In line fuse holder melting

[slowwilly]

12VDC circuit. In line blade style fuse holder. Fuse/holder connection is very tight (difficult to remove fuse). 20A fuse.

When examining this system several weeks ago, I noticed that the fuse/holder showed signs of overheating (some plastic & rubber melting) with the link in the fuse intact. I replaced it with a new holder & fuse.

I got a call from the customer. The fuse holder and fuse had burned up and melted. It melted at the fuse holder, not at the wire connections. I can replace it with another, but don't think I will get a different result.

This system is dual power. It runs off 12VDC truck battery/alternator or (via relays) a Meanwell S-320-12 voltage converter (when plugged in). The load is a fan motor.

Any thoughts on what might be causing this & what to do?

My understanding is that typically when a fuse holder melts, but doesn't blow the fuse, it would be because of loose connections. But due to the tight fit & the fact that we've used thousands of these holders in other applications without the converter, I'm thinking it might have something to do with the voltage converter.

 

[OperaHouse]

I have seen this problem in pulsed applications like microwave ovens where instantaneous current greatly exceeds the average current. Pulsed currents are hard on any electrical transition. How high a current is this fuse holder actually rated for? I suspect the problem is the wire crimp in the holder. This is typically a point of high resistance. For example, operate a vacuum cleaner and the plug will be far hotter than the cord.

 

[slowwilly]

OperaHouse

Per the cataloge, the holder can be used with 30A fuses. I'm using a 20A fuse on a ~12A circuit.

I have noticed that the holders get almost to hot to touch whenever they're being used in a "high" current draw circuit. (We use them with fuses from 5A-20A, they only get that hot with 20A fused circuits).

 

[slowwilly]

Forest through the treees:

I forgot to tell you that the load (DC fan motor) has failed several times. In fact, it seems that we can't keep it alive more than 2 months.

Several other systems have been installed for a while longer with other customers and neither failed a load or burned up a fuse.

I installed the fuse because the wires leading to the load had been overheating (even 10GA THHN 105C) and "burning up".

My thought was that the fan motor starts to fail and draws excessive current (12A nominal). The fuse would blow before the wires get damaged.

This was a new fan, so it shouldn't have been going bad.

There has been no way of knowing if the overheating is occuring during AC (plug in through converter) or DC (truck battery/alternator) because both sets of wires are overheated when the customer contacts me. The customer believes it is the power converter, but has been unable to tell me why they feel this way.

Finally, I can get a power converter with PFC function. Would that be better for a DC inductive load?

 

[Comcokid]

A few loose thought while reading your post.

1) I just want to confirm that when the Meanwell supply is supplying power it is running just the fan - the Meanwell is isolated from the rest of the 12V circuit of the vehicle and is not trying to charge the battery, run the headlights, etc.

2) When the 12V system of the vehicle is supplying power, the Meanwell supply is isolated - the 12V of the vehicle is not trying to supply voltage into the output of the unpowered (or powered) Meanwell.

3) Some switching supplies are really designed for a mostly resistive load, but the fan is quite inductive - the Meanwell supply might possibly have problems. With the high starting current of the fan, the supply might limit at a low output voltage at high current depending upon it's method of handling over current. A large capacitor (1000's of uF at >25V) on the output of the supply will make the load appear less inductive.

4) PFC correction applies only to the AC input of the supply and how reactive the supply is to the power company. Shouldn't affect your problem.

 

[OperaHouse]

"I installed the fuse because the wires leading to the load had been overheating (even 10GA THHN 105C) and "burning up".

This sounds more like a customer or site issue. Could something jam the fan or possibly a sharp edge cutting into the wiring causing a short? Sounds like a mobile application. One other possibility Is an overvoltage condition caused by an alternator. Do the plus and minus of this circuit connect directly to the battery and not to just ground or 12V power line.

 

[itsmoked]

slowwilly; I think your fan is drawing too much current for your wiring.

Remember some fuses are rated at say 10A. This means it can draw 10A forever or 12A for and hour or 15A for 10 minutes, etc. So if you have a cyclical load you could be running a a lot of amps thru the system every now and then that adds up to a lot of heat.

You could have no airflow over your wiring... It will heat up.

Fuse holders can heat up a lot if they have crummy connections. If the fuse is hard to get out it might not be that the connections are tight but some other part is tight.

I would use a temp gun on the whole system. It will allow at a glance where the big heat is coming from.

http://www.infrared-usa.com/Product.aspx?ProductID=143&CategoryID=4158

I bet the fan blades are wrong. Someone changed the pitch or diameter or part number, not realizing that this can greatly increase the hp requirements.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[TurboXS]

I had a similiar experience when I ran a set of thermo fans in a car. Basically the high current flow of the fans plus some small but significant resistance in the fuseholder caused it to overheat and melt.

The main question is how much current is being drawn by the load. If you have say 12A and lets say that the fuse holder has a resistance of say 5 milliohms per contact

http://www.circuitprotection.ca/fuseblocks1.php?category=INLINE FUSE HOLDERS&size=Automotive Medium

and the fuse will have a resistance of say 3 milliohms

http://www.littlefuse.com/cgi-bin/r.cgi/en/prod_series.html?SeriesID=148&LFSESSION=IjngMWGf4G

Therefore the power dissipated in the fuse holder and fuse under good conditions is 1.9W. Enough to feel quite warm or hot.

Now if the current is higher or the fuse contacts are higher in resistance then the power and fuse holder temperature can increase dramatically. A couple more amps and another 10 miiliohms of contact/fuse resistance (don't forget that fuse resistance changes with temperature) will increase the power to 4.5W. If it is mounted in a hot area then this won't help either.

Measure the voltage drop across the fuse holder and wiring, plus a current measurement will help you analysis the situation.

 

[itsmoked]

TurboXs makes a great point!
The power dissipated is:

P = I² x R

Hence a small increase in I is squared or twice the I generates 4 times the power/heating.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[FrankTech]

Dear Sir

First remember allways the when you load a fan to a DC source you will have a pulse shaped current. Not a DC current. The FAN manufacturer will allways give the average current. Not the maximum. Depending on the fan the max current could be much higher indeed. Since the average current is 10 Amps the Power supply will still try to supply the power. The Meanwell S320-12 is equipped with a overload protection circuit. This circuit will switch the voltage low, when the output current rates 105-135% of Imax.
I max is for this type 25 Amps. So in worst conditions the PSU could supply 105%-135% of 25Amps before it switches down. So, to be sure your wiring should be rated for 135%
Which is 33.75 Amps.!! , wiring it for 10 Amps only could indeed lead to burning of wires.
In fact adding an extra fuse to the circuit would be OK but would not be neccesary for protection of the PSU. Because of the pulse shaped current, drawn by the FAN, the fuse will be heating up.

Hope this was of any help,

 

[OperaHouse]

#10 wire can carry 50A in an 86 degree environment. If insulation is melting at other than a connection, you are drawing well over 35A for some period.

 

[MacGyverS2000]

They sell datalogging current meters for quite cheap (Google it)... conside throwing one of those inline with the fuse and see peak as well as average currents.


Dan - Owner

http://www.Hi-TecDesigns.com

 

[slowwilly]

Comcokid:
1&2: It is isolated through relays, when the ignition is moved to on, the "AC" relay is opens and the "DC" relay closes.
3&4: I originally tried a lower output power supply, but could not get the fan through startup (the power supply went into hiccup overload protection mode). I up-sized the power supply and contacted MeanWell. They suggested a SP-320 rather than an S-320. Something about a constant current limiting overload protection when supplying a capacitve load. In looking at the specs (through my ME eyes) the only difference I noted was the SP has a built in active PFC function.
--------

itsmoked:
Almost certianly there is no (or little) air flowing over the wires. They're contained within a small NEMA 4 enclosure along with the relays (power supply is seperate).

According to "touch test", the heat inside the box is coming from the relays and the base of the fuse holder (the fuse itself is warm, but not as hot as the base). As far as an IR gun, our last one got stolen & management is not willing to buy another one. I could thermocouple the components, but not sure how helpful that would be.

And the fan blades are correct. I replaced the entire fan assembly myself to be certian.
---------

TurboXS/itsmoked:
4.5W does not seem like all that much heat to me (of course, I'm used to dealing in more in 10KW increments), do you think it was enough to melt the holder in a high heat environment (say the air temperature inside the box was 120F)?

In any case, when I return from vacation, I can measure the current and voltage drops. I feel reasonably good about the precision of my voltage measurements but not so sure about the precision of my current beyond .25 amps or so (less expensive clamp on ammeter). I'm not sure what to do with the voltage drop measurement. I'll post them when I return.
---------

FrankTech:
I need to learn, not remember With a pulse shaped current, I assume that my DC clamp on ammeter will only give me the average or RMS reading also (it's at home so I don't know which model it is, I think it's made by Extech).

I wired everything 10GA except for the fuse holder itsself, which is 14GA wire attached to the holder. My original thought (after failing the 4th fan or so) was that when the fan started going bad, it's current would go up and the fuse would blow. This would give us a slightly advanced notice to get a fan on order & the customer schedule some downtime when the fan could be replaced. (ie: use a 20A fuse, when it blows order a new fan and instal a 25A fuse until it gets here).
----

All:
Thank you for the help. This thing has been a thorn in my side for over a year now.

 

[slowwilly]

Operahouse:
You might have caught in my last post (just a couple of minutes ago) that the holder itself was 14ga wire and I'd estimate the local air temps to be around 120F inside the enclosure.

-------
macgyvers2000

One thing I've noticed is that there isn't much out there geared towards measureing DC currents. I'll check into it though (right after I sign off here).

----
Thanks again.

 

[FrankTech]

The advice you were given from the meanwell representative was not (exactly) correct.

The SP320-xx is a newer version of the S-320-xx
This is equipped with PFC ( power factor correction ) but has indeed the same overload cirquit ( hiccup mode )

The good advise was that if you have a capacitive/or inductive load you could better use a power supply that has a overload cirquit based on a constant current output.
Could be a SP-200-12 ( 16.7 A) or SP-480-12 ( 43 A !)

Sorry for the "remember" part, sometimes translating it to a foreign language while thinking in another language seems to modify meaning...

Best regards from Holland
If you need info on power supplies, and.or converters
please let me know.

Best regards

 

[slowwilly]

FrankTech:

Your English is quite good, much better than my Dutch! I was mostly making fun of myself, because it's been so long since I thought of wave forms that I'm not sure if I need to remember or learn.

So, the SP-320-xx does not have constant current output but the SP-200-xx and SP-480-xx do? No wonder the guy from MeanWell was confused, what sense does that make???

Anyways, the 200 doesn't seem like it will satisfy my power requirements. So that means the upgrade would be to a 480? Do you think that will help my situation?

 

[FrankTech]

Well, it will not improve your situation I think...
Switching to a bigger power supply will mean only that the power supply will be able to give more power, meaning you have to make the wiring also capable of handling the max.current.

The Sp-200 and SP-480 are later (newer) models.
Most newer models are using this constant current foldback system. Older types stil use hiccup or foldback current limiting.

Did you ever measure the real current of the fan ?
Is it allways running ? 24/24 ?

If you check the LED on the SP-320-12 , does it remain ON ALL the time or does the LED switches OFF and ON now an then? When it switches ON and OFF all the time the power supply is indeed a wrong type and getting a heavier one would solve hiccup problems
If the LED remains on, the power supply is OK.
So the problem is not power supply related.
If you need a contact person in USA that knows the Meanwell line please let me know. I know apeople in the USA that is in the power supply business for a long time now.

 

[TurboXS]

While 4.5W doesn't sound like much I can assure you that you wouldn't want to grab a standard 5W resistor running at 4.5W!

It all depends on how quickly the part can dissipate the heat.

I could find any real data on the thermal resistance of fuseholders however one source indicated that it may be value of between 3 to 50 deg C/W.

In any case let's use the average figure of 25 deg C/W. As a comparison a TO-220 package in free air has a thermal resistance of 50 to 60 deg C/W. 4.5W would mean that the temperature of the fuseholder would be 112 deg C ABOVE the ambient temperature. You suggested that the ambient temperature may be 48 deg C (120 deg F) which puts it at 160 deg C (320 deg F).

 

[TurboXS]

Knowing the voltage drop across each element, fuseholder, wiring, connections and the current allows you to calculate the power dissipated in each part.

 

[slowwilly]

FrankTech:
The real current of a new fan is ~12A as applied.

It does not run 24hours.
I'd say an average working day will be 12 hours spent on battery/alternator power (with intermittent off cycles, totaling about 3 hours of the 12) and 12 hours off the power supply (again with about 3 hours intermittently off). Weekends will be 24 hours using the power supply with approximalty 8 total hours off intermittently.

Diode is now always on (it was cutting out with the old/smaller power supply).
----------

TurboXS:
You're starting to make me wish I hadn't fallen asleap during my EE course Should my power disipations be calcuated using the measured currents or resistance of the parts? I'm thinking that I would trust my resistance reading more, so I'm leaning that way.

-------
Thanks to all again.