Current Sharing with Large LiFePO4 Packs

[eabyrd]

Someone knows why NH00 - NH1 blade fuses (and fuse holders) are not used in US ?

They are rated 500-690Vac, 250-450Vdc, 100-150 kA withstand.
The NH00 is max 160A, the NH1 is max 250A, the NH2 is max 400A.

And they are dirt cheap.
The NH00 3 fuse holder and separator is like $50. (as we saw on many fire cases, best to fuse both poles )
And a fuse is about $7-$10 for it.

Slick looking design. No idea why they aren’t used here

 

[jameshowison]

Can someone point to guidance on sizing the Class T fuse for these larger packs? I'm thinking because the fuse it protecting the wire then it is sized the same way as any other fuse?

 

[upnorthandpersonal]

Can someone point to guidance on sizing the Class T fuse for these larger packs? I'm thinking because the fuse it protecting the wire then it is sized the same way as any other fuse?

In principle yes. In practice, I always design a system based on the parameters of the load. If the load is known to have surges, I want to make sure these surges are within the normal operating parameters of the wiring, BMS, battery, etc. Then I choose a fuse that is rated 25% or so over the maximum surge current that is drawn. This makes sure that a) the wires are protected from currents for which they're not rated b) all equipment such as the BMS operates within their specified parameters, c) if there is a short, the surge capacity of the BMS should make sure the BMS is not destroyed before the fuse pops and d) use a fast acting class T fuse to make sure that any current over it's rated value causes a very quick disconnect saving equipment from potential further damage (e.g. short in the inverter).

 

[JJJJ]

In considering bus bars, I found the following chart helpful. It discusses amp capacity for copper bus bars. Perhaps there is a better source for this:

Electrical: Busbar - DC Copper Busbar Ampacities

The following tables have been provided by the Alliance for Telecommunications Industry Solutions (ATIS), T1 Committee, and represent ampacities for busbar sizes and arrangements typically found in the telecommunications industry.

www.copper.org

 

[Ozark Tinkering]

amp capacity

ampacity

 

[eabyrd]

The question I always had was whether pure copper bus bars needed Anti-ox if you're putting plated terminals against them, or if you're applying them directly to the terminals of cells. Common sense says yes, but my common sense has been wrong before.

 

[mrzed001]

The question I always had was whether pure copper bus bars needed Anti-ox if you're putting plated terminals against them, or if you're applying them directly to the terminals of cells. Common sense says yes, but my common sense has been wrong before.

NoAlox, AntiOx, ... is not the best choice. They are electric insulators. They make the contact even worst. That causes more heat and loss.

In one case a user put so much NoAlox in the terminal that it could not carry the current. And the current gone up in the screw and through the washer into the busbar (almost welding it)

16S Active Balancer just "exploded"?

Hello my friends, quite new to DIY battery packs but I would say, that I am really sure what I am doing from an electrical perspective - until now ;). I finished my first project with building a ~10KWh 16S 48V LiFePo4 battery for my 5KW hybid inverter which is connected to 10 of 310W...

diysolarforum.com

MG Chemicals 846-847 carbon grease is one of the best choices:

MG Chemicals 847 Test

I'm posting this as a response to a comment from Paul F. Note: mR == milliOhm. Hypotheses I proposed: MG Chemicals 847 can lower the resistance in a connection between bus bar and battery terminal, especially in cases where limited torque can be exerted, or where the two connections are not...

diysolarforum.com

 

[JJJJ]

I am considering heat shrink for the sections of the bar that don’t have the bolt connection. This is used on the overkill Bms and in the engine compartment of vehicles so it is effective in very harsh conditions. It will double as a an insulator in case accidental contact is made. Admittedly both a positive and negative contact would be needed but this would add another layer of protection.

 

[HRTKD]

Covering a bus bar with heat shrink would - in theory - affect the ability of the bus bar to shed heat. In that case you would have to down grade the ampacity of the bus bar.

How much of an affect isn't something that I can quantify. For now, my bus bars will remain bare.

 

[Ozark Tinkering]

Noalox has no place around copper.
It is made expressly for use on aluminum conductors and pipe threads.

Electricians use the stuff for 2 things only.
#1 use; coating aluminum conductors at splices and terminations.
#2 use; putting on threads of rigid conduit to stop rust and increase conductivity for continuous mechanical bonding in a conduit run.

Noalox contains no aluminum or copper. It contains zinc which is not a great conductor, but it does stop oxidation on aluminum and steel.
Aluminum is notorious for oxidation and is a continuous process when exposed to air without a protective coating. Connect it to copper and electrolysis really takes off. Back in the 60s thousands of mobile homes were wired new with aluminum wiring and because terminations were made improperly without noalox ended up burning down. Even more burned down when people had to change out receptacles and light fixtures and disturbed already dangerous wiring. Eventually most states banned the sale of mobile homes with aluminum wiring, but the code still allows it for service entrance feeders. Copper doesn't have that issue and nothing more than a thorough cleaning with alcohol before making the term will do the job. Just keep it clean.

Save the heat shrink for your shovel handle.

The question I always had was whether pure copper bus bars needed Anti-ox if you're putting plated terminals against them, or if you're applying them directly to the terminals of cells. Common sense says yes, but my common sense has been wrong before.

 

[JJJJ]

Given the reaction what type of metal would be best for bolt connectors on the bus bar?

 

[curiouscarbon]

Given the reaction what type of metal would be best for bolt connectors on the bus bar?

nickel plated copper is regarded by some as an option to mitigate galvanic corrosion

e.g.

Making BusBars And Nickel Plating

I put together a couple of 25 ah 24 Volt 8S Packs and made my own busbars for this. I chose 1/8" X 3/4" 110 Copper bus bars. •https://www.ebay.com/itm/303841271273 These are rated for a little over 200 amps. Way more than these 25 ah cells will push. I'm deciding to learn how to make my own...

diysolarforum.com

 

[Zil]

Someone knows why NH00 - NH1 blade fuses (and fuse holders) are not used in US ?

They are rated 500-690Vac, 250-450Vdc, 100-150 kA withstand.
The NH00 is max 160A, the NH1 is max 250A, the NH2 is max 400A.

And they are dirt cheap.
The NH00 3 fuse holder and separator is like $50. (as we saw on many fire cases, best to fuse both poles )
And a fuse is about $7-$10 for it.

They are used, or some other brand similar. I see them often on main disconnects. But they are alternating current. We are talking about direct current.

 

[upnorthandpersonal]

There are NH fuses for DC use, such as the gS and gPV categories for Battery and PV applications respectively. However they usually are rated at much higher voltages, and you pay for that. For example, a NH1 gS 690V/200A fuse costs something like €65 or so. They're also huge...

 

[mrzed001]

There are NH fuses for DC use, such as the gS and gPV categories for Battery and PV applications respectively. However they usually are rated at much higher voltages, and you pay for that. For example, a NH1 gS 690V/200A fuse costs something like €65 or so. They're also huge...

For example the NH00 gG is 500Vac and 250Vdc (more than enough for a 48Vdc battery)

https://www.tme.eu/Document/7d253a9842855876dfe27734b478f852/Bus_Ele_DS_2142_NH.pdf

And a 160A fuse costs only $7.44

160NHG00B BUSSMANN - Fuse: fuse | 160A; 500VAC; 250VDC; gG,gL; ceramic,industrial; NH00 | TME - Electronic components Hungary

BUSSMANN 160NHG00B | Fuse: fuse; 160A; 500VAC; 250VDC; gG,gL; ceramic,industrial; NH00 - This product is available in Transfer Multisort Elektronik. Check out our wide range of products.

www.tme.eu

But yeah, they are HUGE ? No chance for an ark to remain in that big sand filled fuse.

 

[upnorthandpersonal]

Very slow though... a 160A fuse like that will take over 30 seconds to blow at 500A. The one you linked to even 50 seconds.

 

[mrzed001]

Very slow though... a 160A fuse like that will take over 30 seconds to blow at 500A. The one you linked to even 50 seconds.

A CLASS-T JLLS 175A needs 20sec to blow in 500A. The JLNN 3 sec.

The NH00 125A fuse with 1000A will blow in 0,6 sec.
With 2000A will blow in 0,015 sec.

The NH00 160A will blow with 1000A in 3 sec
With 2000A in 0,13 sec

And we know that LiFePO4 battery can deliver 5-30 kA in a short circuit.

Class-T is max 20kA interrupt in DC.
For a bigger battery or more parallel LiFePO4 strings ... that can be less than enough (Like Will's Tesla sized battery bank ).

So the fuses need to be designed for the system. How slow or fast acting needs to be planed.


Short info for beginners:
Only the Amps matters if we calculate energy loss in a wire. And this energy loss is heat. That heat blows the fuse.
Does not matter if you use the fuse with 5Vdc or 50Vdc or 500Vdc ... the same heat is generated at 100A.

 

[Ozark Tinkering]

Given the reaction what type of metal would be best for bolt connectors on the bus bar?

We use stainless in the ac world.

 

[CamoGreg]

Fuses like this?

Conext Battery Fuse Combiner Box 160A Fuses

Note: This product is discontinued. Please look at the Victron MEGA fuse 125A/58V and the Victron Fuse Holder 6-way as an alternative option.

www.solar-electric.com

 

[upnorthandpersonal]

The 160A will blow with 1000A in 3 sec

That's what I mean. A class T fuse of the same rating, at 1000A, will blow at something like 0.05 seconds. Remember, you don't always get the full short current when a fault occurs. For example if an inverter MOSFET blows, you don't get the full short current, but you could be at 1000A - and at that stage, you're generating a ton of heat inside the inverter and start burning stuff inside.

 

[mrzed001]

That's what I mean. A class T fuse of the same rating, at 1000A, will blow at something like 0.05 seconds. Remember, you don't always get the full short current when a fault occurs. For example if an inverter MOSFET blows, you don't get the full short current, but you could be at 1000A - and at that stage, you're generating a ton of heat inside the inverter and start burning stuff inside.

I just now updated the post with more data ... sorry took so long

 

[upnorthandpersonal]

A CLASS-T JLLS 175A needs 20sec to blow in 500A. The JLNN 3 sec.

And at your quoted 1000A for the other fuses, in 0.6 seconds for the JLLS and the JLLN at less than 0.01 seconds. That's why I go for Class T, every single time. The fastest ones I can get. As soon as a short (or short-like) fault occurs, I need the fuse to turn it all off immediately. I'm sure you can use the other fuses, but I will never, and never recommend anyone to do so. I've seen what even 1000A can do in 0.5 seconds.

 

[mrzed001]

And at your quoted 1000A for the other fuses, in 0.6 seconds for the JLLS and the JLLN at less than 0.01 seconds. That's why I go for Class T, every single time. The fastest ones I can get. As soon as a short (or short-like) fault occurs, I need the fuse to turn it all off immediately. I'm sure you can use the other fuses, but I will never, and never recommend anyone to do so. I've seen what even 1000A can do in 0.5 seconds.

If there is 500 or 1000A going through the inverter even for 0,1 sec ... does not matter. That is toast.
The fuse main purpose is that you have to protect the wire ... and also you can protect the equipment on the other end.

Do not forgot that CLASS-T is only 20kAIC.
A 16s 304Ah battery pack is already about 10-15kA SC. A 2p16s is possible over it (I never recommend but some use it).

Also if you have 3-4 packs parallel ... way over the CLASS-T scope.
Yes, you could (and should) put a CLASS-T on every single pack ... but you still need a central fuse and disconnector for the inverters.
And for that job the Class-T is not in the right league.

 

[upnorthandpersonal]

Yes, you could (and should) put a CLASS-T on every single pack

I thought pack fuses was the main thing we were talking about, as paralleling multiple packs and current sharing (and the potential for cell shorts etc.) was the topic of the video. You can have an additional non-class T fuse as a main fuse, but in case something goes wrong, the Class Ts wil most likely blow before your main fuse.

If there is 500 or 1000A going through the inverter even for 0,1 sec ... does not matter. That is toast.

The extend of the damage will be different. It can mean the difference between one FET breaking without smoke, and the entire chain going down with that lovely burnt electronics smell.

A 16s 304Ah battery pack is already about 10-15kA SC.

If we assume an internal resistance of 0.5mR, 16 cells in series gives an internal resistance of 0.008 Ohm. At a pack voltage of (rounded up) 60V, you are talking about a short circuit current of 7500A. LiFePO4 short circuit current is less than LiPo for example, and that's a good thing.

 

[JJJJ]

I was thinking about the mounting of bus bars. There has been discussion on other threads about materials, screwing into wood, hardy board, etc.

In the old days, Porcelain was used in the old knob and tube wiring. Admittedly, the positive and negative lines ran separately, but they are still used in fencing and such.

I found this product interesting. I know there are insulators on metal brackets which can then be attached to a surface. The reason I found this interesting is that the screw connection could potentially be separated from the screws which hold the bus bar. A theoretical idea at this time, but it caught my eye.

https://www.amazon.com/uxcell-Tempe...p/B01IX0Z8UA/ref=psdc_306731011_t4_B07Q6KDP9K