H38: What is this Xantrax device under my main DC panel

Jan 11, 2014
12,015
Sabre 362 113 Fair Haven, NY
The one exception is that the middle battery and first battery positive jumpers which both terminate on the Class T fuse block rather than the middle battery positive jumper going to the battery post of the first battery.
Doing this splits the battery bank into 2 different banks, which is precisely what you do not want.

The Class T fuse must be within 7 inches of the battery terminal of the last battery in the parallel bank.

Are you saying that the 18" of 2/0 AWG between the first battery post and the Class T fuse is electrically significant?
First is 11 inches longer than the ABYC standards suggest.

If all of the jumpers are 18 inches there is no problem. If some are 6, some 18, and 48 inches then there will be problems with maintaining even charges across all batteries.

As far as I can tell, the only difference between the ideal setup and what I have is that the positive jumper from the middle battery to the first battery terminates on the Class T fuse, rather than on the positive post of the first battery itself. AFAICT it all boils down to 'is 18" of 2/0 AWG significant enough to make any real-world difference'?

For a 300A load at 13.5V:
Voltage drop: 0.091
Voltage drop percentage: 0.67%
Voltage at the end: 13.409
The problem is not the absolute voltage drop, the problem the batteries will see different voltages when charging and when discharging the first battery will drain faster than the last, leaving the batteries unevenly charged.

EDIT: My guesstimate at 18" is from memory, and thinking about the ABYC standards and the ISO/RCD standards of 7", it could be less than 18" if Hunter tried to be compliant in 2007.
I'm not certain when ABYC adopted the 7 inch standard. If you continue with the factory wiring, any problems would be grandfathered. Since you are going off on your with your own design, any problems that develop are on your shoulders. A good surveyor will note the problems, which may affect any future sale and insurability. I

Your boat, your risk.
 
Jan 11, 2014
12,015
Sabre 362 113 Fair Haven, NY
Just to nitpick - a parallel bank could also have each battery feeding single positive and negative bus bars directly. As long as the cables to the busses are equal I think that’s also generally accepted.
Yes, they would be in parallel, however they would act as 3 separate banks, not one bank. Functionally, this is like putting a 1-2-Both switch on both. I like to follow the electrons as they go to work. In a parallel bank the electron has to visit all batteries in the bank, there is just one road. In the parallel arrangement you describe it is more like 2 or three electrons merging on to a highway.

This kind of parallel wiring is sometimes done with battery chargers in order to increase the charge current. Two or more charges are wired in parallel to a common post with one feed to the bank.
 
Apr 2, 2021
415
Hunter 38 On the move
The folks who are on top of their game with marine electrical systems and LFP batteries are on FB in one of Rod's groups. The folks over on CF can produce more confusing answers the we can here on SBO. :cool:
hahaha indeed but I don't fb and have no intentions of doing so

I have a pretty good handle on who to ignore on CF
 
Apr 2, 2021
415
Hunter 38 On the move
Doing this splits the battery bank into 2 different banks, which is precisely what you do not want.

The Class T fuse must be within 7 inches of the battery terminal of the last battery in the parallel bank.



First is 11 inches longer than the ABYC standards suggest.

If all of the jumpers are 18 inches there is no problem. If some are 6, some 18, and 48 inches then there will be problems with maintaining even charges across all batteries.



The problem is not the absolute voltage drop, the problem the batteries will see different voltages when charging and when discharging the first battery will drain faster than the last, leaving the batteries unevenly charged.



I'm not certain when ABYC adopted the 7 inch standard. If you continue with the factory wiring, any problems would be grandfathered. Since you are going off on your with your own design, any problems that develop are on your shoulders. A good surveyor will note the problems, which may affect any future sale and insurability. I

Your boat, your risk.

All that said, what is important is if it is a MATERIAL difference

A jumper 1mm longer than another will produce a difference. A slightly tighter crimp will produce a difference. Even a slightly different alloy used in a lug or cable will produce a difference. A cable passing through a magnetic field will produce a difference.

But, is it a material difference? That's the important thing. That's what I'm trying to get to the bottom of, and to get there, we need numbers. A quantitative measurement of the difference.

The only measurement that I can find that makes a difference is the voltage drop, and to me, that seems acceptably low.

Now, I will reevaluate whether or not I can move the middle battery positive jumper from the class T fuse to the positive terminal of the first battery, but that might not be possible.

I'm very open to anything that might indicate and quantify material differences.
 
Jan 11, 2014
12,015
Sabre 362 113 Fair Haven, NY
The answer to why it is not a good idea to split a battery bank lies in basic HS physics, which in my case took awhile to surface after a half century of neglect.

The basic issue is resistance in parallel and how current flows through a circuit, not the voltage drop. Electrons are lazy and will take the shortest path with the least resistance.

In a series circuit total resistance is easy to calculate, just add up all the resistance and you get a total. In a parallel circuit resistance is a little trickier to calculate. (BTW, it is all based on Ohms Law I=ER, see link below for the algebra.)

The formula for resistance in a parallel circuit is Rt =1/R1 + 1/R2 + 1/R3 ... where Rt is the total resistance and R1, etc is the resistance of each source, for this discussion, the internal resistance of a battery.

Let's assume the three batteries in the bank have a resistance of 100 ohms each. The total resistance for the circuit will be 1/100 + 1/100 + 1/100 = 3/100 or .03 All the current passing through the bank will be equally shared between the batteries and thus all will be equally drained or charged. So long as the batteries are of the same type and condition life will be good. However, if the batteries are not equal, then the charging will not be equal. This is why it is recommended to replace all the batteries in a bank at one time with the same type of battery. It is also why jumpers need to be of the same length as the resistance of the longer cables will increase the resistance in the circuit resulting in more current going to the battery with lower resistance, the first battery.

The wiring scheme the OP proposes is still in parallel with one important difference, functionally he will have 2 separate banks with one bank having one battery (Bank 1) and the other having two batteries (Bank 2).

Calculating the resistance in the circuit:

Rt = 1/Rb1 + 1/Rb2

Solving:

Rt = 1/100 + 1/((1/100)+(1/100))

Rt = .01 + .02

Rt = .03

Notice Bank 2 now has twice the resistance of Bank 1 which means it will accept less current than Bank 1. If the charging source is allowed to fully charge all of the batteries to 100% SOC, it is probably not a big deal. As Bank 1 approaches 100% SOC the internal resistance will increase and more of the current will then shift to Bank 2 until it is at 100% SOC. However, in practice most of us do not charge our batteries to 100% every time they are partially discharged. If charging is stopped before Bank 2 is fully charged, Bank 1 will discharge into Bank 2 until they are all equally charged. Functionally, this increases the number of charge/discharge cycles for Bank 1 relative to Bank 2.

Because Bank 1 has a lower internal resistance than Bank 2, when discharging more amperage will come from Bank 1 than Bank 2, resulting in deeper discharge levels in Bank 1 than Bank 2 which increases the cycles relative to Bank 2. More cycles and deeper discharges results in shorter battery life, especially for LA batteries.

Unequal jumper cables will add to the resistance calculations. An 18" jumper will have 3 times the resistance of a 6" jumper resulting in less current being sent to the battery at the end of the 18" cable. With all cables the same length the current distribution will be equal, with equal resistance to each battery charging and discharging will be even across the individual batteries in the bank. Which is the goal.

To summarize, restricting current flow because of resistance is the key issue when designing a parallel battery system. Proper wiring with the correct sized cable for jumpers of equal length will yield more efficient performance and increased battery life.



 

Johnb

.
Jan 22, 2008
1,440
Hunter 37-cutter Richmond CA
The formula for resistance in a parallel circuit is Rt =1/R1 + 1/R2 + 1/R3 ... where Rt is the total resistance and R1, etc is the resistance of each source, for this discussion, the internal resistance of a battery.
That formula is not correct. For parallel circuits 1/Rt=1/R1+1/R2+..........
 
Jan 11, 2014
12,015
Sabre 362 113 Fair Haven, NY
That formula is not correct. For parallel circuits 1/Rt=1/R1+1/R2+..........
Yep, an oversight, while the numbers I used ended up being incorrect the basic premise is the same. I wasn't really interested in the total resistance, because in this situation that is not so important, what is important is how the current is distributed between the paralleled circuits that doesn't change even with the error in the formula.
 

Johnb

.
Jan 22, 2008
1,440
Hunter 37-cutter Richmond CA
Yep, an oversight, while the numbers I used ended up being incorrect the basic premise is the same. I wasn't really interested in the total resistance, because in this situation that is not so important, what is important is how the current is distributed between the paralleled circuits that doesn't change even with the error in the formula.
You used incorrect numbers to prove you are right...

What is needed is to use correct numbers to show that the difference between "18" jumper that has 3 times the resistance of a 6" jumper makes a "material" difference.

12 inches of #2 wire has a resistance of 0.0002 ohms
 
Jan 11, 2014
12,015
Sabre 362 113 Fair Haven, NY
You used incorrect numbers to prove you are right...

What is needed is to use correct numbers to show that the difference between "18" jumper that has 3 times the resistance of a 6" jumper makes a "material" difference.

12 inches of #2 wire has a resistance of 0.0002 ohms
No I didn't because I wasn't trying to calculate the total resistance of the battery banks I was trying to show the relative difference between a single battery and a pair of batteries in the configuration the OP proposed. The total resistance is irrelevant, Understanding that the way in which the OP proposed wiring his batteries is not the same as a properly wired parallel battery bank.

The resistance of the jumper cables is in series with the battery, thus the jumper resistance is added to the battery resistance. And it should be doubled because both the - and + jumpers are in series with the battery.

The actual resistance of a piece of 2/0 is immaterial in this discussion. I have no idea if the internal resistance of a battery is 100 ohms or 1000 ohms and for the purpose of this discussion, it doesn't matter. The point is, once again, to show the relative resistance in the OP's proposed design and how it will affect battery life and function. Does the length of the jumper cables make a material difference? People who are much more knowledgeable than seem to think so. I'm not going to debate them.
 
Jan 7, 2011
5,089
Oday 322 East Chicago, IN
In looking at OP’s drawing, particularly the wiring to the Bow Thruster, shouldn’t there be a fuse there to protect that wiring? He has a fuse to the panel, but another unprotected branch to the bow thruster.

I don’t like the overall set up of the batteries…not really clear to me how that 3-battery set up is going to work.


Greg
 
Jan 11, 2014
12,015
Sabre 362 113 Fair Haven, NY
A quick recap. The OP wanted to install a third battery in parallel with 2 existing batteries. However, the third battery would be located some distance away from the existing batteries. Doing so would violate the standard practice of keeping all connecting cables in a parallel arrangement the same length so that charging and discharging would occur evenly and the batteries would remain in balance. The issue with violating this practice is increasing the resistance in the cabling and Ohm's Law which states current will flow through the path of least resistance.

The discussion came to an impasse when it was pointed out the very low resistance that would be added to the circuit and realistically how much effect that would have on charging. The only response I had at that time was deference to authority, every thing I had read about parallel banks said to keep the jumper cables the same length. A slightly lame response because I could not offer a good reason why.

Recently a friend asked for help with a very similar wiring problem. He wanted to add to his house bank but that would require a large separation between his existing 2 battery bank and the location of the additional batteries. I referred him to a document on the Victron website (link below) on basic wiring. In that document at the bottom of Page 17 was the answer.

The concern with uneven cable lengths is not the absolute value of the additional resistance, rather it is the additional resistance relative to the battery's internal resistance. According to Victron the typical internal resistance of a battery is between 10 and 3 milliOhms, thus adding just 1.5 milliOhms of resistance increases the total resistance by 15-50%, which is a significant amount. Twenty cm of 2ga (35 mm2) cable has 1.5 milliOhms of resistance.

One of the OP's questions was how to accurately calculate the effect of uneven cable lengths. The missing piece of the answer to that question is include the internal resistance of the battery in the equation and comparing that resistance to the resistance to the resistance in the existing banks. If the numbers are equal, there should not be much of a difference. Remember the cable and the battery resistances are in series, so they would be added.

How and where the batteries are placed in parallel is also a factor. If the existing batteries are in parallel they will function as a single large battery. Connecting that bank to the new bank at a busbar with only 1 positive cable complicates the calculation a bit. The internal resistance of the existing bank will be much lower than the new battery, leading to uneven charging and discharging. I'll leave it someone more adept at Algebra than me to set up the equation. ;)

Hope this helps clarify the earlier discussion.


 
Jan 19, 2010
12,542
Hobie 16 & Rhodes 22 Skeeter Charleston
Thanks @dlochner

That takes an obscure idea and makes it intuitive. To paraphrase...If your battery has 50% more internal resistance than another, it will accept 50% fewer electrons during charging. If we use the analogy of water in pipes (resistance) that fill two different water tanks (batteries) and the entire system is being filled by the same pump (charger) at the same pressure (voltage) then the water tank with the thinner pipes will fill more slowly and will not ever completely fill until the first tank is full. If the first tank is capable of overflowing, then the second tank never fills completly and the first tank spills (or in this case battery damage). You can repeat this thought experiment when emptying the tanks.
 
May 17, 2004
5,346
Beneteau Oceanis 37 Havre de Grace
A quick recap. The OP wanted to install a third battery in parallel with 2 existing batteries. However, the third battery would be located some distance away from the existing batteries. Doing so would violate the standard practice of keeping all connecting cables in a parallel arrangement the same length so that charging and discharging would occur evenly and the batteries would remain in balance. The issue with violating this practice is increasing the resistance in the cabling and Ohm's Law which states current will flow through the path of least resistance.

The discussion came to an impasse when it was pointed out the very low resistance that would be added to the circuit and realistically how much effect that would have on charging. The only response I had at that time was deference to authority, every thing I had read about parallel banks said to keep the jumper cables the same length. A slightly lame response because I could not offer a good reason why.

Recently a friend asked for help with a very similar wiring problem. He wanted to add to his house bank but that would require a large separation between his existing 2 battery bank and the location of the additional batteries. I referred him to a document on the Victron website (link below) on basic wiring. In that document at the bottom of Page 17 was the answer.

The concern with uneven cable lengths is not the absolute value of the additional resistance, rather it is the additional resistance relative to the battery's internal resistance. According to Victron the typical internal resistance of a battery is between 10 and 3 milliOhms, thus adding just 1.5 milliOhms of resistance increases the total resistance by 15-50%, which is a significant amount. Twenty cm of 2ga (35 mm2) cable has 1.5 milliOhms of resistance.

One of the OP's questions was how to accurately calculate the effect of uneven cable lengths. The missing piece of the answer to that question is include the internal resistance of the battery in the equation and comparing that resistance to the resistance to the resistance in the existing banks. If the numbers are equal, there should not be much of a difference. Remember the cable and the battery resistances are in series, so they would be added.

How and where the batteries are placed in parallel is also a factor. If the existing batteries are in parallel they will function as a single large battery. Connecting that bank to the new bank at a busbar with only 1 positive cable complicates the calculation a bit. The internal resistance of the existing bank will be much lower than the new battery, leading to uneven charging and discharging. I'll leave it someone more adept at Algebra than me to set up the equation. ;)

Hope this helps clarify the earlier discussion.


That is a great explanation! The picture of the parallel resistances in the Victron paper really made it clear to me in a way I haven't visualized it before.
I'll leave it someone more adept at Algebra than me to set up the equation. ;)
I started doing the math, but decided there must be a better way :). So I went to Circuit Lab to build and simulate a circuit. That was a great illustration of how the power comes unevenly from batteries in that configuration. For a 12 amp load the closest battery was contributing over 4 amps, and the furthest less than 1 amp. Certainly that's a significant difference.

Having said that, I'm still not convinced about the real-life effects of the bad configuration. Although more electrons initially get pulled from the "closer" battery, that battery will quickly fall to a slightly lower SOC and then contribute less. Wanting to see the effects more I played around with the voltages of the individual batteries to try to balance out the current from each, and here's what I got:
1693935874539.png


So at a ~12 amp draw the lowest battery will go to just 0.055 V less than the highest before they balance again. Certainly higher draws (or charge currents) will increase the deltas. Maybe later I'll play around with it more to see the effect of, say, an 80 Amp charger. (Apparently a free membership is limited to like an hour per day...)
 
  • Helpful
Likes: rgranger
Jan 11, 2014
12,015
Sabre 362 113 Fair Haven, NY
That is a great explanation! The picture of the parallel resistances in the Victron paper really made it clear to me in a way I haven't visualized it before.

I started doing the math, but decided there must be a better way :). So I went to Circuit Lab to build and simulate a circuit. That was a great illustration of how the power comes unevenly from batteries in that configuration. For a 12 amp load the closest battery was contributing over 4 amps, and the furthest less than 1 amp. Certainly that's a significant difference.

Having said that, I'm still not convinced about the real-life effects of the bad configuration. Although more electrons initially get pulled from the "closer" battery, that battery will quickly fall to a slightly lower SOC and then contribute less. Wanting to see the effects more I played around with the voltages of the individual batteries to try to balance out the current from each, and here's what I got:
View attachment 219563

So at a ~12 amp draw the lowest battery will go to just 0.055 V less than the highest before they balance again. Certainly higher draws (or charge currents) will increase the deltas. Maybe later I'll play around with it more to see the effect of, say, an 80 Amp charger. (Apparently a free membership is limited to like an hour per day...)
Good to see that Circuit Lab doesn't want to distract you from productive work. ;)

The losses in the diagram from Victron are pretty small, then again 20 cm is about the length of a jumper on the typical parallel setup. It would be interesting to see what would happen if one jumper was a meter and the others 20 cm. That would be a closer replication to the OPs initial question.
 
May 17, 2004
5,346
Beneteau Oceanis 37 Havre de Grace
Undeterred by Circuit Lab's limitations I found an open source circuit simulator called Qucs and modeled more circuits. I can't say I fully understand @flightlead404's proposed configuration, so I built a circuit to match my own (sub-optimal) battery configuration. From the dealer our boat came with one house battery in the aft cabin and one in the engine compartment. All loads and grounds went straight to the aft cabin battery, with jumpers (maybe 10' of 2/0) to the one in the engine area. Knowing this was not a great situation every couple years I did a 20 hour load test on each battery and rotated them to try to keep them somewhat balanced. They always tested within 5% of each other and of their labeled capacity. Last year, after 7 seasons our house batteries abruptly went from performing quite well to totally dead (one with a bad cell, the other with 70% state of health). We took that opportunity to add a third battery in a nearby lazarette, about 12 feet of 2/0 away, just jumpered off the aft cabin battery.

At a typical high end sustained load for our usage (12 Amps), this is what that circuit looks like:
1693957998688.png


So, at 12 amps our aft cabin battery is contributing 4.89 Amps, while the others contribute 3.54 A. Not great if it stays like that for an extended period.

But then if the "closest" battery drops SOC just a little, things come into balance again:
1693957958525.png


That's just 0.0076 Volts of difference between the batteries to keep the output balanced. In the real world I'm not worried about that amount of imbalance.

Looking more at a worst case, let's say my alternator is putting out 40 Amps (the most I've seen sustained from my 80 A dumb Hitachi alternator). The amount of imbalance in that case goes up to 0.025 V. That's still such a small margin I'm not worried about it for my use case.

Extending this further, I removed the extra resistances, and played around with the battery internal resistances, considering the common guidance that parallel batteries of different age is bad because of dissimilar internal resistances. What I found is this:
1693958621458.png


At 40 amps a battery with very high internal resistance (the top of Victron's 3 to 10 milliOhm estimate) will be less than 0.1 V away from two other batteries with very low internal resistance. Also not earth shattering IMHO.
 
Apr 2, 2021
415
Hunter 38 On the move
In looking at OP’s drawing, particularly the wiring to the Bow Thruster, shouldn’t there be a fuse there to protect that wiring? He has a fuse to the panel, but another unprotected branch to the bow thruster.

I don’t like the overall set up of the batteries…not really clear to me how that 3-battery set up is going to work.


Greg
I'm adding 300A MRBF to each positive post including the starter battery. However, after reviewing my specs for the BMS on each house battery I believe this will act as a fuse by shutting off in a short situation. Nevertheless I'm adding them.

Not shown is an ANL for the bow thruster already in place.
 
Apr 2, 2021
415
Hunter 38 On the move
This is great stuff, man we're a bunch of nerds lol I just wrote 3,000 lines of code to automate a home mushroom growing environment and control it from my phone lol

To clarify a few points, as is usual some confusion has crept in above

1) From the factory the boat came with two batteries in the house bank that are approx 6' of jumper cable apart from each other, and included that inaccessible class T fuse
2) I intend to add a 3rd battery IN THE MIDDLE (not at the end) but with the same lengths of jumpers by pulling the factory jumpers going to the far end battery back a bit and running them to the middle battery, then adding two new jumpers for the far battery
3) In the original factory config the positive jumper from Bat 2 to Bat 1 terminated on the Class T fuse, not the Bat 1 positive post
4) I intend to leave that positive jumper on the Class T fuse because I can't access it.

The only issue I can see is LEAVING the positive jumper between Bat 1 and Bat 2 both going to the Class T fuse, rather than to the positive post of Bat 1 and thence to the Class T fuse (approx 12-18" of 2/0 AWG)

Based on the above, the difference the resistance of the Bat 1 + jumper is different to the resistance of (bat 2+bat 3)+jumper by the difference in resistance of the different lengths of jumper.

Now, we all know theoretically this makes a difference.

The other thing to throw into the mix on whether this makes a material difference is that I'm using LiFePO4 with internal BMS, whereas all the papers etc I could find were for some form of LA battery. Given that my batteries are effectively "smart" devices and that LiFePO4 itself doesn't suffer from the same issues as LA around less than 100% SOC, in my mind this further reduces the material impact of differences.

In the end though, all I can do is all I can do. If I can access that Class T fuse and move the jumper to the Bat 1 battery post I will. If I can't, I will just have to live with it. If that means battery life goes from 20+ years to 10 years oh well.