Classic running warm, now dead

[wrt]

Hi guys,
I have a problem with my Classic, yesterday during a particularly hot day we noticed the air con turn off for no reason. I went to the shed to check on things and noticed my battery voltage was around 74v (48v system) The inverter had shut down showing over volt.

The charge controller was showing 67c at the FETS, 55c PCB. The fans had stopped and all LEDs were blinking slowly inside and out. All numbers were 0 except for in and out volts they were both at 74v.

I shut down the panels, and reset the inverter to drain the excess charge and removed the CCs case, after this the temp quickly reduced after which all fans started and all lights back to normal. All numbers have remained at 0 except for in and batt volts. Inside I see there is scorching on the PCB beside the middle three FETs but nothing else obvious.

After a few measurements it appears that the charge controller has cooked itself and connected the panels directly to the battery's so they have pulled the panels volts down to battery volts and then just kept filling the battery trying to bring it up to panel volts.

What has happened to cause this and can it be fixed or is this controller now a bin job ?

[offgridQLD]

Bump this one... I heared about this on a local Australian forum and it's got me worried to.

Having his charge controller short his panels directly (unregulated) to his expensive lithium battery and overcharge it to the point of swelling up and cooking -  damaging all his near new lithium cells has me very concerned about connecting my two classic 150's  up to my new $10,000 worth of lithium cells (though fortunately i have over voltage protection) I thought the classic wouldn't allow this by design even in a failure event?

Guys how could this have happened (to me its like total brake failure on a car) very serious!

Kurt

[Resthome]

The normal failure mode for all charge controllers is as stated by Robin in one of his posts. I'm not sure I understand the panel to battery issue or why the breakers didn't trip. Maybe boB over Ryan can shed some like on this.

Robin's post:

"Breakers are sized to protect wiring, not electronics. You can adjust the Classic to do many things, but make sure the breakers and wiring is adequate to protect. When a charge controller breaks, they typically short internally from battery plus to battery minus. This condition puts a direct short across the battery bank. You need a breaker that will trip to keep the wires from burning up. By the way, all controllers work this way."

[mike90045]

I guess you can take comfort in having had an abnormal failure ?

[offgridQLD]

Perhaps short was the wrong word.

The issue was his controller failed internally and some how allowed is Pv to continue to feed power into his battery's but the power flow was unregulated ( like having the pv directly connected to your battery) with high voltage array ( or any array above normal charging range things get overcharged real fast.

Kurt

[wrt]

To resthome,
If the charger had of failed in this manner it would have tripped the HRC fuses and saved the battery.
Because there was no short, or excess flow of current it just kept going.

It was as though the Classic got hot and bothered then just threw its hands in the air, and said "screw you guys I am going home" then just left all the other components to keep going without it 

To Mike, the only thing that I can be thankful for is that it happened while I was home and was aware enough of my system to recognise something was wrong and check it quickly enough to shut it down.

Had it not happened like it did the controller would no doubt have continued to increase volts in the battery to the point of catastrophic destruction, possibly taking the rest of the shed with it.

As I stated in the OP it would be good to get an idea of what could have gone wrong and if it can be fixed.

[kitestrings]

Yes, count your stars that you were there and caught what was going on.  Hopefully your batteries were not damaged.  ~ks

[Westbranch]

I see you said that only Volts were showing, but nothing else...  What is your amperage (configuration) potential from your array?

[offgridQLD]

Yes, count your stars that you were there and caught what was going on.  Hopefully your batteries were not damaged.  ~ks

His cells reached 4.8v per cell 76.8v (20v overcharge) for a 48v bank. They hit 48.7c and swelled up like a balloons there is defiantly some damage.

Kurt

[wrt]

My array is 5.7kw, so has a potential of around 100amps depending on charge state. However because it is a multi orientation array. It's actual peak is 4kw.

Just before the incident it was on float at 700w. The batteries where already full.

[boB]

Is  offgridQLD  and wrt talking about the SAME thing ?  Or are they different ?  If FLA batteries that would be
74V/24 = 3.1 V per cell...   Or if 4.8V for Lithium ?

I am not sure what happened....  I would have said that the processor went nuts or crashed but you  (wrt) mentioned
some scorching near the middle FETs.  One reason I tend to think it was the processor is that
if an input FET shorted AND the relay stuck, the FETs will usually ALL short out and would trip the battery breaker.
That is how most MPPT charge controllers fail.

This is the first time I have heard of THIS happening, BTW, so this would be extremely rare, luckily.
Was it a Classic 150 ?  How long has it been working ?  What serial number or when was it built ?

I would hope that we hear from you, WRT tomorrow at least for RMA info.  Were the batteries hot ?
If this did not go on for too long, then the batteries I would think would be OK.  Nonetheless, this
is scary !

boB

[wrt]

Hi Bob, we are both talking the same incident, they where LiFePo4 pumped to 4.8v. Just about to leave for work so will post a pick of the FETs when I get home and figure out how to. The barcode on the side of the unit is CL02868. Not sure if that is the number you are looking for.

Will post again in 9hrs when I get home. Am driving to Melbourne tomorrow for the distributor to inspect.

[boB]

Hi Bob, we are both talking the same incident, they where LiFePo4 pumped to 4.8v. Just about to leave for work so will post a pick of the FETs when I get home and figure out how to. The barcode on the side of the unit is CL02868. Not sure if that is the number you are looking for.

Will post again in 9hrs when I get home. Am driving to Melbourne tomorrow for the distributor to inspect.

Thank you !  Yes, that was the number I was looking for.    It is older but still not good.

Any charge controller can do this if they break but we all try to make this particular failure mode as rare as possible.
We have now shipped tens of thousands of Classics and yours (AFAIK) is the first and only one to have shown
this particular problem.  Also, 67 degrees temperature is nothing for the FETs...  they can actually run past 100 degrees C
for  a long time but we shut them down before then.   I AM curious how hot the batteries got though ?
Usually, the BMS should protect for over-voltage since there is an active control system.  Did this happen ?

[offgridQLD]

Hi Bob,
        Yes I was talking about WRT's indecent I had some backround on the insedent from another  posting of his on a local forum. His cells are Lifepo4 lithium with a nominal voltage of 3.2v pr cell and he uses 16 cells (typically the cells sit  around 3.3v - cell 52 -53v for 16 cell bank) When charging you usualy pull the pin and drop to float at around 3.5 -3.6v pr cell . 3.8v - 4v pr cell is usually when you should shut things down and see what went wrong befor you kill something. His cells got to 4.8v pr cell.  The cell casing start to swell up like a balloon and typically this turns out to be total loss or reduced capacity (that's if you can successfully clamp the cells flat again)

this is what his cells look like now all puffed up and swollen.


BMS or BCU systems tipically do have large contactors and individual cell monitoring units that pull the contactors in low and high voltage events for any cell. (The BCU system I am installing on my lithium has both high and low) Though WRT's didn't have a high voltage protection he was just trusting the charge controller. (I was thinking of doing the same at first) The classics have been 100% reliable for me for several years and rock solid voltage regulation.

I just mentioned to WRT's to let you guys now about it as it did look like a rare event and one that is worth investigating.

My own personal interest in WRT's event is I run two classic 150's similar climate (though a little lower extreems in the part of Australia where I am) I run a similar largish system (4000w) of pv on each classic and one of my classics has 3s configuration so high VOC and I notice that unit runs a lot hotter than the other classic with 2s panels and lower voltage to down convert to 48v. After this event I am considering reconfiguring the pv to 2s to help that classic run cooler (that's if WRT's classic failure had anything to do with heat) I'm also about a week away from upgrading my old lead acids to 16x 400Ah lifepo4 lithium cells the same as WRT was using so some interest stemmed from there in the event. (lets just say I had a few nights ofter hearing his story when I woke up in a cold sweat and my wife mentioned I was mumbling something about swollen- pregnant lithium battery cells and possessed classic 150's ....

Kurt

[boB]

Hi Bob,
        Yes I was talking about WRT's indecent I had some backround on the insedent from another  posting of his on a local forum. His cells are Lifepo4 lithium with a nominal voltage of 3.2v pr cell and he uses 16 cells (typically the cells sit  around 3.3v - cell 52 -53v for 16 cell bank) When charging you usualy pull the pin and drop to float at around 3.5 -3.6v pr cell . 3.8v - 4v pr cell is usually when you should shut things down and see what went wrong befor you kill something. His cells got to 4.8v pr cell.  The cell casing start to swell up like a balloon and typically this turns out to be total loss or reduced capacity (that's if you can successfully clamp the cells flat again)

this is what his cells look like now all puffed up and swollen.


BMS or BCU systems tipically do have large contactors and individual cell monitoring units that pull the contactors in low and high voltage events for any cell. (The BCU system I am installing on my lithium has both high and low) Though WRT's didn't have a high voltage protection he was just trusting the charge controller. (I was thinking of doing the same at first) The classics have been 100% reliable for me for several years and rock solid voltage regulation.

I just mentioned to WRT's to let you guys now about it as it did look like a rare event and one that is worth investigating.

My own personal interest in WRT's event is I run two classic 150's similar climate (though a little lower extreems in the part of Australia where I am) Irun a similar largish system (4000w) of pv on each classic and one of my classics has 3s configuration so high VOC and I notice that unit runs a lot hotter than the other classic with 2s panels and lower voltage to down convert to 48v. After this event I am considering reconfiguring the pv to 2s to help that classic run cooler (that's if WRT's classic failure had anything to do with heat) I'm also about a week away from upgrading my old lead acids to 16x 400Ah lifepo4 lithium cells the same as WRT was using so some interest stemmed from there in the event. (lets just say I had a few night when I woke up in a cold swett and my wife mentioned I was mumbling something about pregnant lithium battery cells and possessed classic 150's ....

Kurt

I/We are very interested in what exactly happened here.   One thing I was thinking was that if something like this were to
happen again, a noise maker/buzzer or additionally even a relay/contactor of some sort could be connected
to the to one of the Aux outputs in a mode that would turn on when the battery voltage got above some preset
value for at least a second or two or three.  That would keep the plumping from happening.

I am also very interested in the one poster's comment about compressing the plumped battery to get it to
work right again.  The attitude there was that it does work but I have not had enough experience with those
LiFePo batteries to know how well that works.  Might be that is works great ?

67 degrees C is just not very hot.  Remember that the die inside the FETs are rated for 175 degrees C
or 150 C.

boB