I have a set of fuses I need to test the characteristics to see if they will be implemented in one of my design boards that routes power. The fuses need to be rated for 35A, but need to support in-rush current. They will be used on a 24V rail powersupply. I'm slightly nervous about how to set up the experiment, because according to my calculations, I'd need a 0.6Ω load capable of disappating over 900W .

If you had to test a fuse to ensure correct operation, how would you set up a load capable of supporting the high power levels?

 

You might want to look into the matter in more depth. There is more to the fuse rating than the current limit - http://www.circuitprotection.ca/fuseology.html

 

If I had to do this, I'd use an Allen-Bradley pot that I have, as it's rated to 1.6 kW continuous and it adjusts from 0.18 Ω to about 15 Ω. The inductance is less than a μH at all resistances. So such beasts exist...

If I didn't have that pot, then I'd consider making my own load. Since I'd want to do it on the cheap (I'm a home hobbyist), I'd use some Harbor Freight 0.041" diameter stainless steel safety wire that is about 0.1 Ω per 11 cm (it's about $6 per pound). If you want to spend more money, Nichrome or Chromel might be good choices (or Monel if you can find it). I'd find a 1 to 2 cm diameter insulating tubing (ceramic or Pyrex tube preferred) and wind the needed length of wire on the tubing, using an Ayrton-Perry or bifilar winding to minimize inductance. Then I'd immerse the thing in a tub of distilled water to improve heat transfer.

If you have an R&D budget, you could instead use some ceramic fish beads to insulate things. They work well and give the ability to make reasonable-radii bends. I'd still immerse the thing in water though.

 

Sound like a good idea someonesdad, but I would exchange the water with mineral oil. I know you said distilled, but unless you steam cleaned the load before placing it into the water, enough impurities would exist to cause a problem. And if someone was to read this and say "well I dont have distilled water, so tap water it is!" .. Could lead to trouble. Especially at such amperages.

You can get mineral oil about the same price as distilled water these days.

The dollar store has 1 quart bottles for $1.

 

You might want to look into the matter in more depth. There is more to the fuse rating than the current limit - http://www.circuitprotection.ca/fuseology.html

GREAT SITE! Thanks for posting this, good info for me to absorb here.

If I had to do this, I'd use an Allen-Bradley pot that I have, as it's rated to 1.6 kW continuous and it adjusts from 0.18 Ω to about 15 Ω. The inductance is less than a μH at all resistances. So such beasts exist...

If I didn't have that pot, then I'd consider making my own load. Since I'd want to do it on the cheap (I'm a home hobbyist), I'd use some Harbor Freight 0.041" diameter stainless steel safety wire that is about 0.1 Ω per 11 cm (it's about $6 per pound). If you want to spend more money, Nichrome or Chromel might be good choices (or Monel if you can find it). I'd find a 1 to 2 cm diameter insulating tubing (ceramic or Pyrex tube preferred) and wind the needed length of wire on the tubing, using an Ayrton-Perry or bifilar winding to minimize inductance. Then I'd immerse the thing in a tub of distilled water to improve heat transfer.

If you have an R&D budget, you could instead use some ceramic fish beads to insulate things. They work well and give the ability to make reasonable-radii bends. I'd still immerse the thing in water though.

This is kind of around the lines I was thinking. I'm not TOO worried about heat generation, I just need to produce a load capable of withstanding the large power.

Any more ideas anyone has would be awesome too.

 

Sound like a good idea someonesdad, but I would exchange the water with mineral oil.

I purposely left off the suggestion of mineral oil because of the potential of local boiling and a fire. I doubt a kW of power is going to wreak too much havoc, but water is safer than mineral oil. retched, thanks for bringing up the issue of contamination -- the OP should know this cleaning is to keep the low conductivity of the distilled water.

Still, your comment made me go do a quick experiment in the kitchen... A bowl of hot tap water with a 5 or so shakes of our big kitchen salt shaker gave a DMM resistance about 4.5 orders of magnitude higher than the 0.6 Ω the OP wants. This is very gross contamination and would only change the load resistance by a fraction of a percent. A bowl of my normal tap water has a resistance in the 5 MΩ or so region and our water is very hard. I'd have to conclude that it's going to be hard to contaminate distilled water enough to mess up the experiment.

The OP would certainly want to measure the 4 wire resistance of his load before testing and, if it wasn't the 0.6 Ω he wanted, then he'd figure out where the problem was.

 

The DMM measurement of the water's resistance is probably invalid. I doubt it's at all linear.
Just like trying to measure the resistance of your tongue... it will read megaohms with a meter but attach a 9V battery and you get over a milliamp and with a 12V battery with a 5k resistor you get over a milliamp.

 

Agreed. I just wouldnt want a new entrant into electronics to think the water/electricity thing is ok.

 

You should know that fuse manufacturers have lots of engineers on their staff, and that you have a very valid question that needs answering.

Were I you, I would consult various fuse manufacturers for their input. After all, they have conducted a great many tests for a great many fuses under a huge variety of conditions that would be tedious at best for you to carry out, even if you DID have all of the equipment necessary.

My Dad worked for Square D Company, long before they were taken over by Schneider. In the old days, they had a lab in Kentucky where they would routinely subject equipment under test to 10,000 times their designed load limits to see what happened. They would call up the power company to warn them of the test ahead of time, because they knocked the power company offline more than once.

I suggest that you do not have the time, money, or resources to conduct such tests.

Instead, document as best you can EVERYTHING that might relate to your circuits' inrush current that is your cause for concern, and submit it to the staff engineers at a number of fuse manufacturers for evaluation.

That way you will obtain professional advice at little, if any, cost to you.

While this forum is helpful and informative, we simply do not have the resources to conduct such scientific tests for you.

 

Not only do the fuse makers know what they are on about, they publish datasheets!
The data will include a graph showing failure time against current for each specific fuse.

You don't say where in the world you are located?

There are standard classifications for fuse speed and inrush capabilities. The common European markings are:

aM = Motor rated, will withstand very high short term surges.
gL or gG = General Line, normal power and lighting circuits, will stand small surges.
uR or gR = Ultra-Rapid for semiconductor protection, zero surge capability.


For normal fuses, the load graph may give the 'pre-arc time' agains current, which is how long it takes before the element breaks. It will conduct power for some longer time, until the arc quenches - possibly the remainder of the AC cycle.

For higher specified and especially semiconductor protection fuses, The amout of 'let through' power, between an overload starting and the fuse stopping conducting, is given as I2T (Current squared * time).

Again, that will be in the datasheet for the specific fuse.

You also need to check the voltage rating of the fuse, and whether it is rated for AC or DC. High voltage DC ones tend to be much bigger and rather more expensive. The DC rating of a 'normal' fuse will be some fraction of it's AC rating.

If there is no datasheet for a specific fuse, then do not use it in a critical application!

Note that clear glass or open wire fuses, like automotive parts, should only be used on very low voltages and have massive let-through. They only exist to stop the wiring catching fire in case of a fault, and don't always manage that.

 

yes again there is no need to perform testing. The fuse manufacturer already has and published all the data you need in the datasheet for that specific fuse. You want to look at the time curves graphs

 

I agree with SgtWookie and rjenkins to check with fuse companies to see if they will share some of their engineering expertise. However, I also spent numerous decades in corporate R&D and found that some companies could be very closed-mouth unless it was clear you were going to spend a bunch of money with them (or already had). Or, you'd ask for information and not get the time of day until you left a message that you were with a giant corporation with a well-known 2 or 3 letter acronym. Clearly, you never know until you ask.

I still think a water-cooled resistor could be an appropriate experimental load per the OP's question. You don't even need the ceramic core if you don't want it. I did a quickie experiment yesterday and proved to myself that a 0.3 Ω load's DC resistance is unaffected by immersing in my house's tap water. I'd have a working load with a few minutes' worth of work. If you dump a lot of power into such a load, you'll have localized boiling and thermal effects that might make the load behavior not what you want (after all, that's how LLNL and Sandia do exploding wire experiments). That would have to be found out by experiment. I'd have to go hook the thing up to the welder to test it at 100 A or 200 A currents to find this out, but I'm not that interested in doing the work for the OP.

Actually, my scientific curiosity is aroused, so it's a project entered into the logbook. First, I want to finish the study of transformer stuff I'm doing. Then there's the pesky consulting work that has to be done to feed the family. Life would be so much simpler if I had a wealthy relative just give me a few hundred million dollars so I could pursue my interests at leisure (OK, make it a few billion and I'll change my interests). Alas, I don't have any wealthy relatives... Any of you wealthy people want to adopt into my family?

 

The reason I need to perform this experiment is simply some of the datasheets do not have characteristic curves, or simply the datasheets have no "data" lol. And I'm VERY restricted on space, so the fuses I've ordered samples of are already seldom used in the industry it seems.

I haven't tried pesking the companies to produce datasheets with valid information yet, but if I find I can't get a clear conclusion on this experiment I'll probably have no choice.

Thanks for all the help though, gave me some good ideas.

 

I would not, and dont, use fuses without datasheets for anything above a stereo.

If it is an important part that needs protection, get a WELL defined, and documented fuse. I would hold on to the new fuses until data becomes available. If the company does not want to produce a datasheet with the data I need, they dont need me as a customer.