what might cause a rapid rise to Absorb V, with little BULK charging?

westbranch · October 2015

background: There is a constant overnight load on my 900Ah AGM battery of ~3 Amps from the inverter for ,Sat Internet connection + WiFi connection and ,at ~ 2 hr intervals , a Fridge...last night;s readings ranged from 24.0 to 24.1V.... yes a sick battery. I loose a volt or 2 every day due to each 'no charge day'.. Been trying to remedy it for 3 weeks.

My problem is that with a low overnight voltage of ~24.1V at 7 AM, the only BULK charging I am getting is from sunup till about noon, putting in almost no Amps most <1A, max 3.5A.

Then when the sun comes out from behind the shade trees and Voltage rises from ~ 24.5V to Absorb level of 29.3V (temp adjusted from 28.8V). At the same time the amps rise to ~ 5A, and then drop to <1 for the entire Absorb time. I have set End AMPS to 0.0 to ensure I get a full 4 hrs Absorb, and since there is no BULK I even get to float, even less input...

Here is a screen capture of today...

Mountain Don · October 2015

What happens if you run a charge off a generator? The rapid rise to absorb sounds like what I've been told will happen when a battery, or two or more, is dying.

(and the image does not show for me)

westbranch · October 2015

Don,the first graph is in the Sep30.doc shown at the bottom of the page , for some reason below my sig. lines. I don't have a tool to make a JPEG or other like format on this machine...so it is in a Word doc...

the combined charge is shown in days 1, 2 & 3 of the Sep 25 to 29 .doc graph...Day 1 & 2 the charger was in play till after noon ~ 5 hrs each day
when on the CC and charger it poked in 30A from the charger and whatever the CC put out, and chugged along for hrs, with the Voltage very slowly rising, and indicated, on day 3, I was ~ 50%+ charged (25/50/75/100 scale) when we shut it down, and left...Day 1 maxed at ~ 50A, day 2 at 45A, day 3 speaks for it self, a 9.6A max...
see Sep 26,27.28 doc

inetdog · October 2015

I concur that this is a sign of either badly reduced battery capacity or high series resistance. Either one could result from sulfation or other damage to one battery or even one cell. (If one cell gets driven past 100% SOC it will go to a high terminal voltage which will inhibit charging of the other cells in the string.)

I would check battery voltage and cell by cell SG first. Then I would carefully raise the Bulk voltage to see what effect it has.
If one cell is bad it should be gassing noticeably at that point. If it is just a matter of uneven SOC, it will server as a mild equalizing charge.
You want to get the rest of the cells up to full SOC if possible, but if one cell of a battery is bad the whole battery will have to be replaced.

Dave Angelini · October 2015

Really hard to check SG on an AGM is it not? Turn out the lights time....

westbranch · October 2015

inetdog said:

I concur that this is a sign of either badly reduced battery capacity or high series resistance. Either one could result from sulfation or other damage to one battery or even one cell. (If one cell gets driven past 100% SOC it will go to a high terminal voltage which will inhibit charging of the other cells in the string.)
.......
..... If it is just a matter of uneven SOC, it will server as a mild equalizing charge.
........, but if one cell of a battery is bad the whole battery will have to be replaced.

Thanks Inetdog, that first part was what I wanted to confirm, high series resistance...

There is a number for a mild EQ for these AGMs and it is a 16 - 24 hr process... I will up the time and V fora CC delivered Absorb for a few days
I just got a new VMeter, that reads to 0.XXX, the previous one lost the 0.XX and was only 0.X so not too accurate or reliable. Now that the bear has finished ravaging my bee hives here at home I can get back out west..
Dave,I believe you are right in the long run, just want to go through the process and see what I can learn from it.

Dave Angelini · October 2015

I learned once that someone would have to give me the battery free before I would accept a battery (for offgrid life) that did not allow
use of a hydrometer. In most all of the battery failures my 80+ clients have had, the SG is giving a heads up long before a problem like yours surfaces. I do wish you good luck as this is never pleasant.

zoneblue · October 2015

I cant remember the age of that bank, or if they are two volt cells. If so, and its only one bad cell, you might try taking it out. However, bets are they are many/all of them on their way out now.

However, the bank does seem to have accepted some charge on the 26th,27th.

westbranch · October 2015

Right. 2v cells ~7 yrs old.... the intermittent charge is a real puzzle... but 2 different cc's and the same pattern...

westbranch · October 2015

well....I think I have zeroed in on the source of the puzzle....high resistance...

Right after posting the above I went down to do some comparative V readings at the battery and inside the E panel. Was going to do from the Classic but will do that tomorrow....

Whoa! Why are those RED wires BLACK???

The 2 Pos wires that attach to the small POS bussbar, one from the battery post of the main CB , the other from the PV CB, are both charred ~2 inches back from the bussbar.... That should create some resistance to current flow, eh ??

Checked the screw tightness and both hard tight, loosened them and re-tightened and found that the one for the PV now read ~2.0 V low. Loosened and tightened and the voltage jumped around ... stopped when it was within .2V of my DVMM of battery V.

Question now is WHY? when they were both tight?

inetdog · October 2015

A voltage drop of .2V is still higher than it should be for just the terminal connection. (What was the current at the time???)
If you use the wrong size terminal or fail to tighten the screw or bolt initially it will generate heat which will expand the copper. It will flow (slowly) under that pressure.
When you let it cool the pressure on the screw will be lower and the resistance higher. So greater heating the next time you power up the inverter or CC.
Lather, rinse, repeat.

Charred wire insulation will not create a high resistance. It is a symptom of high resistance somewhere close. If the wires are the right size, it will be the fault of a bad connection.

Heavily oxidized copper can show in increased resistance to current flow as a diode. They still make rectifiers for some applications from copper oxide layers in a stack of disks. It takes a lot of layers to withstand a high voltage, but the forward voltage drop can still cause you problems in a low voltage circuit.
Aluminum is much worse since aluminum oxide is a very good and durable insulator.

PNjunction · October 2015

I'm glad you found it!

If i may offer some back-seat driving ... run a temp check with a handheld IR thermometer over your infrastructure from time to time - doesn't have to totally accurate - just enough to be able to find hot-spot trends - like the one hidden by seemingly tight terminals. They don't have to be Fluke quality either just to spot a trend.

westbranch · October 2015

PM, I did have that thought last night when I saw the charring...hind sight is 20/20 I wanted to do that last night but IR meter is 180 miles away... have to get one for here.
This morning the V readings are 2.3V apart. I have to shut it down and steal some wiring from my other Epanel, waiting to be installed and make the other one...

westbranch · October 2015

took some pics. I Used Image Resizer Powertoy Clone For Windows 7, slick, to shrink them... 1 2 are of the bussbar 3 is stain below PV CB

Image: https://us.v-cdn.net/6024911/uploads/editor/uy/bk83uzunt1ke.jpg

waynefromnscanada · October 2015

westbranch took some pics.

WOW! Nasty! I had some automotive fuse holders do this, both 30 amp and 80 amp. The legs on the push in fuses welded themselves into what remained of the melted holders. Happened after a year or two of operating just fine.
Very nearly caused a fire in my case.

inetdog · October 2015

I see that two of the holes with charred wires connected have larger screws than the others. Have you confirmed that those holes are rated for the size wire you are using?
Wire too small for the terminal may not be properly compressed under the screw(s) and may slip to the side giving poor contact.

westbranch · October 2015

Inetdog, the bottom one is the MN factory wire and the same hole size on top was the only other one that would accept the 4 ga. wire... After removing and dismantling the factory original in my 'spare Epanel, I believe that the wire was not sufficiently compressed... had a hell of a time getting the factory installed screw undone , felt like 50 ft pounds!

CLUE: When I dismantled the wiring from the buss bar the upper wire just slipped out, yet last night It felt like it was in as tight as possible... the bottom one was still tight...

more to come

PNjunction · October 2015

Oh man - dissimilar metal right off the bat, copper oxidizing and so forth. Talk about a smoking gun. That's actually a good thing really - since the fix is to run with copper busbars at least.

Dave Angelini · October 2015

PNjunction said:

Oh man - dissimilar metal right off the bat, copper oxidizing and so forth. Talk about a smoking gun. That's actually a good thing really - since the fix is to run with copper busbars at least.

I disagree, the fix is to torque the connections and after a year check the torque. They have been building this to code and the missing ingredient is loose wiring. A specific gravity measurement on a cell or two would have showed this months ago.

westbranch · October 2015

I have been chasing the symptoms for ~ 2 months now.

The first thing I noticed was poor charge state after days of full sun. First I noticed that I was in FLOAT by 13:00 which led me to a short Absorb, and then to maybe 5 minutes of ABSORB ( a result of very short BULK stage) . I was focused on the CC settings at that point.

The burnt wires are a recent result that coincided with new FW and adding another 4S array. The wife did mention a dusty hot wire smell a couple of times , but no visible evidence at the time... the Classic was putting out ~ 40 Amps at the time and the cooling fans were on too. We never had had that much amperage before..

After the a FW update to the 2056 version I saw a big jump in production over all at first and then a degrading of good charging.... back to the Voltage jump and amperage drop...

Lesson learned: CHECK and Tighten all connections Yearly...

mike95490 · October 2015

I had some toasty wires in my epanel last year. Same thing, coarse strands into the buss bar, and it loosened up over time, Mine got a fair amount hotter than yours, even the buss bar supports were browned, I'd been looking for a loose wire for a week, was getting over volt faults, and had checked the batteries several times with the IR thermo gun, but didn't get into the epanel yet.

Aguarancher · October 2015

I have always preferred hex head terminal screws over the slotted head as I feel I can get a better torque on the connection. Another thing I have always done when space allows it, is to wrap the ends of the cables with .003 copper or brass shim stock first as it helps keep the strands confined and makes for a better set on the connection.

Dave Angelini · October 2015

I think you are missing the big picture. The connection is torqued to Midnites spec. The second check is just to verify the wires compressed properly and held at that torque. As bad as loose wiring is, not far down the list would be tightening wiring yearly. It does not need to be done and and can also cause problems reducing the diameter of the copper.
We torque the DC wiring because tightening it is not accurate unless you really have the feel for it. Takes the guesswork out and this common problem is avoided. It is not a bad investment to buy a good wrench!

Aguarancher · October 2015

I always torque screws to oem specs when possible. Never felt I could get an accurate torque with slotted head screws. Re-torque after thermal cycling is standard procedure.

Mountain Don · October 2015

Torque drivers for firearms are often a better price than the electrical specialty ones. I use this one

It's not meant for daily use, but my 20 yr old model still finctions for all I do. You are supposed to back off the spring when placing in storage. The last time I checked the torque against a premium line torque clicker it was close enough for gov'mt work

zoneblue · October 2015

Ok interesting. Weve seen pics like this before, more than once, actually, on those exact same midnite screw terminal bus bars. Me, im no fan of them, and suggest that they should not be used for anything carrying more than trivial current. They are i beleive rated for 80A, but if you look at the way midnite epanels are prewired, for isntance, the charge controller doestn route though those screw bus bars. For, a, reason.

I just dont like them. For a start they are made of alumnium, and plated with tin. The former means you cant really tighten the screws down real well without fear of stripping the threads. That is particuarly so in the case of the larger diameter holes intended for up to 2awg, because it means that there are so few threads left for the screws. For the latter any wear on the terminal will puncture the tin, and expose the alumumium, with all its associated oxidation issues.

Here in NZ brass is used for all bus bars. I have never seen aluminum used for that here ever. Yes i know brass has a higher resistance, actually quite a bit higher, but it is way stronger. I imagine the brass usage here is a product of the 240V AC we tend to use, you can afford to lose a volt here or there. Maybe. Or maybe its a lower resistance grade of brass. Dunno.

Crimp and lug for all PV and Controller cabling is one quite good solution. Maybe thats paranoid, but, it works.

However i think the pretty strong screw terminals on the CBI breakers are OK, if retorqued as specified.

Dave Angelini · October 2015

I could not find the midnite rating easily but I use a similar system from Outback and Schnieder. They have a 1/0 slot and so it can carry 170 ADC. Below is the spec from the store here.

Available in Black, Red, Blue, White, Brown, and non-insulated (for grounding) (specify when ordering). Use for adding more wire terminations or for isolating or grounding multiple positive / negative circuits. Usable in any system, on or off grid - not just Outback Power systems. Also good for RV and marine. All TBB models have three #1/0 to 14 AWG and eight #6 to 14 AWG screw type compression terminals, which means no ring lugs required. Available with black, white and red insulators. The Battery Bus is .25" (7 mm) thick tin plated copper and can be mounted on the back plate of any enclosure. Bus bar with black insulators - use as L1 hot or DC negative TBB-Ground non-insulated - use as a general purpose ground bus bar for systems. TBB-BLUE Bus bar with blue insulators - use as Phase C on three phase systems. TBB-RED Bus bar with red insulators - use as L2 hot or DC positive. TBB-WHITE Bus bar with white insulators - use as AC neutral or DC negative

what might cause a rapid rise to Absorb V, with little BULK charging?

Comments

Categories