down from there) is when the Classic resets.
THAT we are working on still. But we are definitely working on it.
You are working on preventing reset. I understand finding the reset causes is not an easy fix and may take a while and a few tries.
Can you at least make it so that after a reset / reboot it automatically goes to a blinking version of the last known value instead of a solid non-blinking 100%(as suggested in another thread)
Also, is there a place to see the current runtime, iow, when it last reset?
Thanks
In the weekend i added the soc register (4373) to blackbox. I can confirm that SOC is broken with the following combination of settings:
- days between bulk=5
- reset net Ah on charged=yes
In the chart below the stg column is the raw decimal charge state register. Upon waking for the first time this morning, both netah and soc reset.
Quote from: zoneblue on March 16, 2014, 03:35:31 PM
In the weekend i added the soc register (4373) to blackbox. I can confirm that SOC is broken with the following combination of settings:
- days between bulk=5
- reset net Ah on charged=yes
In the chart below the stg column is the raw decimal charge state register. Upon waking for the first time this morning, both netah and soc reset.
I don't quite understand your spread sheet and why it appears to go to 100% at 6:27:55 but I do
see why it resets to 100% when you have days between bulk/absorb set to skip days. (non 0)
The problem is that would happen would be, every new day that it is NOT supposed
to go to Bulk/Absorb, it will go to Float. It would do a force-float and that is where the SOC = 100%.
I think I have just fixed this.
So, NOW, it only resets to 100% when it goes from Absorb to Float because of a full battery.
There is also now a modbus Force Flag that allows one to force the SOC to 100%
Regarding Cybermaus's question about run-time since reset, yes there is. It is called UpTime and
is in seconds and is at modbus register addresses 4348 (Low) and 4349 (High) or registers
4349 and 4350.
boB
PS... ZoneBlue. Please try this version and see if it fixes the bulk-skip SOC = % problem
http://fusion.midnitesolar.com/MidNiteSolarUpdate_1795_3-17-2014_v4.29.exe
I can verify what boB is saying about the reset. I just got back from the lake. Sure which I had an internet connection while I was there (water is too low to get a good signal). My SOC resetting to 100% was being caused by the Classic going to resting and the RFR was 104 - Watchdog Timer. I talked to Matthew at MidNite before I left the lake but he was not aware of the new update. I could not pin down what was causing the Classic to reboot but I suspect as boB said it had something to do with the log data. Did capture the Debug settings, don't know what they mean or if they may be useful.
Hm SOC is undefined that's a new one? I have not seen that one before. boB will have to comment but the new code has one fix for the classic rebooting when the logs are being read and written at the same time.
Ryan
Here is the one with the debug readings after one of the RFR - 104 codes.
Do they mean anything?
Ryan
Not sure what was going on with that undefined SOC. This was in Absorb and it is okay here. Not sure if it go confused with one of the resets.
Bob, do you want the good news or the bad news?
The good news is that with 1795, SOC doesnt reset in the morning.
The bad news is that it also doesnt come out of sleep in the morning:
Classic 150V (rev 4)
MAC: 60:1D:0F:00:1B:95
Device ID: 16F0 BE2
IP: 192.168.0.223:502
Firmware:
- Classic Rev: 1795
- Network Rev: 1674
Absorb Time:03:00:00
Equalize Time:01:00:00
Last Voc:84.2
DB1: 0x0 DB2: 0xffff
DB3: 0x0 DB4: 0x0
Classic 150V (rev 4)
Solar(On)
RESTING (36:Unknown)
Whizbang Jr.
System: -4A Net Ah: 58Ah
Remaining Ah: 322Ah
SOC: 88%
Ive tried setting days betweeen bulk to 0, force float, and bully reboot, still RFR 36.
I will try a lite VMM in case that makes a diff, shortly, and failing further advice, revert to 1779.
Quote
PS... ZoneBlue. Please try this version and see if it fixes the bulk-skip SOC = % problem
http://fusion.midnitesolar.com/MidNiteSolarUpdate_1795_3-17-2014_v4.29.exe
OK, good. I know what the RFR = 36 is.
An Over-Battery-Temperature "feature" was added and so it will need
a Vulcan Mind Meld to set the register. Register 4262 or address 4261
is the reference temperature. Your Classic is probably sitting at a very
low number right now. If you could set it to, say, 1700 (170.0 deg. C)
that should fix it.
I'm not sure if your newer MNGP will VMM it properly or not. The hardware
VMM, placing the two outside jumpers on within 1 minute of re-boot of the
Classic will also fix that but will of course reset all of your settings.
Coming out with another MNGP revision which will also allow you to manually
set the battery SOC to 100% in the WB Jr. --- MORE menu by holding down
the LEFT ARROW key and the ENTER key. Coming right up.
boB
boB
Is it possible to get an updated list of current RFR codes for the latest firmware? I know there have been some additions.
Great, hardware VMM fixed it. Will put days betw bulk back in and let you know how it goes.
Thanks!
BTW could we have 1dp on the SOC register, looks kinda cranky.
Quote from: boB on March 18, 2014, 03:56:12 PM
OK, good. I know what the RFR = 36 is.
An Over-Battery-Temperature "feature" was added and so it will need
a Vulcan Mind Meld to set the register. Register 4262 or address 4261
is the reference temperature. Your Classic is probably sitting at a very
low number right now. If you could set it to, say, 1700 (170.0 deg. C)
that should fix it.
Quote from: zoneblue on March 18, 2014, 04:17:05 PM
Great, hardware VMM fixed it. Will put days betw bulk back in and let you know how it goes.
Thanks!
BTW could we have 1dp on the SOC register, looks kinda cranky.
Cranky ? Not sure what you mean ? I think you want one decimal point added ?
boB
Quote from: Resthome on March 18, 2014, 04:00:47 PM
boB
Is it possible to get an updated list of current RFR codes for the latest firmware? I know there have been some additions.
Sure...
ReasonForResting = 1 Wake state, (Vpv < PreVoc AntiClickSenstvty (MB Addr. 4236)
ReasonForResting = 2 Insane Ibatt on WakeUp state (offset changed from off state)
ReasonForResting = 3 Negative current on WakeUp state
ReasonForResting = 4 dispavgVpv < (dispavgVbatt - 10) Now -25 (RestartTimerms = 1500)
ReasonForResting = 5 Too low power and Vbatt below set point for 90 seconds
ReasonForResting = 6 FETtemperature >= 100C Hot
ReasonForResting = 7 Ground Fault
ReasonForResting = 8 Arc Fault
ReasonForResting = 9 (IbattDisplaySi < -15) (negative current) (MB 4200)
ReasonForResting = 10 (dispavgVbatt < LBDlowV) Battery less than 8 Volts
ReasonForResting = 11 Vpv >= 90% of Voc but slow. Low Light #1
ReasonForResting = 12 Vpv < 90% of Voc Low Light #2
ReasonForResting = 13 Vpv > (Voc + 10V) in PV_Uset || Solar1_OandP
ReasonForResting = 14 Vpv >= 90% of Voc but slow. Low Light #3
ReasonForResting = 15 Vpv < 90% of Voc and taking too long. Low Light #4
ReasonForResting = 16 Normally because user turned MODE OFF... Disabled
ReasonForResting = 17 Vpv > 150V (classic 150)
ReasonForResting = 18 Vpv > 200V (classic 200)
ReasonForResting = 19 Vpv > 250V (classic 250)
ReasonForResting = 22 Average Battery Voltage is too high above set point (RestartTimerms = 2 sec)
ReasonForResting = 25 Battery breaker tripped (Vbatt shot up high)
(If RFR = 25 on Wakeup, check modbus register 4200)
ReasonForResting = 26 Mode changed while running, Vabsorb raised more than
10.0 Volts or Nominal Vbatt changed by modbus command
AND MpptMode was ON when changed...
ReasonForResting = 27 bridge center == 1023 (R132 might have been stuffed old units)
ReasonForResting = 28 NOT Resting but RELAY is not engaged for some reason
ReasonForResting = 29 ON/OFF stays off because WIND GRAPH is insane
ReasonForResting = 30 PkAmpsOverLimit (will change somewhat 1-23-2013)
ReasonForResting = 31 AD1CH.IbattMinus > 900 (peak negative battery current)
ReasonForResting = 32 Aux 2 Logic input is high. Aux2Function 15 (external disable/enable)
ReasonForResting = 33 OCP in a mode other than Solar or PV-Uset (1-10-2013)
ReasonForResting = 34 AD1CH.IbattMinus > 900 Classic 150,200 newer than 1-23-2013
ReasonForResting = 35 Vbatt < 8.6 V (LOW LOW battery)
ReasonForResting = 36 Battery temperature is Greater than reg address 4161 specified
ReasonForResting = 136 Battery temperature fell below MB reg. 4161 - 10 C (Classic turned back on)
ReasonForResting greater than 100... 100 + PowerOnReset, WDT, etc...
ReasonForResting = 104 Watchdog WDT reset (only at boot until first RFR)
ReasonForResting = 111 Normal Power up boot (only at boot until first RFR)
[ 100 + 1 = POR, 2 = Ext. Reset 4 = WDT 8 = Brown Out ]
Thanks boB :)
Quote from: boB on March 18, 2014, 07:21:58 PM
Cranky ? Not sure what you mean ? I think you want one decimal point added ?
Yeah 1dp would be swell if that doesnt complicate anything.
Hey, the SOC reset looks good. Didnt reset this morning, (todays a skip day). I was also suprised to see no reset on any of the reboots both bully and breaker yesterday morning. Appears to saving to eprom not just at midnite...
Excellent ZB ! That's what I wanted to see !
Not sure about extra digits. We'll see though.
Thanks,
boB
Quote from: zoneblue on March 19, 2014, 04:49:16 PM
Hey, the SOC reset looks good. Didnt reset this morning, (todays a skip day). I was also suprised to see no reset on any of the reboots both bully and breaker yesterday morning. Appears to saving to eprom not just at midnite...
Nice job boB.
Hi boB
You did not direct the new download to me (and also I was In Germany for a few days) but I assume I should also load the latest build?
Maurits
Well i can confirm its stable here, just be sure to VMM after you install or it will RFR.
Quote from: cybermaus on March 19, 2014, 06:10:52 PM
Hi boB
You did not direct the new download to me (and also I was In Germany for a few days) but I assume I should also load the latest build?
Maurits
Maurits. Please, now would be the time !
Thanks, guys.
boB
After a week with 1795 SOC remains rock solid.
On the reset front. I havent heard any spontaneous reboots during this time
Quote from: zoneblue on March 24, 2014, 05:36:06 PM
After a week with 1795 SOC remains rock solid.
On the reset front. I havent heard any spontaneous reboots during this time
Good to hear !
Well, I now know how to definitely make the Classic reboot, but it is (hopefully) a very
rare occurrence that Resthome recently found.
I get a "SMA" on screen proceeding bulk,absorb or float. What is that?
Quote from: philb on March 25, 2014, 03:06:36 AM
I get a "SMA" on screen proceeding bulk,absorb or float. What is that?
Oh, you've accidentally entered the SMA mode... It is for when the Classic is being
told what to do from an SMA Sunny Island inverter through one of our comm-boxes.
Did you do a "Vulcan Mind Meld" when you updated your code ? I am assuming
you updated code in the Classic maybe ?
To get rid of that, to into the MNGP --- MISC --- COMM --- MORE menu
and change the SMA to F-ME and then press ENTER.
Re-boot the Classic at some point after doing this when you get a chance. Usually
you won't have to reboot but it won't hurt to make the transition complete.
When did the SMA mode stuff start coming up I wonder ?
boB
The SMA mode came up right after I installed my router and whizbang. That's been about a month ago. VMM was done after reboot.
I was poking around in the MORE area checking the settings for errors. I did have a few in other places. Must have fat fingered that one. I'll fix that tomorrow. It doesn't seem to effect anything that I'm aware of. What does F-ME stand for, BTW?
Thanks boB.
I am having problems updating to 1795. I tried last week, but I was in a hurry and it did not work.
Tried again just now, and it still did not work.
When updating the MNGP, it all works OK. Connected to COM8, runs up to 100%, no issue.
But when next doing the Classic itself, it now sits as if COM8 does not exists.
I tried rebooting PC, etc. The USB driver is there, and it nicely shows up in the device manager with COM8 assigned. But the update software simply does not see the port, no matter how often I restart.
So at this moment, the panel shows 1758 for the classic and 1795 for the MNGP. Not sure if that can hurt.
Also, I called the support hotline a few times. But I got a recording about office hours both times. (and I was well within office hours)
Any hints?
Im not sure what firmware you upgraded from, but some folk had issues about two betas back if they upgraded the MNGP first, it would cause issues., and doing the classic first was the better way. But i dont recall closing the com port being one of them.
Have you tried rebooting the pc, checking the driver and port and ok, then just doing the classic.
My take on it is that it will usually be better for the mngp to not know about new classic code then the reverse. If that makes sense.
Quote from: cybermaus on March 26, 2014, 05:16:29 PM
Also, I called the support hotline a few times. But I got a recording about office hours both times. (and I was well within office hours)
Any hints?
Feel free to call my cell phone 207 416 2006. I am pretty busy lately and apologize. I have another employee starting Friday to help me. Please feel free to call me at anytime on my cell. If you get my voice mail leave me a message and I will do my best to get right back to you. I am on the road all day tomorrow but will have cell coverage all the way.
Ryan
Well, I had left the boat already. Its been a bit crazy at work, so I do not have the time I would like to to test this new firmware.
I hope I can make time for another attempt today.
Thanks for your response though.
Strange. I tried again today, this time it simply worked.
And I am *absolutely* sure I did the same thing.
And yes, yesterday I had also rebooted and checked the driver, several times.
Anyway, I am finally on the latest beta.
FYI: been running 1795 for a week now.
No unexpected fluctuations of the SOC%, no jumping to 100%.
Also, just to test, to day I requested a EQ, and stopped it again 5 minutes later, and it stayed at 95%. Next I rebooted, and again it stayed at 95%.
Very nice, thanks.
Edit: Not that I think it is in any way important, but I did just notice that if you reboot, the daily KWh value is reset.
Quote from: cybermaus on April 04, 2014, 09:21:43 AM
FYI: been running 1795 for a week now.
No unexpected fluctuations of the SOC%, no jumping to 100%.
Also, just to test, to day I requested a EQ, and stopped it again 5 minutes later, and it stayed at 95%. Next I rebooted, and again it stayed at 95%.
Very nice, thanks.
Edit: Not that I think it is in any way important, but I did just notice that if you reboot, the daily KWh value is reset.
Yep and it does not get added to the totals either. One way to avoid that is the do a NEW DAY save and it will get recorded and added to the total before it resets and starts over.