AGM wire sizes more critical?

[booster]

I have our new batteries, and inverter/charger setup on the bench so I could get the inverter/charger programmed and the batteries run a bit to make sure all is well. They won't be installed for a while, so it seemed a good idea to test early. We are talking about four Lifeline GPL4CT six volt batteries wired series then parallel with a Magnum 2000 watt PSW inverter with 100 amp charger.

The charger is set up to exit absorption based on return amps measured by a shunt, so the charging control is very accurate. Lifeline says to use .5% for the return amps so on our 440ah that would be only 2.2 amps. I set it up for that, but it never got there, stopping at 3.2 amps, but what was more concerning was that the two 12v series blocks were charging at significantly different amps and showing different voltages between the batteries.

I called Lifeline, and they said not to worry about inbalances until the bank had been cycled a few times. I ran a couple of 20% discharge cycles with little change.

At this point, I setup a larger load on the batteries by using a hot plate off the inverter so I was getting a 68 amp DC pull from the batteries. I checked the amps on each battery and got 31 and 37 amps. 10% of total load variance-not good. Voltages also varied a lot between them.

I had it wired with separate negative cables of similar length (within 6") to the shunt, but with a single positive to one of the pairs and a clamp on jumper to the second pair. I wanted to be able to easily remove the jumper when the batteries were resting so I could get an accurate state of charge voltage. The wiring was all 2/0 except the jumper which was 4ga on huge battery clamps. I have used a similar setup on our wet cells without seeing anything weird, and the jumper would carrying less than 35 amps so it was way within ratings for the 4 ga.

To test, I moved which battery pair the main positive cable went to, and still used the same jumper setup to the the other pair. The current difference of 6 amps was still there, but it switched sides, so the jumper was causing it. I made a bolt in jumper out of the 2/0 and put it on and the difference went down to 1 amp.

The 4ga has about 4X the resistance of the 2/0, IIRC, so you would get the same issue if you had 2' of 2/0 instead of the 6" of 4ga. Lots of setups I have seen have that much mismatch of cable length, as that is what this is. I am very surprised at this to say the least, as with wet cells I saw almost no mismatches with reasonable cable length differences.

We have all seen the daisy chained batteries with jumpers between the positives and the negatives, sometimes 4-6 in a row. That would mean the 6th battery would have 5 jumpers worth of extra resistance and surely charge and discharge differently than the first battery.

It must be that the internal resistance of the AGM batteries is so low that the tiny variations in the wiring resistance is a big enough % of the total to make a significant difference.

I had been going back and forth on whether or not to put in the extra cabling needed to very closely match the cable lengths for the two pairs, and I think this answered the question very plainly. I will also be checking all the connections and crimps with the millivolt meter, as it seems to easily show if a connection is higher resistance, when the 65 amp load is on.

It probably also helps explain some to the failures folks have, often with only part of the battery bank.

[markopolo]

Interesting discovery. Any chance that it was corrosion on the 4ga jumper or something like that?

When I took apart a second 12 battery pack I bought a couple of the batteries where down to 6 volts and others where 12v and others were at 11 point something. You'd think all would have balanced out as they had been unused for a long time....

When I put the first 12 battery pack in my van I paralleled the entire set two ways. The four corner batteries each have direct connections to two other batteries. The eight middle batteries each have direct connections to three other batteries. All twelve in one long string would have worked but it seemed to me that a "grid of paralleling" might be beneficial. I hope the extra wiring pays off but I guess I shouldn't be surprised if an imbalance occurs over time.

An annual battery rotation might be a good idea.

[booster]

I thought that maybe the jumper was bad initially, even though it is only a few months old, so I checked for voltage drop all along its length. I was worried that either the teeth into the battery cable connection or the cable to clamp connection might be bad. I could get to all the points and even just catch the cable itself just at the end of the lug in the clamp. I couldn't find anything that looked out of place with it and the overall voltage drop was about 4X what the 2/0 cable checked at, and that corresponds with resistance difference, and the current difference. We are talking very low millivolt readings, but they were very consistent. I am in the process of taking the whole bank down to 50% to see how the charger does on full automatic and return amp transition. The discharging amps between the two pairs have been within one amp (out of 30 on each) of each other, but not always the same direction, so that is pretty good. I hope it does as well on recharge.

It sounds like you did the right thing on your setup. You should be able to trace the total wire length for each 12v circuit you have (while they are individual, not including after combined) to see how they match. You have to count positive and negative cables. You might even have room for the clamp on meter to check amperage on each leg under load or charge. I think a long strip would have been a problem.

I did just look at the tech manual on the Lifeline site, and they do take into account total cable length in their diagrams. A series/parallel like I am doing is shown with the positive coming in one end of the bank, and the negative coming out the other end, rather than off the same battery. That evens out the total wire length in the area between where they go to a single line. I had looked at those diagrams many times before, and never noticed they did that. Most banks I have seen in RVs have both coming out of the same end.

As a side spot of information. I found out when I was testing the batteries as individual pairs that two 220ah 6 volts giving 220ah of 12v capacity will max out a 100 amp charger when they are only down 10% (or 22ah). It had made up most of the 22ah in about 10 minutes.

[Boxster1971]

booster - thanks for pointing me to your threads on Lifeline batteries and Magnum inverter/ chargers. I scanned these threads while traveling earlier this summer before I committed to my battery and inverter upgrade. Now this thread provides confirmation that I have to keep the the cable lengths of my series/parallel pairs equal.

I plan to mount the series pairs under my Sprinter on each side in rear where there is space. Then bring the parallel wires up each side through the floor to meet in the center under the lounge/bed so the wires are all of equal length to buss bars that will then distribute 12v loads and charging as needed.

[markopolo]

Sort of related to this topic .... I think.

I have one "house" battery bank at just under 400ah. 180ah is located near the inverter. 216ah is 20' of 4 gauge away

I was running the 5200BTU A/C a month or so ago and checked to see if the banks were sharing the load equally. Something like 2/3rds of the power was coming from the closer 180ah group and 1/3rd was coming from the farther 216ah group. The load was something like 50amps. I made notes at the time but can't find them now.

As the battery voltage decreased the difference narrowed - it might have been 60% / 40% the last time I checked.

This is most probably related to what booster noted above but some of the imbalance in my rig could be the attributed to the different types of batteries in my van.

[booster]

Originally Posted by markopolo

Sort of related to this topic .... I think.

I have one "house" battery bank at just under 400ah. 180ah is located near the inverter. 216ah is 20' of 4 gauge away

I was running the 5200BTU A/C a month or so ago and checked to see if the banks were sharing the load equally. Something like 2/3rds of the power was coming from the closer 180ah group and 1/3rd was coming from the farther 216ah group. The load was something like 50amps. I made notes at the time but can't find them now.

As the battery voltage decreased the difference narrowed - it might have been 60% / 40% the last time I checked.

This is most probably related to what booster noted above but some of the imbalance in my rig could be the attributed to the different types of batteries in my van.

It might be interesting to do that experiment again, and when you are done and at the closer sharing, stop and disconnect the batteries from each other, let them sit and see what the states of charge are. My guess is that the close battery will be significantly lower SOC, and thus lower voltage, so the current starts to drop.

[markopolo]

No doubt you're right. I have a 500 watt electric heater I could use this time of year that would mirror the previous load.

I wonder about the recharge also. I should take current measurements during the charging cycle also. I think the converter/charger in my van maxes out at 14.4v but it would be interesting to see which group gets to 14.4 first and also if both groups actually get there before the converter/charger ends the bulk charge cycle.

[booster]

Originally Posted by markopolo

No doubt you're right. I have a 500 watt electric heater I could use this time of year that would mirror the previous load.

I wonder about the recharge also. I should take current measurements during the charging cycle also. I think the converter/charger in my van maxes out at 14.4v but it would be interesting to see which group gets to 14.4 first and also if both groups actually get there before the converter/charger ends the bulk charge cycle.

As our wet cells got older, the 12v single battery started to accept charge (or was wasting the amps) faster than the two six volt bank, so it would get full first. When they were newer, the difference was much less. In our case, the two bank charger helped as it monitored the current to each, but when one got full first it only reduced the voltage .5 volts so with a 14.7v absorption, the first on done saw 14.2 for quite a while sometimes. Both would eventually get full as we were ending on amps, but it certainly wasn't ideal. If the charger had taken them to float instead of the 14.2, it would have been much better. In hindsight I think I should have set the absorption lower to minimize the gassing.

You will probably see the same thing with yours, depending on the charging rate. Your voltage drop will be less with reducing current, so if the rates are low, both banks will see closer voltages, and hopefully, amps. Are the banks themselves similar batteries, or could that also be contributing?

[markopolo]

Absolute mix of batteries, mixed capacity, age, specs & brands but all AGM on the house side of things - I think of it as my battery experiment.

My thinking is that the only way to keep them all reasonably content & happy is to keep them all charged.

We did a couple of nights parking lot camping in September and the voltage readings were excellent. They were still at 12.7v (no load) when turning in for the night and already showing 12.9v from solar input in the early morning. I realize that is because we really don't use much power in the RV. LED lights, LED TV & propane fridge etc.

I can get full and sustained charging voltage (14.2+) out of the alternator, solar, or converter/charger in this van and that's a key ingredient.

I'm thinking more about it now with a view toward what to expect if my power usage increases because of a compressor fridge or using the A/C regularly for example. I suspect it will be harder to keep the mixed bag of batteries happy.

[booster]

I think you have the right idea of how it is working. The very low usage, and then immediate recharge off solar makes life a lot easier as all the batteries, regardless of mismatch, are very near their full charge and pulling very low amps. That makes the resistance in the wiring essentially go away as a factor, and makes the mismatch much less of an issue.

Bigger loads and deeper discharges will probably make things worse, and the bigger and deeper you go it will be more unbalanced.

Going all the way back to when we had the mismatched batteries and the Tripplite charger, I found I was able to tell fairly well when the first to fill battery was fully charged (on the rare occasions the Tripplite would get there), by looking at the total amps both banks were taking. It was relatively consistent. At that point I opened the disconnect that I had on the battery that was full to prevent more charge on it.