Fuse failure

[ccernst]

I have a question about how fuses fail. I had a 3-phase power strip that uses fuses instead of breakers that had an issue. Fuses were SC-20 and load was less than 6 amps on each phase. The fuses did not blow completely, but manufacturer suggested I replace suspect fuses. Through measurements and observation, the affected fuses were off the "B" phase internal to the strip. Replacement fuses did fix the problem.

I did a continuity test on the failed fuses and they passed. I measured the resistance and it started at 3ohm and after about 10 seconds, it fell down to 0.0ohm. I checked another new fuse, same results. I've only experienced fuses failing by blowing the wire inside. My question is, do fuses fail in ways where they'd pass continuity tests?

 

[Dennis Alwon]

You need to get an electrician in there but generally a fuse either works or it doesn't. There is no sometimes. I have no idea what you mean when you say the fuse blew but not all the way

 

[petersonra]

I have a question about how fuses fail. I had a 3-phase power strip that uses fuses instead of breakers that had an issue. Fuses were SC-20 and load was less than 6 amps on each phase. The fuses did not blow completely, but manufacturer suggested I replace suspect fuses. Through measurements and observation, the affected fuses were off the "B" phase internal to the strip. Replacement fuses did fix the problem.

I did a continuity test on the failed fuses and they passed. I measured the resistance and it started at 3ohm and after about 10 seconds, it fell down to 0.0ohm. I checked another new fuse, same results. I've only experienced fuses failing by blowing the wire inside. My question is, do fuses fail in ways where they'd pass continuity tests?

did you check them in circuit or out?

If in circuit I would guess you are seeing a capacitor charging or discharging when you put the Ohmmeter in the circuit.

If not in circuit, I could not explain why this would be. I suspect operator error of some sort, but it is hard to know just what might be the explanation without seeing it first hand.

did you experience the same thing with the replacement fuses?

 

[ccernst]

You need to get an electrician in there but generally a fuse either works or it doesn't. There is no sometimes. I have no idea what you mean when you say the fuse blew but not all the way

The fuses that failed did not blow, since they pass a continuity test and eventually measure 0 ohms. But they were definitely causing problems. I think right now, I'm more curious to know how they un-commonly fail.

did you check them in circuit or out?

If in circuit I would guess you are seeing a capacitor charging or discharging when you put the Ohmmeter in the circuit.

If not in circuit, I could not explain why this would be. I suspect operator error of some sort, but it is hard to know just what might be the explanation without seeing it first hand.

did you experience the same thing with the replacement fuses?

I tested them when they were removed, out of the circuit. I got the same measurements with new fuses out of the box.

One other thing I did was measure the line voltage to ground from the affected outlets on the strip. The power strips have branch circuit groups of outlets (C19) for 208v. When I measured each voltage from one of the lines from A phase to ground, I measured a steady 122v. When I measured voltage from B phase to ground, it floated around 95 - 105 volts. After replacing the fuses for the branch circuits that use B phase, voltage came right back to 122 volt. The only change was the fuses.

 

[Speedskater]

While this failure probably doesn't fit this case.

One fuse failure mode is:

Failure to clear.
If you put a low voltage fuse in a high voltage circuit and it blows (opens) it may not open completely and still allow some current to flow.

 

[ELA]

Fuses are mostly thermal devices. They open when they get hot enough. Generally they remain open once open. A lot of us have probably seen clear glass fuses that are totally black inside and the element was totally vaporized. In other cases the element only has a vary small gap in it with no black residue.

An ohm meter puts a very small current through the fuse and results in no real heating.

I have not experienced what you have directly but in theory it could happen.
If you are curious you could try putting a variac to a load through that fuse. ( or other variable voltage source) Increase the output voltage while monitoring voltage at the load. It is then possible that the load voltage will follow the variac voltage to a certain point at which the fuse gets hot enough that it opens ( perhaps partially). IN this type of test you are not trying to overload and blow the fuse, just stress it. Kinda like how you use a megger rather than a ohm meter to test cables.

 

[Jraef]

Typically, cartridge fuses like SC (type G) are filled with silica sand. When the link melts, the sand fills the gap and provides the isolation of the fault current. But during the clearing process, there is an arc created at the designed-in weak spots of the fuse link. The arc can, on rare occasions, interact with the metal in the fuse link and the silica in the sand to create a composition called "fulgurite". It is for the most part non-conductive of any reasonable amount of CURRENT, but may on some occasions be able to be read as a conductive element when using a highly sensitive solid state meter. It's kind of like the fact that you can read a voltage from an SCR (called "leakage" voltage) with a Fluke meter but there is very little energy in it and if you used a meter with any kind of burden, it doesn't show. The formation of fulgurite can sometimes be an indicator of the fact that the initial cause of the fuse clearing may have involved a voltage spike in excess of the "Peak Arc Voltage" rating of the fuse. For example, a fuse rated for 300V will typically have a Peak Arc Voltage of something just over 700V. So if you have experience this, it's likely that you had some sort of voltage related event that damaged your equipment.

 

[K8MHZ]

Typically, cartridge fuses like SC (type G) are filled with silica sand. When the link melts, the sand fills the gap and provides the isolation of the fault current. But during the clearing process, there is an arc created at the designed-in weak spots of the fuse link. The arc can, on rare occasions, interact with the metal in the fuse link and the silica in the sand to create a composition called "fulgurite". It is for the most part non-conductive of any reasonable amount of CURRENT, but may on some occasions be able to be read as a conductive element when using a highly sensitive solid state meter. It's kind of like the fact that you can read a voltage from an SCR (called "leakage" voltage) with a Fluke meter but there is very little energy in it and if you used a meter with any kind of burden, it doesn't show. The formation of fulgurite can sometimes be an indicator of the fact that the initial cause of the fuse clearing may have involved a voltage spike in excess of the "Peak Arc Voltage" rating of the fuse. For example, a fuse rated for 300V will typically have a Peak Arc Voltage of something just over 700V. So if you have experience this, it's likely that you had some sort of voltage related event that damaged your equipment.

Fulgurite?

That's what we call sand that's been hit by lightning. I live near miles of sand dunes.

http://en.wikipedia.org/wiki/Fulgurite

When I worked in an aluminum foundry before I saw the light and became a real electrician, we would have fuses that would (shop terminology) 'hang'.

They were in discos mounted up high and were very hot. They would open and close intermittently. (I kid you not). Of course, this made troubleshooting tricky. A fuse that would not carry a load on a hot day would check out OK on a continuity meter, and also carry a load for a while on a cooler day. We were told destroy any fuse we felt was 'hanging'. Simple throwing in the trash wasn't sufficient.

At first, I thought my leg was being pulled when I was told about them by my boss. Fuses don't fix themselves. But, I did find several over the course of the two years I worked there and was really glad I was told about them.

I dissected one and didn't find any fused silica. I did find a crack where the element was soldered to the end cap.

Sometimes, a fuse would hang in a way that would not run a load, show no volts at the load and later close back up after the load was disconnected. They would carry enough current to knock you on your butt, just not enough to start the load.

I don't know why, but most (like 90 percent) of the fuses that 'hung' were run of the mill 30 amp fuses. I never used any from 'the drawer' even if they checked out OK. Any fuse that was in service at that foundry was suspect. I kept a stash of brand new 30 amp fuses to do my troubleshooting with.

 

[ccernst]

The fuse is rated for 600vac and I should only be sending 120v or so through. Though all affected fuses were from the B phase. No other power strips hanging off the same RPP experienced similar problems (yet).

I opened one up and caught all the sand that came out. The element inside seemed as though it was intact. I spun two pliers around and the element inside looked to be twisted apart. There were a few amber-colored "nuggets" in there compared to the off-white sand. I snapped a pic of the largest.

 

[Jraef]

The fuse is rated for 600vac and I should only be sending 120v or so through. Though all affected fuses were from the B phase. No other power strips hanging off the same RPP experienced similar problems (yet).

I opened one up and caught all the sand that came out. The element inside seemed as though it was intact. I spun two pliers around and the element inside looked to be twisted apart. There were a few amber-colored "nuggets" in there compared to the off-white sand. I snapped a pic of the largest.

Fuses don't just "sort of" fail. Fuses protect by melting. All they are is a strip of metal with built-in "flaws" that melt sooner than the rest of it at a specific amount of heat from a specific amount of current for a specific amount of time. If it was not melted, but it was not conducting, then the fuse element was not the problem. So if the element was intact, then the fuse was good.

Are you sure you don't have a problem with poor connectivity with the fuse holder? Sometimes you can get corrosion in the holder that can allow for inconsistent conduction, or if there is a spring force, repeated heating can weaken the holding force. Separation at the ferrule may also have been an issue, but in opening it with anything other than a band saw you may have destroyed the evidence of that.

 

[tom-g]

When you replaced the fuses you also cleaned teh contacts of teh fuse holder. A corroded contact could cause the symptoms that you described

 

[Besoeker]

Fuses don't just "sort of" fail. Fuses protect by melting. All they are is a strip of metal with built-in "flaws" that melt sooner than the rest of it at a specific amount of heat from a specific amount of current for a specific amount of time. If it was not melted, but it was not conducting, then the fuse element was not the problem. So if the element was intact, then the fuse was good.

Usually the case, I agree.
I did come across a problem where two fuses on a VSD input failed. Remedied the problem, replaced the two failed fuses only to discover that the third failed at the next attempt to run the VSD.
Replaced that, only to find that another fuse blew at the next attempt.
It was a while back, prior to 1991 at least, and a loss of a phase for the VSD in question resulted in overcurrent.

After a bit of head scratching, the long and the short was that when any one or two fuses blew, the other(s) were "weakened".
When all three were replaced, the VSD worked just fine.
We guessed at the time that the fuses that did not measure as failed suffered some melting of the element in the pre-arcing phase and that resulted in them having significantly reduced current carrying capacity.