Update : offgrid drama

softdown · November 11

Week long periodic snows have wreaked some havoc with a few things. BB is following this saga so an update is being posted.

I do have 84 volts available at the Outback FLEXPower One. My installer hooked the main DC power cables into the inverter. This is NOT the way it is supposed to be done. But it has worked since 2015.

Tech support at NAWS will only recognize hooking these wires directly into the charge controller. Clearly the preferential way to do it.

Still can’t get system to charge with panels. "Almost no way” that flipping the switches has wreaked system havoc.

So the charge controller is now the chief suspect. If so, what a coincidence that it chose this particular time to fail. Working since 2015 though - could simply be due.

NOTE: flipping a switch can take awhile to be reflected in MATE readings. And the readings fluctuate as well. Mostly “buying" 2.4 kW from the generator. Sometimes reading zero. Doesn’t seem right.

softdown · November 11

Further examination reveals that hooking the main PV combiner wires into the inverter (PV ground wire) and the auxiliary switches (under the inverter on the FLEXPower) would seem to accomplish the same thing. After all it worked fine for almost ten years.

Right now the charge controller is the chief suspect. I have a back up unit so trying that is likely a 1-2 hour chore on the FLEXPower unit.

Charge controllers can usually be repaired. The control board is less than $200 and looks easy. The power board is almost $500, close to the cost of a new charge controller. Having no idea what the issue is - along with a decade old unit - guess I’ll try replacing it. That should fix the issue.

Weird that all batteries suffered pretty catastrophic damage about the same time. About three weeks ago when I was gone for three days. Some are 2024 batteries. They now all read about the same. Resistance between 10 and 13. Or close to junk. Simultaneously suffered a LOT of false alarms with RING “professional security monitoring”. They kept wanting to call the police though my own security dashboard showed no activity. Theirs sure did - according to them. And some cameras have not worked since. They all worked fine before that experience. Aliens? Bigfoot?

softdown · November 12

Last day the CC worked it seemed to work very poorly. Took hours of sunshine to start working then underperformed badly.

BB. · November 12

Further examination reveals that hooking the main PV combiner wires into the inverter (PV ground wire) and the auxiliary switches (under the inverter on the FLEXPower) would seem to accomplish the same thing. After all it worked fine for almost ten years.

A little confusing... All PV Combiner output (+ and -) to inverter DC input, or just the negative (-) ground wires to inverter DC negative lead, then wired to rest of system?...

As far as I know, the Outback products--It is OK to connect PV Ground (negative) "around" the MX 60 (or whatever DC MPPT charger).

For some (few, many, most?) modern MPPT controllers (Schneider/Xantrex MPPT chargers), for example, do monitor/measure current in the DC array negative leg--And if you "wire round" the PV negative lead (I.e., connect negative array cable to "DC Battery Negative Outback" (which is a "don't care" for Outback), will cause the Schneider MPPT charge controller to operate poorly (if at all).

Again, things have been changing dramatically over the last decade+ of years--The above are examples of a couple older models and may, or may not, apply to newer models and or the hundred Asian Charging/all-in-one systems out there.

Over the years, Outback has usually detailed documentation for their many models... You might find something similar to, and or answer your "best practices" questions:

https://outbackpower.com/resources/documents/wiring-diagrams

-Bill

littleharbor2 · November 12

softdown said:

Weird that all batteries suffered pretty catastrophic damage about the same time. About three weeks ago when I was gone for three days. Some are 2024 batteries. They now all read about the same. Resistance between 10 and 13. Or close to junk.

Can you describe how you test battery resistance?

JRHill · November 12

"I do have 84 volts available at the Outback FLEXPower One. My installer hooked the main DC power cables into the inverter. This is NOT the way it is supposed to be done. But it has worked since 2015. "

"Tech support at NAWS will only recognize hooking these wires directly into the charge controller. Clearly the preferential way to do it. "

What is stopping you from correcting the wiring to the 'preferential' way? Especially before you make repairs to the charge controller.

softdown · November 12

littleharbor2 said:

softdown said:

Weird that all batteries suffered pretty catastrophic damage about the same time. About three weeks ago when I was gone for three days. Some are 2024 batteries. They now all read about the same. Resistance between 10 and 13. Or close to junk.
Can you describe how you test battery resistance?

Using a handheld meter (see pic) that purports to show: Battery voltage, aH output capacity, cca, and battery resistance.

Resistance has always been a high useful measurement for me. Very high quality batteries may have a resistance of 1.5 when new. 2.5 is pretty good for a standard lead acid battery when new.

When the level hits 6.5 in a car that battery is lucky to work for one more year. The next year the level is likely to be ~ 8 and the battery is now junk. Half the time I replace a battery showing 6.5 since the end is near.

Turns out my commercial quality Kubota diesel generator has a horrific sine wave for charging batteries. It was quickly destroying the bank via daily usage. Resistance was up to ~ 35 when charging the batteries. Settling down to ~10 - 13 after charging. The batteries were so bad last night that I considered it likely to be their last day of serviceability. Voltage: 53.8 at 8pm quickly dropping to ~43V at 6am. Junk status.

With a new FM80 charge controller they seem to be healing nicely. Resistance is now down around 6.5 and I expect significant improvement on a daily basis. No idea if they will “heal” enough to work into 2025.

Perhaps it doesn’t matter much. The FullRiver bank turns 10 next year and they see heavy demand over the warmer months. Eight freezers and fridges - long story. I got free food from grocery store throw outs for quite awhile. Irresistible for any serious prepper.

Over reacted to covid in 2020 and bought four 98aH 24 volt lithium phosphates weighing 50 pounds each. But only bought two BMS boards. Boards that are no longer available. Panicked when I recently noticed this and ordered four Daly BMS boards - at an irresistible price. I think - discontinued?

Those batteries turn five next spring and I suspect were a couple years old when purchased. So I have very good reason to press them into service as soon as current banks demonstrate weakness.

My experiment with AGM banks (FullRiver 12v’s) in one room and lead acid banks 6v’s) in another room turned out fine considering the very heavy demand on the system. In addition my panels are oriented heavily towards the preferred winter time spectrum - mounted at 45 degrees at a 37 degree latitude. The system would tend to struggle the last two weeks of August - as memory serves.

I was probably thinking that only one BMS board was needed for two 24v batteries in serial. Unsure if that is true now. I literally forgot almost everything in 4 1/2 years.

I also purchased materials to fabricate my own lithium batteries using a lot of cells. Cylindrical cells that may weigh ~ 4 ounces each? Maybe twice the size of standard “computer cells”? Those were interesting times. Never did take their fabled shots fwiw.

Winding back now. Four weeks ago I recall looking at 6am bank voltages of ~ 48.8 and saying “I’ll take it”. Then the problems started.

Perhaps the old charge controller started malfunctioning ~ 3 weeks ago. That now seems possible.

It was a miserable week. So cloudy the sun could not often be located - for a week. Problems, I thought, with solar array, batteries, generator, internet service, security cameras, household power, and foam door that separates me from an area that averaged 42F last week.

Turns out it was mostly the charge controller and the quality of the sine wave from the generator. The many security cameras have mostly started working again. What an unfortunate time for the charge controller to choose retirement. Good thing I had a spare. Replacement on the FLEXPower One system was easier than expected.

I just duplicated the previous installation. It worked for about ten years. Closer analysis indicated that the result appears the same with the FLEXPower One. I could be wrong though.

System voltage is now 53.2 at 12:15pm. Good for a 43v start. Of course weak batteries charge quickly however. They raced up to ~50v then settled down to normal charging rates.

It will be very interesting to see how much the batteries “heal” from my diesel generators evil sine wave.

littleharbor2 · November 12

I have that exact meter. Didn't know it did resistance. Thanks

Marc Kurth · November 13

Just FWIW:

Be careful with broad brush strokes regarding Internal Resistance or Impedance. Various models of Concorde true deep cycle lead acid AGM batteries have an impedance from 0.24 milliohms to 6.77 milliohms when brand new. For example:

12v, 8D = 1.86

12v, U1 = 6.77

2v, L16 = 0.23

Seeing a change from a known starting point in a given battery is certainly valuable information, but variations of 30% to 40% are normal with new batteries from the same manufacturer.

softdown · November 14

I would think that larger batteries would inherently have more resistance. Much like 100’ of wire vs. 25’.

And different chemistries also vary in the resistance department. Resistance may not indicate the quality of a solar battery. But it certainly indicates the degradation of automotive starter batteries quite well.

Battery resistance fell to 3.5 yesterday afternoon. Things seem to have already normalised.

I’m thinking the charge controller may have been going bad for months. Possible?

The new unit starts delivering a robust charge much earlier - even accounting for the brilliant concept of “daylight savings”.

Update : offgrid drama

Comments

Categories