Worried about voltage

[Vern Faulkner]

So ... The Lady and I are planning on going off-grid in three or so years. In advance, we've purchased some assets to take the living room off-grid this winter, to gain a real sense of how much power we consume, how much "real world" juice we can generate from our system.

We currently have a pair of 240w panels, in series, feeding into a Midnite Classic 150 MPPT controller. That's feeding a pair of 6V, 370Ah Trojan batteries wired in series.

The second panel and the controller are a few days new: previously had a single panel feeding through a cheap-o controller into the batteries. That setup was working for about a month, maybe, with light drain.

We have a generic 12v modified sine wave inverter powering the living room computer and a light. The entire batch, at worst, draws 140 Watts.

The last few days, the controller has pushed 700-900 watts into the system daily. However, the voltage reading on the batteries seems to plummet sharply under minimal use.


For example, I left the house at 6:30 p.m., with the voltage at 12.9. The Lady was watching "Bachelor" for three hours, online. Max draw, call it 140 watt-hours. Voltage was down to 12.5. Now, after about four hours continual use, the voltage is down to 12.1, which by my count is quite discharged.

My math says the wife has really only drawn, at max, 600 watt-hours out of the system - probably closer to 500 watt-hours, or, say, 40 amp-hours. That should really only represent about 12-15 per cent of the total charge of the 370Ah batteries, but it seems to be representing almost a 75 per cent discharge, if my knowledge of the voltage readings is accurate.

What's worrying me is that the daily input is far exceeding the daily draw, and yet the batteries seem ill-charged. I've set all the voltages on the charge controller accurately, and it is dumping as much as 380 watts - and usually 20-25 amps - into the batteries around noontime.

Thoughts?

[Sunking]

The last few days, the controller has pushed 700-900 watts into the system daily. However, the voltage reading on the batteries seems to plummet sharply under minimal use.

OK first thought is this number is meaningless did you mean watt hours?

Second thought is what is your location?

Lastly where are you measuring the battery voltage, and are you taking it under load?

[Vern Faulkner]

OK first thought is this number is meaningless did you mean watt hours?

Second thought is what is your location?

Lastly where are you measuring the battery voltage, and are you taking it under load?

a) Yes: 700-900 watt-hours in per day. Almost a kwh today, in fact.
b) The inverter is saying 12.3 volts, the charge controller is measuring 12.1, and the load seems to matter little. Natch, there is a .1 or .2 v rebound after the load is removed ...

Edit: Location is St. Stephen NB. Panels are not in an ideal location, will be re-locating this weekend. Been super sunny, tho, the last few days.

[Sunking]

OK then you are not measuring the battery voltage. You need measure the voltage directly on the battery post with a good accurate DVM. One thing that really strikes me is the Charge Controller output has to be equal to or greater then the batteries. It is not possible to be lower unless the metering is inaccurate.

So how are you calculating the Energy (watt hours) out of the controller?

[Vern Faulkner]

OK then you are not measuring the battery voltage. You need measure the voltage directly on the battery post with a good accurate DVM. One thing that really strikes me is the Charge Controller output has to be equal to or greater then the batteries. It is not possible to be lower unless the metering is inaccurate.

So how are you calculating the Energy (watt hours) out of the controller?

Right from the controller data. The Midnite 150 is a pretty data-heavy beast. There's no issue with the charge controller: it's putting more than enough amps into the batteries. What's worrying me is that the rate of voltage decline seems ... odd ...

[Beanyboy57]

Just a few things for you to check and confirm.
How old are the batteries? Did you buy them new? Could they be heavily sulphated? If they are FLA have you maintained the fluid level?
Which posts have you wired the CC and the Inverter to? (First positive and last negative or vice versa?)
Have you been running the batteries to more than 50% DOD for the past month or more?
Have you measured the individual cells of the battery with a digital meter to see if all of them are functioning correctly?

[Vern Faulkner]

Just a few things for you to check and confirm.
How old are the batteries? Did you buy them new? Could they be heavily sulphated? If they are FLA have you maintained the fluid level?
Which posts have you wired the CC and the Inverter to? (First positive and last negative or vice versa?)
Have you been running the batteries to more than 50% DOD for the past month or more?
Have you measured the individual cells of the battery with a digital meter to see if all of them are functioning correctly?

a) Brand new; b) purchased from a reputable renewable energy retailer.
c) No signs of degradation
d) fluid levels are bang-on
e) Haven't been abusing the battery with discharges ... discharged deeply (12.0 volts) once, then set to charge for a good two weeks without current draw.

[Sunking]

Vern I was no timplying there is anything wrong with the controller. I own and operate the same model. However the voltage you stated on the Controller output is not possible. It cannot be lower than the battery. That tells me the Controller meter is in accurate.

Now with all that said, you cannot get accurate State of Charge voltages on a working system. The batteries have to be disconnected and rested for several hours. Even then SOC measurements are only an indicator and not accurate.

If you really want to know what is going on is obtain a good quality temp corrected battery hydrometer. That is the only true way to determine th ebattery state of charge. You also need a good quality DVM and take your voltage readings on the battery post. With a good meter with th epanels producing full power, and a load on the system you should see voltages something like:

Controller output = 13.9
Battery Post = 13.8
Input at Inverter = 13.7

[Vern Faulkner]

Vern I was no timplying there is anything wrong with the controller. I own and operate the same model. However the voltage you stated on the Controller output is not possible. It cannot be lower than the battery. That tells me the Controller meter is in accurate.

Now with all that said, you cannot get accurate State of Charge voltages on a working system. The batteries have to be disconnected and rested for several hours. Even then SOC measurements are only an indicator and not accurate.

If you really want to know what is going on is obtain a good quality temp corrected battery hydrometer. That is the only true way to determine th ebattery state of charge. You also need a good quality DVM and take your voltage readings on the battery post. With a good meter with th epanels producing full power, and a load on the system you should see voltages something like:

Controller output = 13.9
Battery Post = 13.8
Input at Inverter = 13.7

I think I may have misled you: the battery voltages I'm giving you are after the charging cycle. I do not think the inverter voltage reading is accurate, or in other words, I believe I'll take the Midnite 150 reading over a $99 Canadian Tire 1,500 inverter reading.

However, the drop is the same - 0.8v in four hours, with only 10-12 amps per hour over that time.

I'm not seeing close to 13.8, though. I have the charging voltage set per the Trojan manuals, which, IIRC (not at home) is 14.4v (but I may be recalling wrong). The highest I've seen is 13.7v on the batteries.

Could it be, perhaps, that the things are simply massively undercharged? Should I take the load off for a day or two?

[Sunking]

I'm not seeing close to 13.8, though. I have the charging voltage set per the Trojan manuals, which, IIRC (not at home) is 14.4v (but I may be recalling wrong). The highest I've seen is 13.7v on the batteries.

Could it be, perhaps, that the things are simply massively undercharged? Should I take the load off for a day or two?

Sounds like they are severely undercharged, that is why I keep pounding the hydrometer point. It is the only way to measure SOC.

During the Bulk and and Absorb charge you should be seeing around 14.4 volts out of the controller near the end of Absorb cycle, then fold back to 13.6 in float. If I had to guess is you do not have enough panel wattage to recharge your batteries/ and/or you are over discharging your batteries daily and not able to be fully recharged the next day. Either way it means the same thing.

[Vern Faulkner]

Sounds like they are severely undercharged, that is why I keep pounding the hydrometer point. It is the only way to measure SOC.

During the Bulk and and Absorb charge you should be seeing around 14.4 volts out of the controller near the end of Absorb cycle, then fold back to 13.6 in float. If I had to guess is you do not have enough panel wattage to recharge your batteries/ and/or you are over discharging your batteries daily and not able to be fully recharged the next day. Either way it means the same thing.

Well, two days ago I double-checked the charging points on the controller (which I only got a week ago - and it rained all weekend). The default points were lower than the ones recommended by Trojan.

I've got 480W (labelled) coming in. This aft, when I went home, the controller had 14.0 volts going out, which is higher than I've seen before, at more than 20 amps. At 1 p.m. the logs showed 400 watt-hours of charge put into it.

I sense a hydrometer purchase in my future. I'll keep monitoring.

[Sunking]

Forget what you see going in as it means nothing. HYDROMETER and good DVM.

[obsidience]

You said your wife was watching the Bachelor - what was she watching it on? Laptop? LCD + cable box? Are the units AC or DC?

[Vern Faulkner]

Update: Moved the modules to a much better location. On the deck, leaning against the patio, they really only got full sun for three and change hours. Even now, at 5:30 local time, there's 30-odd watts of power going on.

I obtained both a hydrometer. Well, two - but the first one didn't like slavery and made a bit for freedom, only to die a glassy-shard death. And a DVM. Will poke about later today and report back.

Oh, all (non hydrometer) signs controller, dvm, etc., suggest a full charge at this immediate moment.

I had a sneaking suspicion the inverter, even when off, is drawing current. It is - oddly enough, bouncing from 5 mA to 60mA. Not staggering amounts, but interesting, nonetheless.

[Vern Faulkner]

Couple things to report: Voltage on one battery was 6.51, voltage on the other 6.52, water levels all perfect.

I tested every cell with the hydrometer - and they're all in the range of 1255-1260 (ish) - and those specs are all within Trojan's high end charge (90-100 per cent.) (I wasn't as accurate, in retrospect, as I should have been... lesson learned there.)

However ... when I reconnect the inverter, there's a notable spark.
And if I let the inverter sit disconnecte, idle, for two minutes, then reconnected, there's anther, lesser spark.

Now, the DVM doesn't show much of a current, but I think there's a capacitor in there that's feeding some other circuitry. That, and upon thinking, I'm pretty sure the inverter's power switch looks like it can handle, maybe 15-20 amps.

So I think the culprit, here, isn't the batteries ... it's the godamn inverter, which I suspect is probably pulling something like a couple-six amps, even when "off."

PS: Charge controller's voltage reading is off by .3v - I had assumed it was right, the inverter was wrong. Turns out, the inverter was closer to accurate, and the charge controller's voltage was off: no wonder I thought the batteries weren't charging! I've tweaked the controller offset so the voltage readout is now accurate: The DVM suggestion was wise.