Please help diagnose a near meltdown

[rick_p]

There's two parts to this issue, part one is a crash that damaged the hub motor cable, and part two is the discovery of a near melt down by the battery, which may have been caused by the crash, but might just be coincidence.

Part 1:
On my commute home a driver cut me off, which caused me to go down. Thankfully I wasn't going fast, but the bike went down on the left side and caused minor damage to the cable that goes into the rear wheel hub motor. I did a brief road side inspection and didn't see anything that looked overly concerning so I rode home, and the bike operated normally.

For whatever reason, the wheel didn't come with one of those plastic covers to protect where the cable enters the axle, so I had wrapped it up with electrical tape, which I didn't remove when I did the road side inspection because it actually looked like it helped protect the cable. When I got home I removed the tape and the thin spring thing meant to protect the cable. You could now see that the cable incurred a minor pinch (it was no longer perfectly round), and there was a small tear in the cable's sheathing. Here is a picture:



Part 2:
When I saw the tear in the sheathing, but didn't find any significant damage to the internal wires that I could see (yes, I see there is a tiny exposure on the small yellow wire, but it's the only one, the other scratches don''t expose any wires) I thought the best way to protect this would be to put on a layer or two of shrink wrap and then fabricate a protective bracket to prevent future damage.

So, to get that going I needed to free the other end of that cable from the controller, and I thought it best to remove the battery before I do anything. That is when I noticed some solder had come out of the shrink wrap covering a soldered connection next to the fuse. Oh no I thought!!!!

This is where I need your help. My first thought was there was a brief short circuit when the hub cable got pinched, but then the more I thought about it I realized, wait, this is where the battery power is going into the controller, if there was a short circuit at the hub cable, the melt down (arcing) would have occurred right there at the pinch location, and it would have melted the wire sheathing and it would be noticiable.

So what happened here then? Is this just coincidence and I'm darn lucky to have noticed it before a lot worse happened? Let's take a closer look...



I will never put a solder joint so close to a fuse again, the melted solder could have gotten inside the fuse holder and caused a bridge across the fuse!!!

A bridge did not occur, and clearly it got hot enough to melt solder, yet the fuse never blew. What the heck!!!

Possible clue - I recently changed the front chainring (46 teeth -> 52 teeth) so I could increase my cruising speed. With the smaller sprocket I was peddling too fast, so my top cruising speed was 20mph, with the new sprocket I was able to maintain 24mph and the pedal cadence was just right. Is it possible that the increased speed over long duration pulled too much amperage from the battery and that is what caused the wire to get so hot? Or, do you think the crash had something to do with it? Or, am I overlooking something else??

Thanks in advance for your thoughts on this.

 

[amberwolf]

There's two parts to this issue, part one is a crash that damaged the hub motor cable, and part two is the discovery of a near melt down by the battery, which may have been caused by the crash, but might just be coincidence.

I don't see any worrisome inter-wire damage in the cable. As long as there's no way for the wires to short to each other, then simply repairing the insulation damage and then recovering the unsheathed area to prevent water ingress and corrosion will be sufficient, as far as preventing controller damage from shorted/etc wires.

FWIW, the plastic covers some places provide don't really protect against that kind of damage very well, if at all. But there are derailer guards that bolt to fender mounts (if you have them) that stick out far enough to help prevent this kind of problem.
https://www.google.com/search?source=univ&tbm=isch&q=derailer+guards+that+bolt+to+fender+mounts
just make sure it's actually a tough one made of good hard steel, and that it doesn't require mounting via an axle (because it's hole will be too small for yours, and it may not go up high enough to protect your axle--if you have to drill out the hole to fit your axle it may not have enough strength in the mounting plate to then hold the guard out and protect the axle). If you can get one made in the 1980s it's likely going to be pretty tough; I had one off an old MTB found at goodwill that was very nice; the bike was pretty beat up and so was the paint on the guard, but it was not bent or deformed despite the definite impacts and abrasions that it had survived.

There are other solutions as well, but this one would work for a number of common crash situations.

Part 2:
So, to get that going I needed to free the other end of that cable from the controller, and I thought it best to remove the battery before I do anything. That is when I noticed some solder had come out of the shrink wrap covering a soldered connection next to the fuse. Oh no I thought!!!!

This is probably the very common problem of crappy fuse holders. The fuse connection can be relatively high resistance, so it then heats up during high current usage. This may be enough to reduce the springiness of the contacts and reduce contact pressure and increase resistance further, which may mean that even low current will cause high heat, and the fuse plastic may melt, as may the housing, and solder that is close enough to the terminals may also melt.
https://endless-sphere.com/forums/viewtopic.php?f=34&t=114700

A list of posts in various blade fuse meltdown discussions (some including solutions)
https://endless-sphere.com/forums/search.php?keywords=blade*+fus*+melt*&terms=all&author=&sc=1&sf=all&sr=posts&sk=t&sd=d&st=0&ch=-1&t=0&submit=Search


I highly recommend a bolt-on fuse holder, meaning one where the fuse bolts to the wires via ring terminals on the wires themselves. There are a number of types; I recommend looking at some of the "fuse" threads here on ES for recommendations. Some of them are in this list
https://endless-sphere.com/forums/search.php?keywords=fus*&terms=all&author=&sc=1&sf=titleonly&sr=topics&sk=t&sd=d&st=0&ch=300&t=0&submit=Search
as an example of what to look for, like this.
https://endless-sphere.com/forums/viewtopic.php?f=14&t=114290#p1689990


Note that a short at the hub's phase wires would almost certainly blow up the controller, and while that might cause the fuse to blow from the overcurrent shorted FETs might cause, it would also not allow the system to operate the motor and probably cause the wheel to be VERY hard to turn.

I will never put a solder joint so close to a fuse again, the melted solder could have gotten inside the fuse holder and caused a bridge across the fuse!!!

This is pretty unlikely, given the design of the fuseholder, even if there was a huge amount of solder on the interconnect. Even if it were to happen, if you have *any* connection that causes solder to melt, you have much worse problems to worry about, as heat that intense may start a fire in the wiring.

 

[E-HP]

For a 30A fuse, I'd use a fuse holder rated for at least 40A.

 

[rick_p]

Note that a short at the hub's phase wires would almost certainly blow up the controller, and while that might cause the fuse to blow from the overcurrent shorted FETs might cause, it would also not allow the system to operate the motor and probably cause the wheel to be VERY hard to turn.

Reading this made my day, it means that there wasn't a short at the hub's phase wires, because the controller is fine.

I don't see any worrisome inter-wire damage in the cable. As long as there's no way for the wires to short to each other, then simply repairing the insulation damage and then recovering the unsheathed area to prevent water ingress and corrosion will be sufficient, as far as preventing controller damage from shorted/etc wires.

FWIW, the plastic covers some places provide don't really protect against that kind of damage very well, if at all. But there are derailer guards that bolt to fender mounts (if you have them) that stick out far enough to help prevent this kind of problem.

I triple layered heat shrink over the damaged area. I'll fabricate a metal crash guard for true protection and post pictures when complete.

This is probably the very common problem of crappy fuse holders. The fuse connection can be relatively high resistance, so it then heats up during high current usage.

I highly recommend a bolt-on fuse holder, meaning one where the fuse bolts to the wires via ring terminals on the wires themselves.

If you have *any* connection that causes solder to melt, you have much worse problems to worry about, as heat that intense may start a fire in the wiring.

Reading this also made my day, it means my hypothesis was somewhat correct (increased speed over a long duration pulled more amperage than usual from the battery) and that combined with the insufficient fuse holder and wiring caused the wire to get so hot it melted solder.

This is only temporary, a day or two at most, I don't want to have any sub par wiring, connections, or fuse terminals, but it is 12 gauge wire and rated at 30 amps. I'll go easy on the throttle until I install the upgrade. I ordered the block type you mentioned where you bolt the wires to the block using ring terminals on opposite ends of the fuse. The 10 amp fuse in the picture is just for a test purposes. I'll update the thread after I install upgraded fuse holder.

 

[rick_p]

For a 30A fuse, I'd use a fuse holder rated for at least 40A.

The one I ordered says it's rated for 60 amps, but says 36 volts, will that be OK on my 48 volt setup? My assumption is the small increase in volts is not going to be a problem, as long as the amps are well covered, is that correct?

 

[E-HP]

The one I ordered says it's rated for 60 amps, but says 36 volts, will that be OK on my 48 volt setup? My assumption is the small increase in volts is not going to be a problem, as long as the amps are well covered, is that correct?

Should be fine.

 

[LewTwo]

The one I ordered says it's rated for 60 amps, but says 36 volts, will that be OK on my 48 volt setup? My assumption is the small increase in volts is not going to be a problem, as long as the amps are well covered, is that correct?

Good question. I go by the wattage (which is that actual power passing through the fuse).
60 Amps x 36 volts = 2160 watts
2160 watts / 48 volts = 45 Amps
I am sure someone will correct me if there is madness to my logic :wink:

 

[amberwolf]

The one I ordered says it's rated for 60 amps, but says 36 volts, will that be OK on my 48 volt setup? My assumption is the small increase in volts is not going to be a problem, as long as the amps are well covered, is that correct?

As a warning to anyone in the future ever reading this thread:

If anything that is intended to break a circuit is not made in a way to block the voltage that will be across the resulting gap, current may continue to arc across the gap (exactly like an arc welder, with the same effects), and not only will the circuit not be broken in those cases, intense heating at that gap will occur, including risk of fire, depending on surrounding materials.

There is a good chance that within some tolerance above the rated voltage of the part, it will still break the arc every time. But there is no guarantee of that--it is only guaranteed to do so at/below it's rating, and how much above the rating the part may still break the circuit is unknown without (destructive) testing (and manufacturing variations may change that point with different part batches).

FWIW, the difference in voltage between your 48v system (which will actually be around 54v full) and the 36v rating might seem small in number of volts (54-36=18), but in percentage of difference it is quite large, being half of it's actual entire rating. It might be fine...but it might not. Testing under your conditions would be required to find out what happens when the fuse blows under sustained current draw conditions (which is where the danger typically lies).

If the current flow stops or drops below some critical point, the arc would be extinguised from that, but if the current doesnt' stop being drawn thru the circuit at the sufficient rate, the arc will continue until either it has eaten away enough material to increase the gap size beyond that the voltage will push across, or burned away the holder, etc, for the parts creating the gap so the same thing happens, or enough material is melted by the arc and deposited across it to bridge the circuit with a lower resistance than the arc itself creates. (or any resulting fire engulfs enough of the circuit to prevent further current flow :lol: )

I use a "24v" system's circuit breaker on my SB Cruiser trike's main 14s (~58v full) battery-to-controller line, as well as 32V bolt-inline fuse on the battery output terminal, simply because I had those and did not then have money to spend on correct ones. But I don't really expect them to break the circuit in the event of catastrophic overcurrent...I just hope they do.

(the breaker is actually there more as an extra "internal" switch, in addition to the external removable-handle battery-shutoff switch I use every time I turn the trike on or off...but it *is* still a breaker).


Regarding current flow thru the fuse before it blows, that depends on that specific fuse brand and model and part number's specific design, and should be specified on it's rating sheet from the fuse manufacturer. If there is nothing like that available for it, then you'd have to test it yourself under your specific usage conditions to find out what it blows at, for different current amounts vs length of time that current is sustained. An example was actually recently posted here by Mxlemming:
https://endless-sphere.com/forums/viewtopic.php?f=6&t=111159&hilit=fus%2A&start=250#p1724611
Each color curve is for that specific fuse rating size. Time before blowing is on the left, and current before blowing is on the bottom.
I've attached the diagram from that post below:

 

[LewTwo]

....
FWIW, the difference in voltage between your 48v system (which will actually be around 54v full) and the 36v rating might seem small in number of volts (54-36=18), but in percentage of difference it is quite large, being half of it's actual entire rating.
....

Off course one might suppose that anything rated for 36 volts actually means three 12 volt lead acid batteries in series which is about 13.5 (float voltage) per battery or 40.5 volts (vs 13 x 4.1 = 53.3 volts for Li-Cobalt).

 

[amberwolf]

Except that fuses, etc, are not rated based on the average voltage of battery systems, they are rated for specific voltages and currents. You'd need to check the manufacturer's datasheets for exactly what they rate them for, but the number on it should be the actual voltage limit itself.

If it was a fuse developed by an "ebike company" specifically for such applications, then perhaps they would rate them that way...but nothing I've used in various electronics repairs or applications over the years has been rated for anything other than it's actual true limits.

(FWIW, if electrical / electronic components were generally based on battery average voltages, then 100v caps and FETs would actually be designed to handle say, 118v or so...but they aren't (although they may tolerate it for some time depending on other conditions, they aren't guaranteed to do so and they may fail at any point above that max rating).

Here's one manufacturer's documentation regarding various parameters, if it's useful:
https://www.eaton.com/content/dam/eaton/products/electrical-circuit-protection/fuses/protective-device-ratings/bus-ele-voltage-rating.pdf
View attachment bus-ele-voltage-rating[1].pdf
One critical statement they make regarding voltage ratings:

The voltage rating of a fuse must be at
least equal to or greater than the circuit voltage. It can be higher but never
lower.

While the document mainly explains AC systems, this applies to DC, and should apply even more strictly since current will never drop to zero on a regular basis as it would in an AC system.

Other manufacturers have similar documentation available, some of it specific to specific fuse types and part lines.

....
Off course one might suppose that anything rated for 36 volts actually means three 12 volt lead acid batteries in series which is about 13.5 (float voltage) per battery or 40.5 volts (vs 13 x 4.1 = 53.3 volts for Li-Cobalt).

 

[LewTwo]

Except that fuses, etc, are not rated based on the average voltage of battery systems, they are rated for specific voltages and currents. You'd need to check the manufacturer's datasheets for exactly what they rate them for, but the number on it should be the actual voltage limit itself.

Problem being that we are dealing with a lot of pieces-parts produced in China. Most of those manufacturers do not have data sheets. The ones that do exist are likely in Chinglish and may not be related to the actual product being delivered

 

[E-HP]

To my layman's eyes, it's a standard green 30A automotive blade fuse, so 30A @ 32V.

 

[amberwolf]

Problem being that we are dealing with a lot of pieces-parts produced in China. Most of those manufacturers do not have data sheets. The ones that do exist are likely in Chinglish and may not be related to the actual product being delivered

Then the only safe thing to do is either not use those parts (if they are safety-critical and you need the data sheet to be sure how they will operate to ensure the safety factor required), or test them yourself to ensure they will perform as required in the application in question.


Short of that, for fuses, breakers, etc., using them well below whatever voltage rating they claim *might* be fine, but to know what current they will actually blow at, to ensure a dead short in a high current application won't cause a wiring fire, etc., would still require buying more fuses than you will need and sacrificing at least one to such a test.

Otherwise it's just hoping for the best and not planning for the worst.


For non-safety-critical parts, it doesnt' matter that much...but for anything with the specific purpose of protecting against fire, etc., like fuses and breakers, it's pretty important to be able to depend on their specific ratings being correct in order to choose the right one for an application.


For myself, if I were building something for someone else, and I suspected a safety-critical part might not be what it claims to be, or didn't have a data sheet or specifications I could rely on, I wouldn't get it in the first place--I'd pick a different manufacturer (not just a different seller of the same thing, since the manufacturer would still have the info in the case of a seller that didn't know what they had, and if the manufacturer doesn't have reliable info on their parts then no seller possibly could, short of one that did complete testing and characterization of everythign they sell...and I've never heard of one of those.)

If I was building it for myself and it was a fire-safety-critical part, the same thing applies. (even if I break my "rules" by using parts I *know* are not really capable of protecting something, and just hoping they do anyway... :/ )

 

[rick_p]

Well, first and foremost, I thank everyone for their posts. I’ve learned a whole lot about fuses, fuse holders and the importance of using properly rated ones. If for no other reason than it being better (safer) to be equal to or over the voltage rating, I will find one that is.

What I find interesting is, some in-line blade fuse holders are rated for 600 volts but only 15 amps, while others are rated only 32 volts but at 30 amps.

I plan to return the one I ordered and find one rated for at least 48 volts and 30amps or greater, and install a 30 amp fuse in it.

 

[rick_p]

Of all the people on this thread, I am the least knowledgeable. That is both a disclosure and a warning to anyone who is reading this for the purpose of deciding on what kind of fuse and fuse holder to use for their battery. I strongly advise that you read what others have written in this thread, and use the links they provided for additional information.

From the research I've done, some if not most fuse holders are not rated for voltage, but they are ALL rated for amperage. I think the reason for that is, the fuses themselves ARE rated for voltage, and when a fuse blows, by design it opens a gap large enough so that an arc (which burns at a very high temperature) can't form between the fuse terminals, so using the correct fuse for your application is critical, because if you don't you are setting yourself up for an electrical fire.

With that said, a cheap fuse holder, even one rated for the correct amperage with the "correct" fuse in it can still overheat. That is precisely what happened to me, and I'll quote @amberwolf here:

This is probably the very common problem of crappy fuse holders. The fuse connection can be relatively high resistance, so it then heats up during high current usage. This may be enough to reduce the springiness of the contacts and reduce contact pressure and increase resistance further, which may mean that even low current will cause high heat, and the fuse plastic may melt, as may the housing, and solder that is close enough to the terminals may also melt.

So, if you take nothing else away from the information provided in this thread, use a fuse holder designed to grip the fuse much tighter than a slip connector can, and use a fuse properly rated for your application.

I haven't been able to find what looks like a good fuse holder for blade type fuses, so I'll be switching to either an ANL type or something similar that is designed for use in high powered audio applications in autos.

 

[rick_p]

https://endless-sphere.com/forums/search.php?keywords=fus*&terms=all&author=&sc=1&sf=titleonly&sr=topics&sk=t&sd=d&st=0&ch=300&t=0&submit=Search

https://endless-sphere.com/forums/viewtopic.php?f=14&t=114290#p1689990

I am the type of person who does more research than is necessary, especially when I ask for advice on a forum. I read the threads linked above and then went searching for products. I was not able to find an ANL or MIDI fuse holder that is rated above 32 volts, or a 30 amp ANL or MIDI fuse that is rated above 32 volts.

If anyone knows where to buy an ANL or MIDI (preferably MIDI) fuse holder rated at 48 volts or higher, please share that with me.

At this point I'm seriously considering bolting a 30 amp MIDI fuse inline without using a holder at all, because a single fuse holder only seems to serve one purpose as far as I can see, to hold the wires in place whether there is a fuse in it or not, but I can accomplish that in a safe way without the holder.

 

[E-HP]

So the discussion on arcing is interesting, so as a sanity check it's easy to find some tools to calculate the min and max distances to maintain that can produce an arc, at a specified DC voltage, in air. So for 3000 VDC, the max arc distance is 3mm; for 400 VDC, it's 0.03mm, for example.

Just eyeballing it, but looking at the gap of a blown blade fuse, while "rated" for 32V or 42V, would present no potential for a sustained arc, unless you're planning on running over 1000 VDC or higher.

Fuse voltages only come into play after they blow, otherwise are pretty arbitrary. Your issue was a crappy holder, not fuse, and the voltage rating carries no real world weight in selecting a decent quality holder.

https://www.cirris.com/learning-center/calculators/50-high-voltage-arc-gap-calculator

 

[thundercamel]

My 35 amp ATC fuse did the same thing with a 32 amp controller. I switched to a low profile J-Case fuseholder when space matters, or a Maxi fuse when space does not matter.

 

[E-HP]

Seems like there's a bottleneck from all 30A having to flow through these thin pieces of copper of the connector, so they probably get pretty toasty. It's odd that it's open on both sides; seems like the one on the left should be closed to increase the area for the current to flow across.

 

[rick_p]

Fuse voltages only come into play after they blow, otherwise are pretty arbitrary. Your issue was a crappy holder, not fuse, and the voltage rating carries no real world weight in selecting a decent quality holder.

This is consistent with what I read, and the holders I was finding were barely what I would consider decent, which is why I decided to make one out of High Temp PVC electrical conduit.

My 35 amp ATC fuse did the same thing with a 32 amp controller. I switched to a low profile J-Case fuseholder when space matters, or a Maxi fuse when space does not matter.

Those are good choices, but I decided to go with a MIDI fuse so I could bolt the fuse to ring connectors. I wasn't impressed with the holders I was finding online or in stores, so I decided to make one.

The current location of the fuse is inline near the battery, and there's no room to move it to inside the battery case, so I still need to fabricate a cover for it until I rewire things and move it inside the controller box where it can be mounted.

 

[rick_p]

I fabricated a cheesy but safe cover and rode the bike to work today, which is 30 miles round trip. In between lights I maintained speeds as fast or faster than when I melted my fuse holder. I checked (touched with bare hands) the wires very close to the new fuse holder several times while in route and it never even got warm. I think we’re good.