Fusing Question

Herbi

Member
Joined
Sep 2, 2017
Messages
34
I am currently using 2A glas fuses. Like these:
https://de.aliexpress.com/item/200pcs-Lot-Glass-Fuse-3mm-x-10mm-3-10-250V-0-5A-1A-2A-3A-5A/32638019006.html

If I run 16A from 8 parallel cells, I get the following voltage drops over those fuses:
68, 67, 69, 83, 70, 67, 59, 65 mV. Take 1 or 2 mV for the soldering spots.
At ambient room temperature, they get about hand warm.

If I run 24A from them, I get average 160 - 190 mV and they get hot to the touch, but not unbearable. They dont blow yet, even its about 3A each for them.

1/6Watt-Resistor legs (about 1cm, decreases linear if length shortend):
1A=12mV
2A=30mV
3A=40mV
4A=60mV
5A=100mV
6A= up to 500mV, glows but does not blow

3A glas fuses:
1A=25
2A=50
3A=80
4A=160
4,5A= glows

I tried 0,2mm diameter copper wire, but its blows at about 15A only. Too much.


Resistor legs seem about right for what I need. I will try 5A glas fuses and report back.
 
The fuse wire VS the glass fuse dilema.

In a way I think we are looking at "fuses" in a bit of a wrong way. Fuses are meant to protect a device against an overload or short circuit.
Not to impede the electrical flow with a voltage drop. Voltage drop indicates resistance and resistance = heat.

So as we are testing these fuses we ramp up the amperageslowly to find the "melting"point.
Fuses have different characteristics of operating time compared to current. A standard fuse usuallyrequires twice its rated current to open in one second, a fast-blow fuse usuallyrequires twice its rated current to blow in 0.1 seconds, and a slow-blow fuse usuallyrequires twice its rated current for tens of seconds to blow.
Notice the twice its rated current specifications. The reason is that we do not want to impede the circuit with a voltage drop.

Rather than slowly ramping them them up I think the best way to test a fuse is to have a set amperage ready and available and short circuit the fuse.That is what the fuse was designed for. Overload and short circuit protection.
No matter what mA per cell we are running our powerwalls at, we need to look at our fusing criteria for our safety but yet not impede the current flow with excessive voltage drop per cell.

Ok enough about voltage drop.

I personally am looking at several scenarios on how and what type of fuse to implement into a powerwall design.

First of as we test all of our reclaimed batteries I hope we all are able to weed out the bad cells initially so we don't have the SD cells to begin with.
Let's say we have a 100p pack and out of those 100 cells we have 1 cell that is starting to go rogue on us. The problem I see is that this cell will gradually get worse and worse and not instantaneous. As this cell is slowly drawing down the pack I hope we would see this in our BMS system before it becomes a real problem. But if we don't seem to notice it then eventually we would have a "fuse wire" that is slowly getting hotter and hotter and eventually starting to glow. I don't know about you but a glowing wire in the open is somewhat worrying to me.
Looking at the Tesla fuse wiring I am wondering if that is what could be causing some of the spontaneous fires?

image_qooara.jpg

Hopefully our BMS system would alert us to this pack being way out of balance and we could address this in a timely manner and avoid any meltdown.

Another condition could occur we have a short high amperage draw and our endpoint breaker has not tripped as it has not reached its tripping point and all fuse wires glow a little. If the system is properlydesigned that probably shouldn't happen.

And of course a cell that spontaneously direct shorts.There would not be a lot of glow just a snap and the fuse wire has done its job.
I have seen that in action the only problem there was some cardboard nearby and the splatter ignited the cardboard. luckily I was close by and put out the smoldering cardboard.

For me that incident alone has steeredme toward a fast blow glass fuse. I prefer to keep my glow inside a glass container. :p


Wolf
 
Those values of voltage drops I gave above where of course only to estimate how your system would behave within nominal tresholds. E.g. assessing waste heat and the total voltage drop at the output of your system. Nobody said anything about purposely limiting the flow of current.
Should be obvious, we are not here to build a heating system, we want to build a battery.
 
Herbi said:
Those values of voltage drops I gave above where of course only to estimate how your system would behave within nominal tresholds. E.g. assessing waste heat and the total voltage drop at the output of your system. Nobody said anything about purposely limiting the flow of current.
Should be obvious, we are not here to build a heating system, we want to build a battery.

@Herbi

I think you took myresponse the wrong way :)I was just commenting on the fuse dileman we all have. We want to have a fuse to protect the pack but yet allow a low voltage drop across it. It is somewhat a balancing act of how low or high of an amperage fuse we go with.
Your voltage drop comparison was a great indicator of what happens when a "wire" or "fuse" gets close to its amperage threshold.
The point I was trying to make is that if that fuse is carrying an amperage just below its threshold it gets hot and potentially glows but does not blowas you discovered.
This is where the balancing act comes in. We want a fuse that is just the right amperage rating to stay cool during normal operation but if an adverse situation transpires it will protect us and blow.The problem I see is that if the unusual incident occurs of a cell slowly turning into a SD that the fuse will not blowin the beginning and only get hot hence my thinking is to keep that heat and "glow" in a glass container.
Good work by the way getting the mV voltage drop readings on a glowingwire without toasting the meter leads. :D

Wolf
 
Interesting results there Herbie.
Like Wolf is suggesting, one on side we need to protect the pack by having fuses on the cells in case of an accident like a major short (has happened to the best of us).
On the other side a tiny amount of heat dissipated is a necessary part of a fuses operation - it has to have some resistance or it would never blow!

It's interesting in your results how the voltage drop suddenly rises (=sharp temp rise of the fuse wire).

Another plus point for the glass fuses vs resistor legs is when they blow, the metal particles are contained, vs blobs of metal rolling around that can short the cells.

For your higher current needs, maybe consider higher rated fuses so you're further from the heating "knee" point.
 
Ahhh, the eternal FUSING dilema... I've come to the conclusion that we only need to fuse the + side of the cells. The other side could be made with a solid plate (like Tesla does on their packs) I'm scheming a way to come up with something similar. I'm getting tired of soldering both sides already!!! I'm looking at etching a board with tracks for the + sides and a whole plate for the -.
 
Yes, sorry guys. I had a bad day. You speak my mind. It is a difficult thing to achive a good compromise between making sure the fuse actually blows and an acceptable voltage drop.
I also discovered, that the lenght of a fuse wire determines when the wire blows. 1cm resistor leg blows at about 6-7A, 0,5 cm does not even blow at 15A, if current is ramped up slowly.
I guess because the heat can not build up quickly enough in the wire. That proofs the point Wolf made, that it is also a function of how fast the current is ramping up. If it ramps up fast, the heat has no time to dissipate into the soldering spots.
I came to the decision that I will not use resistor legs for fuse wire. It just leaves to much residue and debris after its blown. Next I will try 0,7mm of 35 AWG or 36AWG.
 
I believe it is better to fuse the positive side of cells in a battery pack, but with the current pack I am building it would be a much neater install if I fused the negative side. Since it is primarily for dead shorts or other major catastrophe ...and since they are all new cells does anyone see a problem with this approach?
 
It doesn't matter which end you fuse.
You only need one fuse per cell.
If you're using wire for the fuse, then -ve end is likely better as any blown bits won't get into the cell +ve end caps or short to edges, etc.

Note that proper fuse wire is different to copper wire & is designed to melt at a lower temp.
 
Hi ..thanks Redpacket. Actually I am using glass fast blow fuses designed to blow at a lot higher current than is normally used by others. That is because I am using new cells and I do not want to limit in any way the current they can put out. I just want an extra level of safety in case a short happens or some major catastrophe happens.
 
Any change to post what kind of glas fuses you use, where to get and cost? Maybe even a link.


Herbi said:
I am currently using 2A glas fuses. Like these:
https://de.aliexpress.com/item/200pcs-Lot-Glass-Fuse-3mm-x-10mm-3-10-250V-0-5A-1A-2A-3A-5A/32638019006.html

If I run 16A from 8 parallel cells, I get the following voltage drops over those fuses:
68, 67, 69, 83, 70, 67, 59, 65 mV. Take 1 or 2 mV for the soldering spots.
At ambient room temperature, they get about hand warm.

If I run 24A from them, I get average 160 - 190 mV and they get hot to the touch, but not unbearable. They dont blow yet, even its about 3A each for them.

1/6Watt-Resistor legs (about 1cm, decreases linear if length shortend):
1A=12mV
2A=30mV
3A=40mV
4A=60mV
5A=100mV
6A= up to 500mV, glows but does not blow

3A glas fuses:
1A=25
2A=50
3A=80
4A=160
4,5A= glows

I tried 0,2mm diameter copper wire, but its blows at about 15A only. Too much.


Resistor legs seem about right for what I need. I will try 5A glas fuses and report back.

I promised to come back for more:

For 5A fuses (3x10mm) I get about 25mV running 2A through them and about 65mV pushing 5A. Acceptable losses.
 
Back
Top