Are Tesla Style Fuses Enough?

cstanley

Member
Joined
Jan 2, 2017
Messages
43
I question this because I have heard of 4P 18650 charging at 1 amp (divide by 4 per cell) 3.7v get so hot you couldn't touch it. The fuse didn't blow because it didn't break the amp threshold but still got super hot. What about using Thermal fuses? They break at a certain amp AND temperature. What's your thoughts? Link below:
Thermal fuse 72C Degrees 250V 10A Thermal Cutoff
http://s.click.aliexpress.com/e/Z7YjmqN

This is a quote by Steve McMechan on YT:
I personally tested using the fuses this style (resistor) and I don't believe it offers enough protection personally. I setup a simple 4P 18650 battery holder that I'd soldered cell fuses onto (using the same ones you used that Jehu recommended), and there were 2 bad cells out of 4 in the 4P set, and I was only charging at 4v 1amp current, and within 5 minutes the 2 bad cells got so hot I could hardly touch them without burning my fingers, yet the cell fuses DID NOT BLOW (as the cells were not sucking more than 5amps to blow them). So I replaced the cell fuses with mini 10a 70C thermal fuses and that did the job perfect, protecting the pack from the bad cells when they over-heated above 70C they blew the fuses nicely (even though they were only sucking around 1amp of current). I strongly recommend needing to use thermal fuses not just the resister legs as you've used, as that's a fire hazzard as I found from my own personal testing sorry. The Thermal fuses are super cheap anyhow, and just as easy to solder in place but offer much better protection for each cell.?

What are your thought everyone?
 
I agree that the fuses probably won't do much in a 4p pack. I think most of us are building much larger packs though. I'm using 120p packs, many here are using 80p packs. If you charge the pack at 20A (which is common in my setup), that is 0.16A per cell. The fuses shouldn't get warm at all. Now let's say each cell can supply 2A of current, admitting that this is probably the extreme low end since most can do 5A+ very easily. If you assume 2A times 119 cells, that fuse on the bad cell is now going to carry 238A and will surely blow. The idea of fuses in my circuit is supposed to prevent a short cell, not a hot cell.

The more strings you add, the less the load per cell is. The goal in my setup is to get this as low as possible. If I have 4 strings of 120p packs charging at 20A, now that's 0.04A per cell. There is certainly no problem with fuses there :)

A good test would be to have some bad cells that get hot and put in a large pack to see what happens. Maybe I'll try to get some test together considering I have a huge case of bad cells.
 
Fuses will do their job whatever size pack you have - big or small.

But if folks are just cutting legs off resistors and expecting them to "work" they might be in for a unpleasant supprise - any wire that is intended to be used as a fusable link must be tested - properly - including work to determine what length is to be used, and when installed that length must be kept consistent with your test results.

Thermal fusable links are a good idea, alongside traditional individual cell protection fusing.
 
This is why i dont use fuses at all, after testing over 3000 second hand cells from over 500 laptop batteries, none of the bad ones were short circuit. Less than 50 cells got hot when charging (even cells that tested 0v) The only cells that got super hot (70c+) were about 8 of the red sanyo cells, and these got really hot @ 1000mAh charge, and barely warm at less than 500mAh, at the 100-200odd mAh in the power wall, they didnt do anything abnormal. BTW, the evil red sanyos were not <3v when initially recycled, but they did consume abnormal current without going much higher than nominal.

mike said:
A good test would be to have some bad cells that get hot and put in a large pack to see what happens. Maybe I'll try to get some test together considering I have a huge case of bad cells.

Having done this, I can confirm that your initial hypothesis is confirmed, when charged at <500mAh cells didnt get hot, but some never got to 4.20v, they hovered at 4-4.1v and consumed 50-100mAh, warm but not hot.
 
3nergE said:
mike said:
A good test would be to have some bad cells that get hot and put in a large pack to see what happens. Maybe I'll try to get some test together considering I have a huge case of bad cells.

Having done this, I can confirm that your initial hypothesis is confirmed, when charged at <500mAh cells didnt get hot, but some never got to 4.20v, they hovered at 4-4.1v and consumed 50-100mAh, warm but not hot.

So they were just burning off the energy as heat, thus reducing the pack's capacity, not actually making it catch fire. Interesting. I guess it depends on the charge per cell and if it can dissipate the amount of heat it's generating.
 
mike said:
So they were just burning off the energy as heat, thus reducing the pack's capacity, not actually making it catch fire. Interesting. I guess it depends on the charge per cell and if it can dissipate the amount of heat it's generating.

Yes indeed, and at 4v ive yet to see a cell discharge at more than 50odd mA, that equates to less than a quarter watt, not enough to get warm. The sanyo red devils were the exception, they were dissipating around 3 watts each being held at 4v, thats enough to get too warm.
 
3nergE said:
This is why i dont use fuses at all, after testing over 3000 second hand cells from over 500 laptop batteries, none of the bad ones were short circuit. Less than 50 cells got hot when charging (even cells that tested 0v) The only cells that got super hot (70c+) were about 8 of the red sanyo cells, and these got really hot @ 1000mAh charge, and barely warm at less than 500mAh, at the 100-200odd mAh in the power wall, they didnt do anything abnormal. BTW, the evil red sanyos were not <3v when initially recycled, but they did consume abnormal current without going much higher than nominal.
I had the same thoughts as you, INITIALLY ! but found that fuses clearly save me from memory I've had 5 Fuses blow on me with three cells going dead short. These cells WOULD have damaged my pack either by creating a huge amount of heat having 79 other cells dumpingevery available ampinto it trying to balance the cell, or just vented.
So not only will I use fuses, personally from my experience & what happened to Rohan I'll be double fusing every cell from now on. If Rohan's had of been double fused it would have been a non-issue.
 
With the thermal fuses is it visible that they have failed when they fail? My fuses even tho it takes a long time I can visually check every fuse in the pack with out taking it out. Does that little cylinder break showing you there is an issue?

I've also seen the resettable thermal fuses - same issue with them you don't know if there is an issue.
 
mike said:
I agree that the fuses probably won't do much in a 4p pack. I think most of us are building much larger packs though. I'm using 120p packs, many here are using 80p packs. If you charge the pack at 20A (which is common in my setup), that is 0.16A per cell. The fuses shouldn't get warm at all. Now let's say each cell can supply 2A of current, admitting that this is probably the extreme low end since most can do 5A+ very easily. If you assume 2A times 119 cells, that fuse on the bad cell is now going to carry 238A and will surely blow. The idea of fuses in my circuit is supposed to prevent a short cell, not a hot cell.

The more strings you add, the less the load per cell is. The goal in my setup is to get this as low as possible. If I have 4 strings of 120p packs charging at 20A, now that's 0.04A per cell. There is certainly no problem with fuses there :)

A good test would be to have some bad cells that get hot and put in a large pack to see what happens. Maybe I'll try to get some test together considering I have a huge case of bad cells.

While I agree with everyting you said, I still think it's possible for a cell to overheat (and potentially catch fire) even when charging at 4V 250mA (1A/4cells) - it has been tested by one person, I would like to see of we can reproduce this occurrence.

hbpowerwall said:
With the thermal fuses is it visible that they have failed when they fail? My fuses even tho it takes a long time I can visually check every fuse in the pack with out taking it out. Does that little cylinder break showing you there is an issue?

I've also seen the resettable thermal fuses - same issue with them you don't know if there is an issue.

This is a very valid question. I am not sure if there is visually. I will contact the manufacturer.
 
cstanley said:
While I agree with everyting you said, I still think it's possible for a cell to overheat (and potentially catch fire) even when charging at 4V 250mA (1A/4cells) - it has been tested by one person, I would like to see of we can reproduce this occurrence.

Its not about belief friend, its physics, originally you said 4p@1000mA, so the cell getting hot was likely 1 cell getting 1000mAh, thats about 3-4 watts, thats enough to get hot. Divide this by 4 and you get less than a watt, which just cant really get hot, thats why I showed my math earlier, which was also reflected in my testing.

While I agree fuses are a great way to increase safety, its also a way to decrease reliability / safety if you cascade the failure and all of a sudden cells are getting overcharged at an over-current rate / situation.

Keep in mind that most people fuse for draw rate, which is double the rate that most cells can be charged at.

Horrible things can go wrong if you overcurrent cells, even good ones

This is why I dont fuse.

kind regards
 
3nergE said:
cstanley said:
While I agree with everyting you said, I still think it's possible for a cell to overheat (and potentially catch fire) even when charging at 4V 250mA (1A/4cells) - it has been tested by one person, I would like to see of we can reproduce this occurrence.

Its not about belief friend, its physics, originally you said 4p@1000mA, so the cell getting hot was likely 1 cell getting 1000mAh, thats about 3-4 watts, thats enough to get hot. Divide this by 4 and you get less than a watt, which just cant really get hot, thats why I showed my math earlier, which was also reflected in my testing.

While I agree fuses are a great way to increase safety, its also a way to decrease reliability / safety if you cascade the failure and all of a sudden cells are getting overcharged at an over-current rate / situation.

Keep in mind that most people fuse for draw rate, which is double the rate that most cells can be charged at.

Horrible things can go wrong if you overcurrent cells, even good ones

This is why I dont fuse.

kind regards

I never said I didn't believe you?

All I said originally was 4P@1 amp. You keep saying 1 amp per cell and that's not right. 4P@1A would be evenly split between all cells. In this case it's 4. 1 Amp divided by 4 is 250mA.

That's what I am saying, now that we have this cleared up we can go to the original question.

If you read the last part of the opening post I pasted a quote from a person that claims to have been charging 4 cells in parallel giving 1Amp total (which is evenly split) so 250mA per cell - he said one of his cells got so hot that it was getting dangerous.

In a post I made after the opening post I stated that I wanted to see if we could reproduce this happening. Perhaps it was a damaged cell. I am unsure.

It's hard to convey what we mean over text, so my apologies if I am not conveying my message clearly.

It appears by what you are saying is that it is impossible for it to overheat at such a low amp draw? I was just asking the question because someone claimed they had it happen and I wanted to see if it was possible to reproduce. It sounds like the answer is no.
 
My concerns with these thermal fuses involve the re-settable variety and the non reset able metal cased ones.
The former creates a drain on the battery as the bad cell is reconnected briefly when the fuse cools down.
The Later is just a short circuit hazard, and impractical to use as a cell fuse due to it's size and shape.

I don't see a single cell disconnecting a problem when charging a powerbank with the usual hundreds of cells.
If it is, or using a pack with less cells, then you need to have a system to monitor the series voltages and disconnect the charger.
 
station240 said:
My concerns with these thermal fuses involve the re-settable variety and the non reset able metal cased ones.
The former creates a drain on the battery as the bad cell is reconnected briefly when the fuse cools down.
The Later is just a short circuit hazard, and impractical to use as a cell fuse due to it's size and shape.

I don't see a single cell disconnecting a problem when charging a powerbank with the usual hundreds of cells.
If it is, or using a pack with less cells, then you need to have a system to monitor the series voltages and disconnect the charger.

Can you please explain how these fuses are a short circuit hazard if you use one fuse per cell soldered to the bus bar? It would definitely be doable. Just want to understand how it's a hazard. Thanks!
 
View attachment 1

This style of thermal fuse is the one I'm concerned about.
The outer metal casing is connected to one wire, wouldn't take much to bridge the fuse.

Other styles with insulating (plastic ?) outer casing would be better in battery packs.

image_fhuigl.jpg
 
hbpowerwall said:
With the thermal fuses is it visible that they have failed when they fail? My fuses even tho it takes a long time I can visually check every fuse in the pack with out taking it out. Does that little cylinder break showing you there is an issue?

I've also seen the resettable thermal fuses - same issue with them you don't know if there is an issue.
I contacted them and it appears there is no visible way to tell if the fuse is blown. Bummer.

station240 said:
This style of thermal fuse is the one I'm concerned about.
The outer metal casing is connected to one wire, wouldn't take much to bridge the fuse.

Other styles with insulating (plastic ?) outer casing would be better in battery packs.
Ah yes, of course. In the event that the casing is metal I can see were problems could occur.
 
The price of those thermal fuses would prohibit their use on a per cell basis, for our application the correct way to use them is to trigger isolation from load and charge current, and trigger an alert.
 
I am dutch so pls forgive my bad writing......

Taking apart old computerbattery's I had now and then a battery with has 0.0 volt and even the ohm meter did nothing.
First I trow them away but now I was curious why I messure noting ad all.
So making the cell open ad the top I saw this was build up in layers, the bat has a little copper wire with stand up to a round flat plate and this plate is one with the top what you see wen you look ad the bat.
Messure on the little copper wire i found out the bat was 3,126 V. ????
But wen I first measure he was empthy 0.0 V
Looking good to the top I found out that the plate wil popup to the top so it wil disconnect from the copper wire, I suppose this wil happen wen the batt wil charge to fast or wil must deliver to much power and get to hot.

I had more batt with has 0.0 V and wen you take a little screwdriver and put it to 1 off the 4 holes then you can pusch on the little plate, you here a click wen it go down and then the batt wil function again.

So I suppuse that this is also a build in savety future from the battery ???

just my 2 cent.. :)
 
Gandyman, the device you are talking about is called the CID, or Current Interrupt Device one of a few internal safety features integrated into MOST 18650s. The CIDgoes open circuit when the pressure in the cell exceeds a certain value. As you have found out the CID is not self resetting. Another safety device integrated into most 18650s is the PTC or Positive Temperature Coefficient device which inhibits current flow when the cell gets too hot. The PTC is fortunately self resetting although itsactivation is can permanently increase that cells internal resistance.

When something goes wrong most of the time a cell can be saved from exploding by its own internal safety features alone. However, these features are not always reliable and catastrophic failures do still happen regularly. The additional protection being discussed here is still a must for large batteries.

The Tesla cells, or at least the ones used in their cars do away with both the PTC and CID to cut costs. Thus using both normal 18650's and Tesla style fuses would presumably provide more electrical protection to each single cell than a Tesla battery. However, Tesla batteries have additional physical safety features, such as active cooling and an "intumescent material" coating (you can read about this below). Rather than just being made fail safe through fuses these Tesla batteries are also designed to be safe when they fail. In the case of the cars, gases & flames from catastrophic battery venting are directed away from the exits of the vehicle. For a power wall it is important that even if the battery does vent it resides in a safe location to vent every cell in the pack without starting a larger fire.

I wouldn't take the risk re-using salvaged cells with reset CID's but.....I have seen quite a few bad cells with popped up CID's which when pushed down reveal an excellent cell. Does any one know to what degree the CID is manually resettable? Would this safety feature to work again reliably?


Cont... (Can't add URL links in my 'first post' on the forums)
Sources etc:
CID and common failure modes of unprotected cells: https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20100037250.pdf
Information on the PTC: http://www.nrel.gov/transportation/energystorage/pdfs/45388.pdf
Information on Tesla cells: https://teslamotorsclub.com/tmc/threads/amazing-core-tesla-battery-ip-18650-cell.17456/
Diagram of simple Tesla cell without PTC or CID: http://www.patentsencyclopedia.com/imgfull/20100136421_01
There seems to be some debate if Tesla power-wall cells have CID's and PTC's : https://arstechnica.com/civis/viewtopic.php?f=26&t=1176353&start=5280
 
Battery said:
Gandyman, the device you are talking about is called the CID, or Current Interrupt Device one of a few internal safety features integrated into MOST 18650s. The CIDgoes open circuit when the pressure in the cell exceeds a certain value. As you have found out the CID is not self resetting. Another safety device integrated into most 18650s is the PTC or Positive Temperature Coefficient device which inhibits current flow when the cell gets too hot. The PTC is fortunately self resetting although itsactivation is can permanently increase that cells internal resistance.

When something goes wrong most of the time a cell can be saved from exploding by its own internal safety features alone. However, these features are not always reliable and catastrophic failures do still happen regularly. The additional protection being discussed here is still a must for large batteries.

The Tesla cells, or at least the ones used in their cars do away with both the PTC and CID to cut costs. Thus using both normal 18650's and Tesla style fuses would presumably provide more electrical protection to each single cell than a Tesla battery. However, Tesla batteries have additional physical safety features, such as active cooling and an "intumescent material" coating (you can read about this below). Rather than just being made fail safe through fuses these Tesla batteries are also designed to be safe when they fail. In the case of the cars, gases & flames from catastrophic battery venting are directed away from the exits of the vehicle. For a power wall it is important that even if the battery does vent it resides in a safe location to vent every cell in the pack without starting a larger fire.

I wouldn't take the risk re-using salvaged cells with reset CID's but.....I have seen quite a few bad cells with popped up CID's which when pushed down reveal an excellent cell. Does any one know to what degree the CID is manually resettable? Would this safety feature to work again reliably?


Cont... (Can't add URL links in my 'first post' on the forums)
Sources etc:
CID and common failure modes of unprotected cells: https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20100037250.pdf
Information on the PTC: http://www.nrel.gov/transportation/energystorage/pdfs/45388.pdf
Information on Tesla cells: https://teslamotorsclub.com/tmc/threads/amazing-core-tesla-battery-ip-18650-cell.17456/
Diagram of simple Tesla cell without PTC or CID: http://www.patentsencyclopedia.com/imgfull/20100136421_01
There seems to be some debate if Tesla power-wall cells have CID's and PTC's : https://arstechnica.com/civis/viewtopic.php?f=26&t=1176353&start=5280



This is great information Battery thanks for explaining!

This helped me (visually) understand the concept:

image_ocoyld.jpg
 
Back
Top