Interesting... I had looked at that earlier so I went back to look again. Actually the dimensions I see for Victron are 9.4H x 12.7L x 6W. That 12.6" is too long to fit in the undermount tray that Hein sells, for use in front of the wheel wells. The Battle Born GC2 batteries are 11 x 10.3 x 6.9 if you use them with the terminals on the side, and it is the 10.3" length that fits in the tray.
Airtime's electrical schematic for review please
- Thread starter Airtime
- Start date
I've done some analysis of a "Method 3" approach to cabling up my Battle Born batteries in undermount trays on left and right. For those of you who like to geek out, I'd really appreciate any feedback on my approach here, from a safety standpoint and anything else. While I've spent some time with small signal electronics, I haven't spent much time in the world of batteries, fuses, and big cables.
Method 3 is with individual cables from each battery to common bus bars. Here again is the link that was helpfully provided by Cheyenne:
http://www.smartgauge.co.uk/batt_con.html
To have balanced batteries, the cables need to be balanced. In my scenario, I experimented with matching resistance using larger guage, longer cables for the right bank, and smaller guage, shorter cables for the left bank. All while trying to meet ampacity, voltage drop, etc.
Jumping to the end... DC simulation shows good results with current imbalance of 0.4% without too much fine tuning, and total system voltage loss of less than 0.7% at 200A load and 1.5% at max 400A load. The cables seem smaller than I would have thought - 4 AWG for the left bank (each handles up to 100A continuous). But based on my understanding of the Ampacity specs and the Ancor wire specs, it seems to work. Comments?
Following are the details.
View attachment Method 3 simulation.pdf
Battle Born specs
- 100A continous
- 200A for 30 seconds
- 4X bank can supply 400A continuous
Inverter specs
- 2000W continuous
- 4000W surge
- 90% peak efficiency (assume peak is at full load)
- Calculates to 185A continuous, 370A peak @ 12V
- Kisae requires minimum 300A DC fuse or breaker, plus DC disconnect
Battery bank cabling and circuit protection
- 100A MRBF terminal fuse on each battery positive terminal
- Individual cables from each battery terminal to positive and negative bus bars
- 400A Class T fuse on output of positive bus bar feeding inverter
Wire sizing (would like feedback on these)
- Matched cable resistances between right and left banks for balanced battery current
- Each cable sized to have
- ABYC ampacity rating exceeding the fuse rating
- assuming 75C insulation spec (Ancor specifies 75C for wet, vs 105C for dry)
- and further derated by 0.857 for bundles of 4-6
- Overall system designed for
- 1% voltage loss from battery terminals at 200A load
- 2% loss at 400A maximum load (fuse rating)
Actual cable sizes
1) The left bank cables are selected to be minimum guage meeting the above constraints. This works out to 4 AWG and 3ft length. Resistance is 0.72 milliohms.
2) The right bank cables are sized to match the resistance of the left bank cables, using larger guage to compensate for longer length. Result is 1/0 AWG and 7ft length, 0.70 milliohms.
Simulation results at 200A load:
- Current imbalance between batteries less than 0.4%
- Voltage drop from battery terminals to load (assumed at bus bars) of 0.7%
Method 3 is with individual cables from each battery to common bus bars. Here again is the link that was helpfully provided by Cheyenne:
http://www.smartgauge.co.uk/batt_con.html
To have balanced batteries, the cables need to be balanced. In my scenario, I experimented with matching resistance using larger guage, longer cables for the right bank, and smaller guage, shorter cables for the left bank. All while trying to meet ampacity, voltage drop, etc.
Jumping to the end... DC simulation shows good results with current imbalance of 0.4% without too much fine tuning, and total system voltage loss of less than 0.7% at 200A load and 1.5% at max 400A load. The cables seem smaller than I would have thought - 4 AWG for the left bank (each handles up to 100A continuous). But based on my understanding of the Ampacity specs and the Ancor wire specs, it seems to work. Comments?
Following are the details.
View attachment Method 3 simulation.pdf
Battle Born specs
- 100A continous
- 200A for 30 seconds
- 4X bank can supply 400A continuous
Inverter specs
- 2000W continuous
- 4000W surge
- 90% peak efficiency (assume peak is at full load)
- Calculates to 185A continuous, 370A peak @ 12V
- Kisae requires minimum 300A DC fuse or breaker, plus DC disconnect
Battery bank cabling and circuit protection
- 100A MRBF terminal fuse on each battery positive terminal
- Individual cables from each battery terminal to positive and negative bus bars
- 400A Class T fuse on output of positive bus bar feeding inverter
Wire sizing (would like feedback on these)
- Matched cable resistances between right and left banks for balanced battery current
- Each cable sized to have
- ABYC ampacity rating exceeding the fuse rating
- assuming 75C insulation spec (Ancor specifies 75C for wet, vs 105C for dry)
- and further derated by 0.857 for bundles of 4-6
- Overall system designed for
- 1% voltage loss from battery terminals at 200A load
- 2% loss at 400A maximum load (fuse rating)
Actual cable sizes
1) The left bank cables are selected to be minimum guage meeting the above constraints. This works out to 4 AWG and 3ft length. Resistance is 0.72 milliohms.
2) The right bank cables are sized to match the resistance of the left bank cables, using larger guage to compensate for longer length. Result is 1/0 AWG and 7ft length, 0.70 milliohms.
Simulation results at 200A load:
- Current imbalance between batteries less than 0.4%
- Voltage drop from battery terminals to load (assumed at bus bars) of 0.7%
Thanks everyone for all the advice. I've made quite a few changes and I'm posting my updated schematic. I'll leave out the details on battery bank cable sizing, if you are interested in that please see the analysis I just sent in my prior post.
Here is the updated schematic:
View attachment Sprinter electrical schematic v3.pdf
Changes include:
1) Moved the positive lead to the BMV-712 shunt to one of the battery terminals, and added the temperature sensor
2) Eliminated the rooftop solar fuse and added a disconnect breaker/switch inside
3) Added 100A MRBF terminal fuses at each battery positive terminal
4) Rewired the battery bank for individual matched resistance cables feeding into positive and negative bus bars
5) Inverter fed by 400A Class T fuse followed by a remote battery switch--so my wife can have a button in the kitchen rather than a manual switch underneath the bench seat
6) Added a connection from House negative bus bar back to the ground post at the throttle pedal. Sized to match the DC-DC converter supplies, although I think it is more of a ground reference and won't actually carry much current? Depending on how balanced the supplies are on the charger.
Here is the updated schematic:
View attachment Sprinter electrical schematic v3.pdf
Changes include:
1) Moved the positive lead to the BMV-712 shunt to one of the battery terminals, and added the temperature sensor
2) Eliminated the rooftop solar fuse and added a disconnect breaker/switch inside
3) Added 100A MRBF terminal fuses at each battery positive terminal
4) Rewired the battery bank for individual matched resistance cables feeding into positive and negative bus bars
5) Inverter fed by 400A Class T fuse followed by a remote battery switch--so my wife can have a button in the kitchen rather than a manual switch underneath the bench seat
6) Added a connection from House negative bus bar back to the ground post at the throttle pedal. Sized to match the DC-DC converter supplies, although I think it is more of a ground reference and won't actually carry much current? Depending on how balanced the supplies are on the charger.
Justin Laureltec
I poke electronics
As a couple of notes, Airtime: your house bank chassis ground connection needs to be on the "load" side of the shunt, not the "battery" side, and the temp sensor is a two-wire connection with one wire being fused; it replaces the individual power lead that comes with the BMV-712. You may be showing that, but with the individual labeling it looked like you might be anticipating still using the individual power wire, so I just thought to clarify that a little.
Kevin.Hutch
2011 Mercedes 313 906
Yes a classic mistake with the alt charge bypassing the shunt and the load going through it.
I missed your point here the installation manual shows it as he has drawn it.
100 amp battery fuse on each battery - that works from a technical viewpoint. Consider if these trip during a vacation or if your wife is using the system and the ability to swap them when it is raining.
The simulation results are interesting. I am wondering about the effect of contact resistance at connections vs copper conductor losses. It doesn't take much for a wire lug to bus bar connection to to be imperfect, especially at those currents. As an example, the lug of a 1/0 wire is going to have nearly 2x the surface area connection to the bus bar as the 4 awg. I have no data on the contact resistance of this connection point but if your design has a risk area this will be the place that needs to be verified on a bench or after the build.
Fortunately it is relatively easy to build this all up on a bench and verify the results before installation. Worst case you just buy a few new wires. A clamp meter will be your friend.
I have never attempted to down size the wire to match impedance but that makes sense. The traditional method used is to just make the wires all the same size and length and coil up the extra length. Not saying that is right or wrong, just common.
My approach has been to wire the batteries in series so that I don't have to match wire impedance, but it is hard to argue with the simulation results - at least from a bulk conduction perspective.
The simulation results are interesting. I am wondering about the effect of contact resistance at connections vs copper conductor losses. It doesn't take much for a wire lug to bus bar connection to to be imperfect, especially at those currents. As an example, the lug of a 1/0 wire is going to have nearly 2x the surface area connection to the bus bar as the 4 awg. I have no data on the contact resistance of this connection point but if your design has a risk area this will be the place that needs to be verified on a bench or after the build.
Fortunately it is relatively easy to build this all up on a bench and verify the results before installation. Worst case you just buy a few new wires. A clamp meter will be your friend.
I have never attempted to down size the wire to match impedance but that makes sense. The traditional method used is to just make the wires all the same size and length and coil up the extra length. Not saying that is right or wrong, just common.
My approach has been to wire the batteries in series so that I don't have to match wire impedance, but it is hard to argue with the simulation results - at least from a bulk conduction perspective.
Last edited:
Justin: Thanks, I need to remember that the shunt is really just part of the battery, and the load side of the shunt is the House Negative.
Kevin: Actually the alternator charging current has a return path on the alternator input from the Kisae DC-DC converter to the chassis ground. And then on the output side, battery charging return current is connected to the Load side of the shunt.
The connection between house ground and chassis ground is a separate question. I think current really shouldn't be flowing in that except in a fault condition--depending on the isolation in the internal design of the charger and inverter. I've got some questions into Kisae support. But either way, connecting it to the load side of the shunt--the House Negative--makes sense.
Harry: Yeah, good points on the fuse swap inconvenience. But that really shouldn't happen unless I have a cable short, and then I want it to trip rather than have a fire. Say if I hit something in the road while driving. Otherwise it would take a 400A load to reach 100A on one battery, less if one battery was very imbalanced.
I do think I'll lower the Class T main fuse inside the van to 300A, so it trips first if (someone) runs the microwave and the induction cooktop at the same time and then makes an espresso while waiting...
On the contact resistance, I was trying to find info on that with no luck.
As you said, nothing that can't be bench tested.
Kevin: Actually the alternator charging current has a return path on the alternator input from the Kisae DC-DC converter to the chassis ground. And then on the output side, battery charging return current is connected to the Load side of the shunt.
The connection between house ground and chassis ground is a separate question. I think current really shouldn't be flowing in that except in a fault condition--depending on the isolation in the internal design of the charger and inverter. I've got some questions into Kisae support. But either way, connecting it to the load side of the shunt--the House Negative--makes sense.
Harry: Yeah, good points on the fuse swap inconvenience. But that really shouldn't happen unless I have a cable short, and then I want it to trip rather than have a fire. Say if I hit something in the road while driving. Otherwise it would take a 400A load to reach 100A on one battery, less if one battery was very imbalanced.
I do think I'll lower the Class T main fuse inside the van to 300A, so it trips first if (someone) runs the microwave and the induction cooktop at the same time and then makes an espresso while waiting...
On the contact resistance, I was trying to find info on that with no luck.
As you said, nothing that can't be bench tested.
Kevin.Hutch
2011 Mercedes 313 906
I think you will find the input and output negative in the Kisae DC-DC converter are connected together, either way the added wire is a return path to the alternator so this way they are anyway.
It would be interesting what your BMV charge current would look like with a short across the shunt. Ohms law knows no difference between intended an fault current, it takes the path/s of lowest resistance.
Not sure regarding your ampacity for 4 awg cable. You're using 4 awg for individual 100A max battery cables.
75 C, 4 awg cable is rated for 85A. It's close, but OK.
Looking at battery balance:
Cable resistance:
0/1 = 0.0983 m ohm/ft
4 = 0.2485 m ohm/ft
ratio = 0.2485/0.0983 = 2.528
That means if you're using 7 ft 0/1 cable, the 4 awg cable needs to be 7/2.528= 2.76 ft long. Your 4 awg cable is 3 ft in length. They are not balanced.
75 C, 4 awg cable is rated for 85A. It's close, but OK.
Looking at battery balance:
Cable resistance:
0/1 = 0.0983 m ohm/ft
4 = 0.2485 m ohm/ft
ratio = 0.2485/0.0983 = 2.528
That means if you're using 7 ft 0/1 cable, the 4 awg cable needs to be 7/2.528= 2.76 ft long. Your 4 awg cable is 3 ft in length. They are not balanced.
Thanks for replies!
http://assets.bluesea.com/files/resources/reference/21731.pdf
75C (wet), 4 AWG (21 sq mm) is rated for 125A for single wire, multiply by 0.857 for bundles of 4 to 6, works out to 107A for wet, bundled 4 AWG cable.
Balance--yes not perfect with round numbers in feet, but close. I could tweak the 4 AWG cables as you suggest.
But since I'm down in this rabbit hole, I'm looking at another alternative that has larger copper and perfect balance, at the expense of extra connections. I'll post that.
I checked. Yes, all three negative returns (alternator, solar, battery) are connected in common. The Chassis ground is a safety ground for the case and is isolated from all positive and negative current-carrying lines.
I see somewhat different numbers in the Ampacity table I got from the Blue Sea website:
http://assets.bluesea.com/files/resources/reference/21731.pdf
75C (wet), 4 AWG (21 sq mm) is rated for 125A for single wire, multiply by 0.857 for bundles of 4 to 6, works out to 107A for wet, bundled 4 AWG cable.
Balance--yes not perfect with round numbers in feet, but close. I could tweak the 4 AWG cables as you suggest.
But since I'm down in this rabbit hole, I'm looking at another alternative that has larger copper and perfect balance, at the expense of extra connections. I'll post that.
You'll need to protect the battery cable in a sheath. The Blue Sea ampacity table for 4 awg drops from 125 to 88A when cable is in a sheath at 75C insulation.
My data from:
https://www.cerrowire.com/products/resources/tables-calculators/ampacity-charts/
Battery balance will always be dependent on cable lengths, or equal cable resistance. Your new approach shouldn't be any different.
My data from:
https://www.cerrowire.com/products/resources/tables-calculators/ampacity-charts/
Battery balance will always be dependent on cable lengths, or equal cable resistance. Your new approach shouldn't be any different.
I mentioned this before, but what about a different battery / location? Is there no way you could fit 2 batteries inside the van? The Battleborns you are planning on using are the same price per Ah as the 200 Ah Victrons. Each Victron has twice the capacity in 15% less volume and saves 39lbs for 400Ah. For an extra $118, you also get an external BMS.
Putting 2 Victrons with 400 Ah total capacity inside solves the following problems:
Simple cable balancing / no need to for long cables running from one side of the van to the other.
Batteries are in the habitable space so no need for battery heaters for cold weather charging.
No holes in the floor for stiff, large-gauge battery cable to chafe on.
Simple on-battery fuse replacement (no lying on your back, in the rain, trying to change a fuse).
Fewer, shorter and therefore less expensive battery cables
Fewer connections
No need to derate cable rating from 105 to 75 deg C due to cable in dry location
Not that you plan is in any way a bad one - in fact it seems very well researched and thought out. However, for me, I'm willing to make space inside the van for these really expensive batteries. I never considered putting them anywhere but inside, directly adjacent to the inverter. On the other hand, if I were going lead acid, I'd definitely put them under the van.
Putting 2 Victrons with 400 Ah total capacity inside solves the following problems:
Simple cable balancing / no need to for long cables running from one side of the van to the other.
Batteries are in the habitable space so no need for battery heaters for cold weather charging.
No holes in the floor for stiff, large-gauge battery cable to chafe on.
Simple on-battery fuse replacement (no lying on your back, in the rain, trying to change a fuse).
Fewer, shorter and therefore less expensive battery cables
Fewer connections
No need to derate cable rating from 105 to 75 deg C due to cable in dry location
Not that you plan is in any way a bad one - in fact it seems very well researched and thought out. However, for me, I'm willing to make space inside the van for these really expensive batteries. I never considered putting them anywhere but inside, directly adjacent to the inverter. On the other hand, if I were going lead acid, I'd definitely put them under the van.
One other comment that may help in your design / simulation. Marine how to suggests a resistance of .00025 ohms per connection:
"Over years of field examples and tests. I tend to use a sizing factor of *0.00025Ω per connection point, as the resistance.. *This number assumes the connections are clean, tight and not made with a Dollar-Store crimp tool. You can see better performance than this with good tooling but I find 0.00025Ω pretty safe for sizing purposes."
This suggests that with very large wires and lots of connections, the circuit resistance is driven more by the number of connection points than the length of the wire. With, say, one foot of 4/0 wire it seems reasonable that there is more resistance in the crimp and bolt interface than the actual wire. At the very least the value above is something you can add to your voltage drop calcs and see if it makes a difference.
https://marinehowto.com/fusing-termination-voltage-drop/
"Over years of field examples and tests. I tend to use a sizing factor of *0.00025Ω per connection point, as the resistance.. *This number assumes the connections are clean, tight and not made with a Dollar-Store crimp tool. You can see better performance than this with good tooling but I find 0.00025Ω pretty safe for sizing purposes."
This suggests that with very large wires and lots of connections, the circuit resistance is driven more by the number of connection points than the length of the wire. With, say, one foot of 4/0 wire it seems reasonable that there is more resistance in the crimp and bolt interface than the actual wire. At the very least the value above is something you can add to your voltage drop calcs and see if it makes a difference.
https://marinehowto.com/fusing-termination-voltage-drop/
You make some really good points there... I'm going to take a closer look at this option.
If you do decide to go with the Victron, PKYS lists the 200 Ah Victron at $2217, but will lower the price to $2101.20 if you email them.
Midwestdrifter
Engineer In Residence
Note that some victron batteries do not have a built in disconnect, so you need to buy one in addition to the battery. They may need an external BMS control unit as well.
OK I learned a few more things.
First, 99sport there is no US stock of the Victron batteries that you suggested. Everyone is quoting 4-6 weeks, and I'm sure there's risk in that with the virus situation. Still thinking about it though.
As for my existing design...
Ampacity in a wet, sheathed environment (under van) is about half of what works for cool, dry, open air. I.e. go up about 2-3 sizes. That plus the batteries spaced 5 feet apart really makes cabling difficult.
I thought about putting a couple of isolated posts (like Blue Sea Power Post) on the frame in the middle and sending equal cables there, then one 4/0 each for pos and neg up inside. See updated schematic here:
View attachment Sprinter electrical schematic v5.pdf
But having those extra connections in a wet, corrosive, and hard to inspect place for up to 300 amp currents may not be the best idea.
I thought about going 24V with batteries series connected on each side, and then parallel the two sides, it would cut currents in half. But Kisae units are currently only available in 12V. I do see that Victron has 24V.
On the plus side, I do have a better handle on chassis grounding for the Kisae units after talking with Ricardo at Kisae. And I found that the Mercedes BEG in 8.4.2 specifies three ground bolts, on the frame, for use for in grounding. Each can have up to 4 terminals. I think this is the place to land the chassis safety grounds (i.e. green wire type of ground that don't carry any current in normal operation) rather than run ground cables back up to the starter battery ground.
Thanks everyone for all the help. At this point I'm leaning toward making space inside the van.
First, 99sport there is no US stock of the Victron batteries that you suggested. Everyone is quoting 4-6 weeks, and I'm sure there's risk in that with the virus situation. Still thinking about it though.
As for my existing design...
Ampacity in a wet, sheathed environment (under van) is about half of what works for cool, dry, open air. I.e. go up about 2-3 sizes. That plus the batteries spaced 5 feet apart really makes cabling difficult.
I thought about putting a couple of isolated posts (like Blue Sea Power Post) on the frame in the middle and sending equal cables there, then one 4/0 each for pos and neg up inside. See updated schematic here:
View attachment Sprinter electrical schematic v5.pdf
But having those extra connections in a wet, corrosive, and hard to inspect place for up to 300 amp currents may not be the best idea.
I thought about going 24V with batteries series connected on each side, and then parallel the two sides, it would cut currents in half. But Kisae units are currently only available in 12V. I do see that Victron has 24V.
On the plus side, I do have a better handle on chassis grounding for the Kisae units after talking with Ricardo at Kisae. And I found that the Mercedes BEG in 8.4.2 specifies three ground bolts, on the frame, for use for in grounding. Each can have up to 4 terminals. I think this is the place to land the chassis safety grounds (i.e. green wire type of ground that don't carry any current in normal operation) rather than run ground cables back up to the starter battery ground.
Thanks everyone for all the help. At this point I'm leaning toward making space inside the van.
You mean with regards to my comment about Kisae not having 24V? The thought of a 24V system was actually something I first thought about today, as I was wrestling with the implications of my undermount approach. I know you are a proponent of 24V, from the threads I have read.
I like the Kisae components, they seem to hit a good balance of functionality and cost-effectiveness. A total of less than $800 plus the various cables and breakers gets me 50A MPPT solar charging, 60A DC-DC alternator charging, 2000W inverter/charger with 55A shore power charging. Not magical, just a good price/performance point vs my needs.
Going to 24V is an interesting proposition, and if I ever wanted to do it now is the time. The fridge, future Rixen D5 setup, and (maybe) future Nations alternator setup are all available in 24V. Batteries can be configured for 12 or 24. Small 24 to 12 DC-DC converter for the small 12V-only loads. But it also means a whole different system when I had settled on Kisae for core components.
But yes I am thinking about this option because my current undermount design is stretching the limits of 12V. I either need to back off a bit on system power goals, or go to a different design.
Last edited:
Are you perhaps over-thinking this? You're not going to pull 100A from each battery. Using 4 awg should be OK.
The batteries will be balanced when the 0/1 length is 7.6 ft long while the 4 awg cable is 3 ft long.
The Blue Sea ML-RBS switch is very pricy ($250). Is it used to remotely turn off the inverter? This inverter has excellent reviews and the 2500W version works well for me.
https://www.ebay.com/sch/i.html?_from=R40&_trksid=m570.l1313&_nkw=reliable+power+inverter&_sacat=0
You can turn it off remotely by connecting a second switch in series with the inverter's power switch. I checked it out. The power switch conducts little current. That's a simple mod.
The batteries will be balanced when the 0/1 length is 7.6 ft long while the 4 awg cable is 3 ft long.
The Blue Sea ML-RBS switch is very pricy ($250). Is it used to remotely turn off the inverter? This inverter has excellent reviews and the 2500W version works well for me.
https://www.ebay.com/sch/i.html?_from=R40&_trksid=m570.l1313&_nkw=reliable+power+inverter&_sacat=0
You can turn it off remotely by connecting a second switch in series with the inverter's power switch. I checked it out. The power switch conducts little current. That's a simple mod.