Being the holidays, I don't have much time to get into this. Without a doubt, a balanced connection is best. You ran a simulation of discharging a battery to 50% in 2 hours. That is not a typical scenario. What about using the microwave for 5 min?
I did a similar battery sharing calculation several years ago. No spice program. I did the calculation the old fashion way. 4 batteries, each with 10 mohm internal resistance and 1 mohm connecting cables. Load current is 105A. Each battery supplies:
B1 = 23A
B2 = 25A
B3 = 27A
B4 = 30A
Yes, we can do better by placing load cables at opposite ends. Does it matter that much that B1 supplies 7A less than B4? After the load is turned off, the charge will equalize in each battery. No harm, no foul.
Internal resistance varies greatly with the state of charge. As B4 gives up more of it's charge more quickly than the others, its internal resistance will go up in relation to them and its voltage will roll off. The four batteries will tend to equalize with each other to a large extent even as they are discharging. The degree to which this happens is unknown. That phenomenon is one reason why you need an accurate non-linear model to do the battery simulations.
Salvo wrote: Being the holidays, I don't have much time to get into this. Without a doubt, a balanced connection is best. You ran a simulation of discharging a battery to 50% in 2 hours. That is not a typical scenario. What about using the microwave for 5 min?
I did a similar battery sharing calculation several years ago. No spice program. I did the calculation the old fashion way. 4 batteries, each with 10 mohm internal resistance and 1 mohm connecting cables. Load current is 105A. Each battery supplies:
B1 = 23A
B2 = 25A
B3 = 27A
B4 = 30A
Yes, we can do better by placing load cables at opposite ends. Does it matter that much that B1 supplies 7A less than B4? After the load is turned off, the charge will equalize in each battery. No harm, no foul.
LScamper wrote:
So it seems that a balanced connection is best.
IIRC, your results appear better than smatgauge's. Atleast you show a figure for interconnect resistance. Smartgauge never bothers to mention that. I have never been able to fathom why this neglect.
Now what would happen if you figured 12" lengths of 2/0 or better?
Cheers,
Kendall
1986 Winnebago Chieftain 22RC
Our Camper (with no payments)
Cedarhill wrote: Internal resistance varies greatly with the state of charge. As B4 gives up more of it's charge more quickly than the others, its internal resistance will go up in relation to them and its voltage will roll off. The four batteries will tend to equalize with each other to a large extent even as they are discharging. The degree to which this happens is unknown. That phenomenon is one reason why you need an accurate non-linear model to do the battery simulations.
A few of us agree with your assessment, though I. Employ method #2 with 2/0 interconnects on my bank of 3 because it's no more difficult or costly than method 1. I doubt method 3 would make much of a difference... especially for a guy with no larger draw than a 300w inverter.
“So my plan to use a fifth battery as an AH pail to put AHs into the bank of four isn't going to be as fast a process as I had hoped.”
You already know the answer to that. Think of the charged battery as a constant voltage battery charger with a voltage set point of 12.6V. If you try to charge a low battery with a 12.6V charger you won’t get very far! It is even worse than that. As the low battery charges up the charged battery voltage goes down.
I didn't want that answer, so I ran a little test to prove LScamper is wrong! He just has to be, or my AH pail idea won't work.
So I ran a pair of 6s down to 50% approx and connected up the full 27DC using my 25 ft 1 ga cu clad-al jumper cables that are supposed to be like 3 ga copper, and watched the Trimetric for amps.
Time, 27's volts, 6's volts, amps
Before start, 12.60, 11.95, n/a
1 min- 12.13, 12.10 , 10amps (tapering down fast)
after 3 minutes, amps settled at 1.8
5 min- 12.16, 12.12, 1.8a
10 min- 12.17, 12.13, 1.8a
15 min- 12.19, 12.15, 1.7a
No rise in AH measured yet in the 6s, and the SG in the 27s not dropping either.
Well phooey some more! LScamper was right! (except about the battery voltages--both go up after the initial hit) Drat.
For fun, now detached the 27DC and used the other pair of 6s (still full) with the jumper cables.
Amps 14.5ish and tapering fast, dropped to 4.5 amps after a few minutes. Better, as to be expected, but no cigar.
So the AH pail idea is a total failure as I had planned it.
2003 Chev 2500HD Gas, 2003 Komfort 26FS 5er
See Profile for Equipment List
"Yes, we can do better by placing load cables at opposite ends. Does it matter that much that B1 supplies 7A less than B4? After the load is turned off, the charge will equalize in each battery. No harm, no foul."
You beat me to the punch line. No need to finish this thread! Also was going to show that even in a balanced system the internal battery resistance will do more to unbalance than the wiring. You will never get two batteries that are exactly matched. We agree I think but this could be fun anyway.
"Being the holidays, I don't have much time to get into this. Without a doubt, a balanced connection is best. You ran a simulation of discharging a battery to 50% in 2 hours. That is not a typical scenario. What about using the microwave for 5 min?"
Agree this is not a typical scenario. At 5 min there will be very little difference in SOC. Not a very dramatic show! That is the only reason for 2 hours.
I would like to see what happens over a 6 or 8 hour discharge cycle, which is typical of my battery usage. My guess is the undesirable effects of differences in resistance paths of a multi-battery system would be less.
I have been away. Not yet ready to give up on this.
Thanks for the constructive comments, they are appreciated.
Cedarhill wrote:
"download and plug in one of the many lead-acid battery models that are available on line"
I found much discussion on lead acid battery modeling but not much in the way of a useable model. I have added Peukert to the simulation with a Peukert number of 1.3. This still does not model charging. I know there are many more variables but hope this will make a much more real simulation. If you know of any specific models I would be interested.
"battery voltages--both go up after the initial hit"
Did you let the 6s recover after running them down to 50% before connecting them. Both battery voltages going up makes me think that the 6s were just recovering after discharge. The important thing is the difference voltage between the two batteries, that is what drives the current.
Offline
