Class T vs ANL fuse

[John Frum]

These days, fuses are also specifically useful to protect semiconductors (google "semiconductor fuse"). With a fast acting fuse, your BMS will likely survive (I tested this). With a slower fuse, your wires will be fine, but you will likely fry the BMS. Keep in mind that short circuit current with a BMS that starts to smoke/burn might be much lower than a direct short on the battery terminals outside the BMS. LiFePO4 by the way also has a more limited short circuit current compared to e.g. LiPo or other chemistries.

The BMS should trip on short circuit current in microseconds.
The semiconductor fuse should blow in milliseconds.
Therefore if the semi fuse blows it means the BMS has already failed.
The question I have is, will the really fast blow fuses prevent the already failed BMS from causing a fire?

 

[upnorthandpersonal]

The BMS should trip on short circuit current in microseconds.
The semiconductor fuse should blow in milliseconds.
Therefore is the semi fuse blows it means the BMS has already failed.
The question I have is, will the really fast blow fuses prevent the already failed BMS from causing a fire?

So, from personal experimentation: I have used a 100A JK BMS, with a 225A class T fuse, and shorted the whole thing (48V LiFePO4 battery). The class T fuse blew before the BMS knew what happened. The BMS survived in all tests (I did this multiple times; this was when I could get a class T fuse for €20).

 

[Hedges]

I found "Technical Data Sheet" at Eaton that has the DC Breaking Capacity rating: 50 kA at 500 V d.c.

"High Speed Fuse", 350A FE, FM fuse at 2200A clears in 0.01 second
(Clears? or "virtual pre-arcing time", whatever that is?)

https://www.littelfuse.com/media?resourcetype=datasheets&itemid=78225034-cd5b-4006-8871-38e7378ac8f3&filename=jlln-fuse-datasheet

These have a graph for "Total clearing I^2 t", don't also see what current is guaranteed not to clear or damage them. Maybe it is "pre-arc" in the table.

Class T JLLN 300V is similar.

https://www.littelfuse.com/media?resourcetype=datasheets&itemid=2925cc6d-2278-44e0-8fdd-106f087f707e&filename=jlls-fuse-datasheet

Class T JLLS 600V, needs 4200A to clear in 0.01 second. 2200A would clear in 0.3 seconds

(You may need to search Littlefuse site to get datasheet; those links may have been served to me.)

https://www.farnell.com/datasheets/712509.pdf semiconductor fuses

350FM, 2200A "virtual pre-arcing time" 0.01 seconds, below that time (half a 50 Hz line cyle) curve extends to 12,000A 0.0001 seconds
These also have "pre-arc" and "clearing" in a table (energy given as current x current x time), I think only for AC not DC. Time to clear DC might be longer, at least where time works out to more than half a line cycle (which helps stop AC current)

I had expected class T to be slower, time delay compared to high speed semiconductor which is used to protect transistors. Looks like class T is fact acting, confirmed by Up North's tests. An by people blowing them with capacitor pre-charge.

Whichever you choose, make sure intended surges are well below clearing current/time. Deal with precharge of capacitors if needed.

 

[John Frum]

So, from personal experimentation: I have used a 100A JK BMS, with a 225A class T fuse, and shorted the whole thing (48V LiFePO4 battery). The class T fuse blew before the BMS knew what happened. The BMS survived in all tests (I did this multiple times; this was when I could get a class T fuse for €20).

What was the short circuit current duration set to in the BMS?

 

[upnorthandpersonal]

What was the short circuit current duration set to in the BMS?

That I don't remember. Whatever the default was - I did not change that.

 

[brum]

These days, fuses are also specifically useful to protect semiconductors (google "semiconductor fuse"). With a fast acting fuse, your BMS will likely survive

Good to know. Thanks!

The Bussmann 100FE fuse that I'll be using is a pretty fast one. It clears 500A in 0.01s.

 

[eemarty]

I just discovered this post from a few months ago which discusses using "BS88" style fuses in the LET and LMT series including the idea of bolting directly to the battery terminal. This forum is great. There are actually lots of options in the BS88 Style and all have at least 40kA interrupt ratings.


In fact almost every series in this whole Bussman "IEC High Speed Fuses" catalog is rated for >100VDC and usually 50kA interrupt. There are lots of cheaper options out there, they just require a lot of digging any may have low availability. Many are high cost. Any then you have to have a proper fuse holder....I wonder if any of these sizes happen to be compatible with the dimensions of Class T holders? I know there are already two sizes of Class T holders for <=200A and >=225A which seems to cause lots of confusion and ordering mistakes.

For class T fuses: The blue sea systems product pages are vague. They seem to just be selling other manufacturers fuses but they don't list the actual part numbers or give the datasheet so you can't actually look up the specs of the Class T fuses they sell.

I also found this cool cross reference guide from Edison. The 300V class T are the ones I believe blue sea systems sells. Maybe they change manufacturers based on availability and price.

 

[upnorthandpersonal]

I know there are already two sizes of Class T holders for <=200A and >=225A which seems to cause lots of confusion and ordering mistakes.

True, but for a good reason: so you can't put a 500A fuse in a system that needs to be protected under 225A.
As far as I know, those BS88 fuses (I have not checked/tried them all) don't fit a class T fuse holder. Even if the space between the holes matches on some (160 LET iirc), the size of the fuse itself makes it not fit in the Class T holder.

 

[brum]

Yep, holders are an issue with the Bussmann bolt-on fuses. I'll be using two single stud busbars for that purpose.

 

[hoosierrun]

I have accidentally shorted out my Battleborn batteries before and they are quick to open up with their internal built in resettable breaker. That is probably why they are supplying ANL fuses in their package.

 

[eemarty]

True, but for a good reason: so you can't put a 500A fuse in a system that needs to be protected under 225A.

I mean that's one reason. It provides a small amount of protection from putting a wildly wrong sized fuse on a system, but you can still put a 400A fuse on a system designed for 225A. And there are actually 5 sizes in the Class T fuse lineup catalog. Blue sea systems only carries 2 of the sizes. On the smallest size they range from 0.5A to 60A so you could potentially fit a 60A fuse into a system designed for 0.5A. I think the primary reason they have different physical sizes is that the engineering requirements of the fuses demand different physical sizes of fuse. They struck a balance between having a different fuse holder for every amp rating and having one huge fuse size which would make all the smaller rated fuses bulky and overpriced.

 

[eemarty]

As far as I know, those BS88 fuses (I have not checked/tried them all) don't fit a class T fuse holder. Even if the space between the holes matches on some (160 LET iirc), the size of the fuse itself makes it not fit in the Class T holder.

It actually looks like the BS88 LET series (25A to 180A) would fit in the 101-200A class t holder. Am I missing something?



Bussman class T holder:

 

[tigerwillow1]

I found "Technical Data Sheet" at Eaton that has the DC Breaking Capacity rating: 50 kA at 500 V d.c.

And I found a Mersen data sheet that gives the DC breaking capacity of 40 kA at 240 volts. Either way, on paper this is superior to class T.

https://ep-us.mersen.com/sites/mersen_us/files/brochure_BS88_Low-Voltage-Fuses-and-Fuse-Holders.pdf

On the Bussman spec, it's showing the 100 amp fuse breaking at over 150 amps after infinite time. The Mersen specs show the same thing. I had always assumed a 100 amp fuse would open up very close to 100 amps over a long time period.

 

[eemarty]

And I found a Mersen data sheet that gives the DC breaking capacity of 40 kA at 240 volts.

It's possible that the Bussman fuse has superior performance to the Mersen one and yet they both meet the BS88 requirements set by the queen herself back in 1988. (/s) It's also possible that they are equivalent and were simply tested/qualified at different conditions.

On the Bussman spec, it's showing the 100 amp fuse breaking at over 150 amps after infinite time. The Mersen specs show the same thing. I had always assumed a 100 amp fuse would open up very close to 100 amps over a long time period.

I find those time vs current breaking curves to be very interesting. People are talking about fast blow vs slow blow, but for example, in response to a 300A overload, a 150A fast blow might perform similarly to a regular 100A fuse. The Class T are described as "extremely" fast. But you also have to consider the Amp rating and look at the charts.

 

[curiouscarbon]

thanks everyone! learning a lot here.

 

[time2roll]

View attachment 119990
It appears they could be bolted directly to battery lugs and the cable bolted to the other side of the fuse which seems safer to me than having a separate fuse block because the fuse is closer to the battery. Then, it's less likely that a short would happen between the battery and the fuse.

I am not real excited about skipping the fuse holder. The fuse holder should have a cover that reduces a fault right at the fuse. Also the holder will keep the wires from moving or dropping to the frame should the fault separate the fuse halves.

 

[eemarty]

I bought a pair of 180LET fuses off Aliexpress for $8 ea including shipping and tax. I had to file out the hole a tiny bit to get the M8 bolt to fit. I think I could cover the whole assembly with something to protect from accidental shorts. A large enough heat shrink would work. Do you think this would battery terminal cover from blue sea systems would fit?

 

[John Frum]

I bought a pair of 180LET fuses off Aliexpress for $8 ea including shipping and tax. I had to file out the hole a tiny bit to get the M8 bolt to fit. I think I could cover the whole assembly with something to protect from accidental shorts.

That fuse assembly is a lever that could potentially put some bad stress on the battery terminal.
Might be better to orient the fuse over the battery, so a bakelite shim or similar can support the fuse.

 

[John Frum]

Perhaps kapton tape

Kapton® Tape - 1 Mil, 2

Maximum resistance to high temperatures. Recommended for masking circuit boards and splicing wires. Temperature range up to 260 C (500 F). Polyimide film with silicone adhesive for easy release. UL 510 flame retardant.ULINE offers over 38,500 boxes, plastic poly bags, mailing tubes, warehouse...

www.uline.ca

 

[FilterGuy]

OK.... I am going to wake up this old thread.

Everyone says we must use a class-t fuse. It is pretty much accepted as gospel on this forum and as far as I know, it is true that a Class T is needed.

Has anyone seen any actual studies of short circuit currents on real LiFePO4 systems with BMSs?

Victron reference designs don't use Class T fuses, instead, they use mega-fuses with AIC of less than 5KA. Victron is a very well-respected supplier with millons of deployments. How do I reconcile this?

 

[upnorthandpersonal]

If you want to take my word for it, I've tested Class T fuses in comparison with Mega fuses and am currently also testing BS88 fuses since Class T is almost impossible to get.

The main thing a Class T (and BS88) fuse offers over Mega for the same AIC is that they are faster. This means that in addition to protecting the wiring (typical use of the fuse) they can also protect the silicon (like a BMS). You'll often see this symbol for this type of fuse:

A 'fast acting' Mega fuse, say a 200A rated one, will take 0.3 seconds to blow at 1000A. A similar rated Class T fuse will break in something like 0.02 seconds at 1000A - an order of magnitude faster.

 

[FilterGuy]

If you want to take my word for it, I've tested Class T fuses in comparison with Mega fuses and am currently also testing BS88 fuses since Class T is almost impossible to get.

The main thing a Class T (and BS88) fuse offers over Mega for the same AIC is that they are faster. This means that in addition to protecting the wiring (typical use of the fuse) they can also protect the silicon (like a BMS). You'll often see this symbol for this type of fuse:

A 'fast acting' Mega fuse, say a 200A rated one, will take 0.3 seconds to blow at 1000A. A similar rated Class T fuse will break in something like 0.02 seconds at 1000A - an order of magnitude faster.

That is an interesting perspective that goes beyond fire protection. Even if a megafuse prevents fire, it may not protect the circuitry.

 

[upnorthandpersonal]

That is an interesting perspective that goes beyond fire protection. Even if a megafuse prevents fire, it may not protect the circuitry.

Yes! And from experience, it does work: just recently I shorted a 48V pack (with BMS) to test BS88 fuses as alternative to Class T. They succeeded - BMS still lives, and the fuse blew before the BMS could register the event as short. With a Mega fuse, no way. The BMS would register the short and disconnect, or worst case scenario, blow the fets if the BMS is not up to it/having a bad day.

 

[Hedges]

If BMS is able to cut off over-current, then fuse doesn't matter much.
Have to figure out what conditions could cause BMS to fail shorted. It's resistance might be enough to keep fault current within Mega fuse capability.
When I debugged a board with failed inrush limiter (back to back MOSFET), the higher voltage drop was across the melted MOSFET, lower voltage across the one turned on by gate voltage.

If fast fuse saves BMS, that is a win.

 

[John Frum]

Yes! And from experience, it does work: just recently I shorted a 48V pack (with BMS) to test BS88 fuses as alternative to Class T. They succeeded - BMS still lives, and the fuse blew before the BMS could register the event as short. With a Mega fuse, no way. The BMS would register the short and disconnect, or worst case scenario, blow the fets if the BMS is not up to it/having a bad day.

Do you have a product link for a bs88 fuse and carrier?
Preferably in North America.