EF3 Fuse as a Class-T Alternative?

[Nightguest]

I'm curious as to what those who have more knowledge on this matter would think about this.

I recently came across an interesting fuse type, the Adler EF3
Link with spec sheet: https://www.adlerelectric.com/products/detail/264

Until now, I've always been using Class-T fuses in my system, however, looking at this fuse, am I right in thinking that it could be a viable alternative?

From what I can see (and looking at the 150A version of the EF3 fuse, the breaking capacity is good (50kA@200Vdc).
However, it's the time-current curve that leaves me wondering if it is a viable alternative.

I'm basing my comparison on the info from the datasheet for the Littelfuse Class-T fuses
Link with spec sheet here: https://www.littelfuse.com/products/fuses/industrial-power-fuses/class-t-fuses/jlln.aspx

Looking at the time-current curve, my brain feels like it short circuits, so it would be awesome if anyone could either confirm or correct me.

Am I correct in concluding that at lower overload currents, for example, 500A, the EF3 fuse would be a tiny bit slower than the Class-T Fuse?
With the class T being around 0.2s and the EF3 being around 1.7s
*Strangely, a 150A Bluesea system Class-T fuse shows 0.02 to 0.1s trip time which further confuses me.
Context here: https://www.bluesea.com/products/5114/Fuse_A3T___Class_T_150_Amp

However, we tend to use the Class T fuse as wire protection against a short circuit in the cells, so herein lies my second question.
Given that the short circuit current is crazy level (6000 to 16000A depending on the internal resistance) I'm mostly interested in how long the fuses would take to blow at those currents.

Am I right in understanding that both the EF3 and the T-Class fuse would trip really fast, (quicker than 0.01s) however, the EF3 would likely be slightly "slower" ?

Unfortunately, the T-Class fuse charts end at about 720A@0.01s so I can only "guess" what the time for 6000A would be, however, the EF3 chart seems to indicate around 0.004s to trip for 6000A.

If I am correct:

• Would the difference in trip time be negligible enough, to make the EF3 a safe alternative to the Class-T fuse?
• Is that slight time-to-trip difference going to have any impact?
• Does the fact that the EF3 fuse is much slower at lower overload currents (e.g. 100%) end up being an issue, or can that be ignored as the primary use here is to protect the wires from a short circuit?

A big part of why I'm interested, other than the huge cost difference in the Netherlands (around 50e for Class T vs the 20e for the EF3) is that the EF3 Fuse fits in the Lynx Distributor, and based on the images here (https://lithiumaccus.nl/product/adler-ef3-zekering-200v-dc-ir-tot-50ka/) the fuses are also filled with sand, so they do have that arc protecting feature as well.

Would love to know what the others here think!

 

[Suijkerbuijk]

Hummm.

Yes you can.
But wy ?
Only if you go hyper mode with 100kwh battery pack with a high volts you can go for it .
For normal 12/24v no use
For a 48volts its really over use
Than how easy van you have a replacement or is it easy to order or store.

Than i see you come from the Netherlands.
And yes its more easy than a t-fuses
For usa and the rest not that easy.
Looks like it base on the nh00 fuse

 

[ricardocello]

I've looked at those, they appear to be great.
The EV market has spawned useful fuses.
I could find no standard fuse holders, probably because you bolt them into assemblies.

 

[robbob2112]

Looking at the time/current curves it looks like more than a tiny bit slower....I'm not sure I would use them.

 

[rhino]

As @robbob2112 mentioned you can see that in the charts below. Look at 2000A. For the T fuse it shows it is instantaneous for all of them whereas those EV ones have a slight delay. Since we don't really know how the stuff we have connected will handle that I'd lean towards the safer route.

 

[Nightguest]

Thanks a lot for looking into this as well!

To be clear, I've just brought (another) class T fuse for my system, so for me, from here on this is just for my own knowledge, and I have a feeling that others may want to know this as well.

Important: if at any point someone notices that what I'm writing is pure BS, please please please call me out (and please correct me, or at least give me some clue on what to study so that I can learn it).

The reason I'm writing this is that I'm not sure if what I write below, especially the calculation section, is correct or not. It frustrates me to no end that I have this fact in my mind (use a class T fuse), I know/trust that this is right, but if I were to be in a situation where someone asks me to explain why (in more detail) to justify it, I'd have no clue what to say.

When it comes to using a Class T fuse, the justification breaks down into two parts.

• We use a class T fuse due to its arc-extinguishing features.
  • This one is easy to understand, the AIC and the general knowledge about arcs covers this point very nicely.
• We use a class T fuse due to how fast it reacts.
  • It's this point that leaves me with a few questions.

First Question:
On these forums, I've read that the fuse protects the wires, nothing else (many times). Is this an absolute fact, or, is there more to it?

1. If it is purely true, does that mean that we're essentially accepting that should the worst case happen, and a short circuit triggers in a cell, the BMS and potentially the inverter etc are fried?
2. OR, are we saying that the Class-T fuse triggers fast enough that, if placed before the BMS, it could actually protect the BMS and the inverter + any other upstream components as well?
3. Or is there something else entirely that I've missed somehow?

Second Question
At what point is a fuse fast enough, and how do we measure this?

If I look at the EF3 fuse and the Class T fuse (150A each), once we hit the short circuit currents (I'm going with 0.4 milliohms of internal resistance, giving 8,000A for the current), both are faster than 0.001s.
If I drop the internal resistance to 0.2 milliohms, the short circuit current jumps to 16,000A, and the reaching time is even faster.

That's great, and it's "logical" that faster is better. Or, at least it should be logical, but there has to be a point where it's simply "fast enough". And I don't mean the upper limit/just quick enough, I mean fast enough to just accept that it's quick enough, provided the rest of the system is well spec'd (e.g. cables are large enough).

On top of that, how do we measure fast enough?

I've tried to think about this with my (limited) physics knowledge, but I have no clue if this is correct or not, and even if it is, I get stuck as it quickly becomes too advanced.

What I've done is calculate the energy dissipated in the system when a short happens using this forumla:

E (energy dissipated in joules) = I^2 x Rt (current squared times by Resistance times by time).

Let's say I use these values:

• 0.4 Milliohms for internal resistance.
• 0.0004s time (Class T)
• 0.0006s time (EF3)

Note: for this part, the "time" that I used for the EF3 and the Class-T fuses are both pure guesswork, it's mainly for the calculation - don't take the times as "accurate".

That gives me:

• The Class T fuse will let 25.6J through before popping.
• The EF3 fuse will let 38.4J through before popping.

That shows me that very little energy is let through.
However, this isn't enough (I guess), and it brings me back to the first question. Are we really just protecting the wires from overheating/causing a fire, or, are we (hoping) that other components are protected as well?

Because, if it really is "just the wire" as I've read then isn't this overkill by...a lot?
This is where it gets too complicated for me as my thought process was:

• If just wire, then we want to stop the wire from getting too hot to ignite the surroundings.
  • And melt, but, if the wires are spec'd correctly, then the surrounding material would ignite before the wires melt, I think.
• Obviously, it's a lot faster to get hot than to cool down, so we can't look at just "max temp before copper melts".
  • Melt point depends on the wire size, but I read that for a 50mm2 wire, it's 225C (no idea if wrong/right, but I'm ignoring it anyway).
  • For sanity's sake, I'd say 60C for the wire before it can start causing issues (this is probably far too aggressive, but hey).
• From here, an old friend of mine who studied physics at uni that I poked told me that I'm messing with the laws of thermodynamics and if I dug into it, I'd get a headache. He then pointed me at something called the Adiabatic Heating Equation.
  • He was right, this did give me a headache as I got to:

Time = (mass of the wire x heat capacity of copper x Temperature rise) divided by (current squared x resistance of the wire)

• Of course, it's not "that easy" as the mass of the wire is the density of the copper x cross-sectional area x length of the wire.
• And then the resistance of the wire is (resistivity of copper x length of the wire) divided by the cross-sectional area.

At this point I stared into space and he said "0.28s" to go from 25C (starting temperature) to 60C.

Remember my disclaimer at the top - I have no idea if this is BS or not, I wanted to write this out so that people don't think that I'm not (at least trying) to work things out myself first. However, my friend was right, I did get a headache out of this, and I'm not sure if this is the right way to look at this. Naturally, my curiosity is killing me.

Also, assuming that is somewhat right, it still only looks at the "protecting the wires" part of things. If that is wrong, then obviously the "time" will be based on how the survivability of the components. However, I'll stop here as just writing this out and thinking of it again is making my head spin!

So, does anyone here happen to know if any of this is correct, on the right path, or pure madness? I'd seriously love to know!

 

[robbob2112]

When it comes to using a Class T fuse, the justification breaks down into two parts.

• We use a class T fuse due to its arc-extinguishing features.
  • This one is easy to understand, the AIC and the general knowledge about arcs covers this point very nicely.

Yes

• We use a class T fuse due to how fast it reacts.
  • It's this point that leaves me with a few questions.

The right class T reacts fast - the JLLN series to be exact.

First Question:
On these forums, I've read that the fuse protects the wires, nothing else (many times). Is this an absolute fact, or, is there more to it?

1. If it is purely true, does that mean that we're essentially accepting that should the worst case happen, and a short circuit triggers in a cell, the BMS and potentially the inverter etc are fried?

We are hoping that the electronic survive - by using a fuse to protect the wiring we are preventing a house fire that will take out the electronic regardless

1. OR, are we saying that the Class-T fuse triggers fast enough that, if placed before the BMS, it could actually protect the BMS and the inverter + any other upstream components as well?

The BMS is about 100 times faster than a class T fuse - if it doesn't malfunction it should shut things down. If a MOSFETs inside the BMS are turned off while carrying a large amount of current they can potentially fry. When they die it is typically as a short that draws a lot of current then if the current is still flowing because of the lack of a fuse that short will become an open violently and potentially cause a problem. A class T stops the process about the time the short happens so you need to check over the battery and ensure normal function if you blow a fuse.

1. Or is there something else entirely that I've missed somehow?

Inverters - they also have mosfets inside them - the class T on that positive line will prevent a short in the inverter from causing issues for everything connected to it. There may be a temporary high current, but the fuse breaks the link before anything catches fire or smokes.

Second Question
At what point is a fuse fast enough, and how do we measure this?

All fuses have a time/current curve if they are real instead of fakes. You can look that up in the datasheet of the fuse. For a typically JLLN class T it will run for 10 minutes at double rated current but will blow in 0.01sec at 6 x rated current. There are also slo-blow versions of the class T that take several times that long to blow. So make sure to get the correct series. you want the ones that are very fast or fast not the slow version.

If I look at the EF3 fuse and the Class T fuse (150A each), once we hit the short circuit currents (I'm going with 0.4 milliohms of internal resistance, giving 8,000A for the current), both are faster than 0.001s.
If I drop the internal resistance to 0.2 milliohms, the short circuit current jumps to 16,000A, and the reaching time is even faster.

The idea behind the fuse rating is that it allows the rated current with no problem and if there are surges from the inverter demand it will allow those through without issue. The wires are sized so they won't melt or have a problem in the time it takes for a surge to happen. Only with an actual short does it blow super super quick. In the time it takes a fuse to blow your wire will get warm and may even get hot but it shouldn't melt the sheath.

That's great, and it's "logical" that faster is better. Or, at least it should be logical, but there has to be a point where it's simply "fast enough". And I don't mean the upper limit/just quick enough, I mean fast enough to just accept that it's quick enough, provided the rest of the system is well spec'd (e.g. cables are large enough).

On top of that, how do we measure fast enough?

I've tried to think about this with my (limited) physics knowledge, but I have no clue if this is correct or not, and even if it is, I get stuck as it quickly becomes too advanced.

What I've done is calculate the energy dissipated in the system when a short happens using this forumla:

E (energy dissipated in joules) = I^2 x Rt (current squared times by Resistance times by time).

Let's say I use these values:

• 0.4 Milliohms for internal resistance.
• 0.0004s time (Class T)
• 0.0006s time (EF3)

Note: for this part, the "time" that I used for the EF3 and the Class-T fuses are both pure guesswork, it's mainly for the calculation - don't take the times as "accurate".

That gives me:

• The Class T fuse will let 25.6J through before popping.
• The EF3 fuse will let 38.4J through before popping.

That shows me that very little energy is let through.
However, this isn't enough (I guess), and it brings me back to the first question. Are we really just protecting the wires from overheating/causing a fire, or, are we (hoping) that other components are protected as well?

Because, if it really is "just the wire" as I've read then isn't this overkill by...a lot?
This is where it gets too complicated for me as my thought process was:

• If just wire, then we want to stop the wire from getting too hot to ignite the surroundings.
  • And melt, but, if the wires are spec'd correctly, then the surrounding material would ignite before the wires melt, I think.
• Obviously, it's a lot faster to get hot than to cool down, so we can't look at just "max temp before copper melts".
  • Melt point depends on the wire size, but I read that for a 50mm2 wire, it's 225C (no idea if wrong/right, but I'm ignoring it anyway).
  • For sanity's sake, I'd say 60C for the wire before it can start causing issues (this is probably far too aggressive, but hey).
• From here, an old friend of mine who studied physics at uni that I poked told me that I'm messing with the laws of thermodynamics and if I dug into it, I'd get a headache. He then pointed me at something called the Adiabatic Heating Equation.
  • He was right, this did give me a headache as I got to:

Time = (mass of the wire x heat capacity of copper x Temperature rise) divided by (current squared x resistance of the wire)

• Of course, it's not "that easy" as the mass of the wire is the density of the copper x cross-sectional area x length of the wire.
• And then the resistance of the wire is (resistivity of copper x length of the wire) divided by the cross-sectional area.

At this point I stared into space and he said "0.28s" to go from 25C (starting temperature) to 60C.

Remember my disclaimer at the top - I have no idea if this is BS or not, I wanted to write this out so that people don't think that I'm not (at least trying) to work things out myself first. However, my friend was right, I did get a headache out of this, and I'm not sure if this is the right way to look at this. Naturally, my curiosity is killing me.

Also, assuming that is somewhat right, it still only looks at the "protecting the wires" part of things. If that is wrong, then obviously the "time" will be based on how the survivability of the components. However, I'll stop here as just writing this out and thinking of it again is making my head spin!

So, does anyone here happen to know if any of this is correct, on the right path, or pure madness? I'd seriously love to know!

I've done a lot of 'what if' calculations from current in the time it takes a fuse to blow... If the length of the wire is less than a foot there isn't enough resistance to heat it red hot in the short amount of time to blow unless it is vastly undersized. If the wire is very large and has a long run, say 12ft, the amount of copper acts as a buffer to contain all the heat being produced and then cool off again. It may get red hot but no ill effects as long as it didn't exceed the rated sheath capacity.

If you calculate out over time the buildup of heat in a wire you have to also take into account the passive cooling of being in an enclosed space, or free air or however it is installed along with ambient temperature. Basically any current over time calculation that is longer than around 10 seconds becomes very complicated.

Here is the one to boggle your mind -

8 parallel 100amp capable batteries - all hooked to a bus bar with 3ft of wire. Now short one of the batteries so it has a voltage not of zero but of about 2 volts low. As much current as you pour in it will always stay 2 volts low because of a shorted cell.

Now if you do the typically and put a fuse on the battery post you have major troubles - the 7 other batteries are going to dump current to their maximum into bad battery - which means the wire that was sized for 125amps on the bad battery now has 700amps coming at it.... in 10 seconds or so the wire is red hot with the sheath melting off and in 30 seconds the copper is vaporized... In theory the fuse at the battery should blow but the input could be a lot of amps and as the filament burns away it sustains an arc. This is where a high AIC or a fuse like the class T filled with sand is a good thing.

Now redo the calculation for a wrench shorting the 3ft cable at the point where it attaches to the fuse.... now there is no way to break the connection and you have 700amps until the steel melts in about 4 minutes. But the cable is gone in only a couple ... depends on how much oversize it is.

Now redo the calculation for a 10ft cable and hardly anything happens except the wrench melts

And if you fuse at the bus bar where all the batteries are connected - nothing happens except you blow the 1 fuse....

I've thought about this a bit.

 

[ricardocello]

I've thought about this a bit.

Well written and explained.

But we are dying to know if any of those events actually happened to you?

 

[J.H.]

I use blue seas class T…and blue seas/ bussman MRCB switch’s . I really don’t care about a small discount on important stuff.… Im baffled by all this ….
Class T fuses are tried and true…a known product with a great history of performance.

I try to save money when I can, but not on fuses, parachutes, self defense ammo , rappelling rope, heart surgeons, medications my life depend on….etc..

Jus sayin…

J.

 

[robbob2112]

Well written and explained.

But we are dying to know if any of those events actually happened to you?

I'll never tell... never ever

I use blue seas class T…and blue seas/ bussman MRCB switch’s . I really don’t care about a small discount on important stuff.… Im baffled by all this ….
Class T fuses are tried and true…a known product with a great history of performance.

I try to save money when I can, but not on fuses, parachutes, self defense ammo , rappelling rope, heart surgeons, medications my life depend on….etc..

Jus sayin…

J.

99% of the time the Blue Sea class T are JLLN Littelfuse. I but those directly and save a few bucks. I get them from Zoro, pkys, or other sources where I am certain I will get the real thing.

 

[J.H.]

I'll never tell... never ever


99% of the time the Blue Sea class T are JLLN Littelfuse. I but those directly and save a few bucks. I get them from Zoro, pkys, or other sources where I am certain I will get the real thing.

I have used Zoro for class T awhile back .. they had some good deals.. so did some of the marine supply places… if anyone is too cheap I back off.. somthings wrong.

J

 

[robbob2112]

The marine supplied ones I looked at were not JLLN, they were different time/current curve and slower. The ends were a beryllium copper plating verse silver... the beryllium copper is more corrosion resistant and the different class was meant for engineering spaces to start engines and the like. So high surges are expected and you don't want to blow the fuse doing it. But they were 1/2 the cost of the JLLN version. I ordered a fuse for $15 to see what the deal was because I trusted the source not to be trying to rip me off.

Ok, I admit it... I have blown a fuse from careless handling of a wire. The MRBF on the post went pft and was gone.

 

[J.H.]

The marine supplied ones I looked at were not JLLN, they were different time/current curve and slower. The ends were a beryllium copper plating verse silver... the beryllium copper is more corrosion resistant and the different class was meant for engineering spaces to start engines and the like. So high surges are expected and you don't want to blow the fuse doing it. But they were 1/2 the cost of the JLLN version. I ordered a fuse for $15 to see what the deal was because I trusted the source not to be trying to rip me off.

Ok, I admit it... I have blown a fuse from careless handling of a wire. The MRBF on the post went pft and was gone.

I got it….I have done rather careless things too.. it’s a great teacher if it don’t kill ya..
Just call me lucky Jim.

J.

 

[Nightguest]

Thanks for taking the time to answer!

I use blue seas class T…and blue seas/ bussman MRCB switch’s . I really don’t care about a small discount on important stuff.… Im baffled by all this ….
Class T fuses are tried and true…a known product with a great history of performance.

I try to save money when I can, but not on fuses, parachutes, self defense ammo , rappelling rope, heart surgeons, medications my life depend on….etc..

Jus sayin…

J.

I've got nothing against getting product A because it's known to work (in this case a class T fuse), and, as I wrote, my system uses only Class T fuses for each battery.

However, I do not believe that this means that people (I'm only talking for myself, but I suspect others think similarly), should just settle on it for all eternity. Availability fluctuates a lot, I've seen people on these forums and in others commenting that they often cannot find them in stock anywhere, so knowing what alternatives are safe is useful. And for that, we do need to understand what makes it safe.

@robbob2112
Thank you so much for that write-up, it did make me realise that I have overlooked and confused a lot of things.

I've done a lot of 'what if' calculations from current in the time it takes a fuse to blow... If the length of the wire is less than a foot there isn't enough resistance to heat it red hot in the short amount of time to blow unless it is vastly undersized. If the wire is very large and has a long run, say 12ft, the amount of copper acts as a buffer to contain all the heat being produced and then cool off again. It may get red hot but no ill effects as long as it didn't exceed the rated sheath capacity.

This was my biggest issue, I didn't really consider "longer wires" (and even that, by longer I mean...more than 2m).
However, after that, my question really remains.

You mentioned that it won't get red hot in a short time to blow - but then what is a "short time"?
This is the crux of my entire post and the source of my curiosity - we're using Class T fuses because they blow fast, but I cannot find any information on what is fast enough.

From the calculation that I've made, anything that is faster than 0.2s to blow is fast enough (provided that it has the arc extinguishing material as well.

*Big assumption here that my calculation/formula/theory was correct - of that I'm not sure, but as of writing this, I've not been told it's wrong)

If you calculate out over time the buildup of heat in a wire you have to also take into account the passive cooling of being in an enclosed space, or free air or however it is installed along with ambient temperature. Basically any current over time calculation that is longer than around 10 seconds becomes very complicated.

I've intentionally ignored passive cooling and looked only at the change in temperature from ambient (25C) to my given upper limit, 60C.
This was intentional because I wanted a "worst case", for me, in the Netherlands, 25C is not that common (unless it's mid-summer), and passive cooling would end up meaning that the fuse could remain intact for longer and be safe.

What confuses me here is, where did you get 10 seconds from?
Everything I've mentioned and calculated caps out at 0.28s.

The next section answered this, turns out I made a big mistake!

If someone suggested a fuse that would take 10s to pop, I'd call them insane and probably force someone to inspect their system out of fear for their own lives.

Here is the one to boggle your mind -

8 parallel 100amp capable batteries - all hooked to a bus bar with 3ft of wire. Now short one of the batteries so it has a voltage not of zero but of about 2 volts low. As much current as you pour in it will always stay 2 volts low because of a shorted cell.
Now if you do the typically and put a fuse on the battery post you have major troubles - the 7 other batteries are going to dump current to their maximum into bad battery - which means the wire that was sized for 125amps on the bad battery now has 700amps coming at it.... in 10 seconds or so the wire is red hot with the sheath melting off and in 30 seconds the copper is vaporized... In theory the fuse at the battery should blow but the input could be a lot of amps and as the filament burns away it sustains an arc. This is where a high AIC or a fuse like the class T filled with sand is a good thing.

Ok, maybe this part explains the 10s thing.

And, here I thank you because it seriously fixes a misconception that I had.
I failed to make a difference between a cell short-circuiting and a pack short-circuiting.
In your example, a single cell short circuits, resulting in the 700A / 10s that you mentioned.
I was assuming that the full pack short circuits, resulting in the 8000A.

I do find this to be a really important point.
It might be just me having selective memory, but, all of the guides, instructions, videos etc, seem to only focus on the "pack short-circuiting" not the cells. Even if I've misunderstood, it seems to be that no distinction is being made.

This is one of the reasons that I really wanted to make that larger post, as I could not find this information anywhere. It was always buried under "Class T is the best because it reacts fast", but turns out there are a lot of layers to it.

So yeah, honestly thank you very much!
I'm still mega curious about what is "fast enough" I still would love to know the theory behind that and what the target speed is.
But I see that I approached it incorrectly as it has to be fast enough not just for a pack level short, but also a cell level short - and these 2 types of short have a very very different current load (700A vs 8000A).

 

[Nightguest]

Just coming back to this one as I've learnt a lot but there's one nagging question in my mind, "what is fast enough"?

This all comes down to a massive hole in the data/knowledge that I can find:

Previously, I was considering a short circuit current of 8000A, but @robbob2112 very kindly corrected me, so it's 700A (which makes a massive difference to the reaction time of many fuses).

Going on with everything I've learnt thanks to you lot here, the fuse is here to stop my wires from melting.
In my case, the wire is 2m (longer cables would increase the time so).

If I assume 60C as the upper limit to what the wires can get to, then:
50mm2 = 36.6s
35mm2 = 17.8s
25mm2 = 9.1s

*Assuming the ambient temperature is 25C, and the wires are not going to cool in the air, making this a worst-case scenario.
Of course, no one in their right mind would want a fuse that takes that long to trigger, regardless of the size of their cables.

Searching on the forums, I see that NH fuses are accepted as good, provided that they're not the gG variants, looking at a few I-T curves, 700A gives us:

• gG/gL: about 1.5s
• gPV: about 1s
• gS: about 0.02s (I've seen this advertised as a battery fuse)
• aR seems to be very close to the Class-T fuse - and I've read on these forums that these are a viable alternative to Class-T.

There also seems to be a gBat fuse, but they make Class-T fuses look cheap so hey.

Going back to the EF3, 700A seems to be around 0.2s

My question is, what is fast enough?
If we go by heat alone, all of them are fine.
I thought we'd want to protect the other components, but, my understanding is that we kind of want to just pray they survive, even with a Class T fuse.

Just to reiterate so I'm not giving off the wrong impression here, I have a Class-T fuse in my system, I simply want to learn more about the speed of the fuse.

 

[robbob2112]

The aR is fast enough with nearly the same curve. Because you are worried about wire sheath melting and shorts, you make sure your mounting surfaces are not flamable. Also, ideally non-conductive. So cement board is the simplest solution. Then secure the cables with metal clamps so if the sheath does melt it doesn't short and cause a fire.

In the US all DC wires are supposed to be in metal conduit inside a dwelling. Sheath melts and they short and the fuse blows in a few microsecond if it hasn't already blown.

Where it isn't in metal conduit it should be mounted on the non-combustable surface and secured.

One other tidbit to make it harder - there is a thermal curve/coefficient for the fuse. Higher current for longer times means the fuse blows faster than when it is cold and has nothing flowing through it.

There are hydraulic/magnetic breakers that don't care about the temperature like your standard thermal breakers.

Another style are the MCCB breakers with a shunt(relay)-trip added. The shunt-trip is just a coil with a solenoid attached to a lever to trip the breaker off. The ones I looked at have 12v or 24v coils. So you power that could from a PSU on the load side of the breaker. Then you just need something with a dry contact to trip the breaker remotely. Connect to a cerbogx or Rpi or inverter... those connected to the shunt and if current exceeds your programable limit the power is cut. These on aliexpress run $40 for the breaker and $15 for the trip mechanism that is added afterwards. Note, in my answer thread (first signature link) there are pictures of the guts of a mccb breaker opened up.

Example of MCCB with relay trip
Schneider and midnite solar make din mount breakers with the same sort of trip. The main thing is to power it from the load side of the breaker so it turns off the relay once the breaker is tripped and it doesn't burn out the coil. Or you can have a control circuit that just sends a pulse to trip it then turns off.

There are others with the psu and limit that is a dial so the breaker is self contained, but those are $$$.

Now, anything mechanical is slower that a fuse in a short situation but in an overcurrent situation where the current is double the class T takes 10 minutes, but a shunt-trip takes only a second or so.

The other thing is you can hook the RSD system to the shut-trip and cut PV power and battery power from one button.

My point is you can build layers for safety. I am considering How I want my final design but this will figure into it along with some NO DC contactors to do pre-charge. When this looses power it would disconnect the batteries. Hook that to RSD and it could cut the power. There are many ways to do things.

Layers to things with the fuse either the first or last line of defense depending on what happens. The number of layers depends on risk tolerance and funds. Each layer adds more connections and loss because of the added resistance.

And we hope the gear and the building survive.

 

[Kornbread]

Throwing this out there. a3t seems to be 'very fast acting' while jlln is 'fast acting'. Grainger says jlln, jjn, a3t, and tjn are equivalent. That opens up offerings by Bussman, Mersen, Shawmut, Blue Sea, as well as LittleFuse.



xxx From Grainger ...

xxx From Grainger ...

xxxx

 

[PyroRider]

We just had these fuses in a large victron project here in germany, they fit perfectly in Lynx distributors from victron as mega fuses are definitly not suitable for the short circuit currents from LFP batterys.

 

[PyroRider]

I've looked at those, they appear to be great.
The EV market has spawned useful fuses.
I could find no standard fuse holders, probably because you bolt them into assemblies.

they fit into standard mega fuse holders

 

[ricardocello]

they fit into standard mega fuse holders

It does appear that indeed they do
EF3


MEGA


Beware of the height difference, however.
And they are a little longer.

 

[Kornbread]

FWIW, just received two 175amp class t fuse holders and fuses from Don Rowe dot com. Both had a3t LittleFuse fuses instead of the jlln LittleFuse. This is the first time they sent a class t fuse kit with something other than a LittleFuse jlln.

 

[ricardocello]

FWIW, just received two 175amp class t fuse holders and fuses from Don Rowe dot com. Both had a3t LittleFuse fuses instead of the jlln LittleFuse. This is the first time they sent a class t fuse kit with something other than a LittleFuse jlln.

My impression is that A3T is brand-specific (Mersen), and JLLN (Littelfuse), but they are equivalent.

A3T - Class T - Fast-Acting | Mersen Electrical Power: Fuses, Surge Protective Devices, Cooling & Bus Bars

ep-ca.mersen.com

 

[OTRwSolar]

Voltage drop? Blue Sea Class T vs. Blue Sea/Littelfuse MEGA (of the same rating)?

 

[ricardocello]

Voltage drop? Blue Sea Class T vs. Blue Sea/Littelfuse MEGA (of the same rating)?

I just got a new 4-wire YR1035+ meter to measure milliohms, and I'm measuring everything I have.
It's better than digging through data sheets.
I don't have your exact comparison (but Blue Sea uses Littelfuse for both Class T and I believe).

So for example:
125A Class T JLLN Littelfuse (20 kA AIC): 0.65 mΩ
125A MEGA 70V SF51 Littelfuse (2.5 kA AIC): 0.43 mΩ

Use V=IR to compute your voltage drop. All measurements taken at 25 deg C, zero current flowing.
Is the Class T extra resistance worth it for 8x the interrupt capability? It sure is for me.

 

[OTRwSolar]

Thanks for the reply/info, @ricardocello. Does a Class T offer significantly more interrupt capability at 12v (vs. a MEGA)? I can certainly see the advantage of a Class T with a very high current, 48v system. For the time being, our biggest concern is voltage drop. We get a fair amount of heat when pushing 150a through a 300a Littelfuse MEGA (35mv drop). I was hoping a Class T might help us in this regard.