Resistor Wattage - How Long Sustainable?

[leachdor]

Hi All,

I've got a project that uses a rectifier, and outputs to 5 locations.

I need to build a load cell for each output. The resistance value will be 13Ohms. Power dissipation through each load will be 32.5KW for roughly 15 seconds.

My question is, how would you go about selecting a resistor capable of this. I have checked around and it looks like 1000W resistors are about the max of what I can find. I assume I wouldn't need a bunch of 33kW resistors....but how do you know the minimum you can get away with without burning up the components for x amount of time?

Thx much

 

[dpc]

Have you tried a Google search for "power resistors" or "wirewound resistor"?

There are short time ratings, but it's going to depend on the resistor construction.

Ohmite and Post-Glover come to mind....

 

[sreid]

Power resistors are rated for short term overload but usually for one second. Large power resistors will have a lead time as they are made to order. I suggest you search on "loadbank." For example

http://www.avtron.com/load_banks.htm

Is your voltage 2 kV? If so you'll need to consider arcing, etc.

 

[dpc]

Grounding resistors are typically rated for 10 seconds, so I'm sure a 15 second rating can be quoted. You just need to discuss with the resistor manufacturer.

 

[itsmoked]

You need to use a test load at each separate place?

What is the point of this test? Is there no alternative? Are you just trying to test the rectifier function? That would take 17ms not 15 minutes.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[danw2]

You've got a 50A draw at 650V for 32.5Kw through 13 ohms.

What about a bunch of electric hot water heater elements wired up for your test?

Wire up a series string of three 4500 W heater elements, with each element drawing about 18.7 A across 240V (~12.8 ohm element), or in your case 16.9A at 217V drop.

Use 3 'strings' in parallel for 50.7 amps, give or take, depending on the actual resistance of the elements.

(I don't know what nominal voltage elements are spec'd at 240V or 230V or 220V?)

I do know that hot water heater elements are very inexpensive, so inexpensive that I can't recall what I paid two years ago for one.

Dan

 

[ScottyUK]

Check out the link for some examples of larger resistors. I know Cressall used to have short leadtimes unless you were looking for something very unusual.

http://www.cressall.com

The applications guys at Cressall are normally very helpful.


----------------------------------

I don't have an attitude problem. You have a perception problem...

 

[oldfieldguy]

danw's idea about using electric heater strips is a good one. I've used them in the past to make load banks for battery bank testing. The strips are easily obtainable, relatively inexpensive and of known quantity.

We put together a bunch of them in a cabinet with a shunt for current measurements, and with switches to enable us to switch strips in and out of the circuit to vary the load as needed.

old field guy

 

[Compositepro]

Light bulbs are about the least expensive and most convenient load resistors. They can dissipate energy continuously without overheating if you don't enclose them too much. You can get a 1000 watt halogen utility light for as little as $10. The main drawback is the relatively large resistance change from cold to hot filaments.
A heater from a hotwater heater must be submerged in water to keep from burning out. Turning electricity into heat is often the easy part. Getting rid of the heat makes things more complicated. If you generate heat in any resistor without providing cooling it will evetually become a light bulb!

 

[leachdor]

To answer a few questions...

We have 480 3phase going into the rectifier assembly, 640Vdc out per built in rectifier. Currently the rectifier assembly has 2 rectifiers built in, with the potential for 5 down the road. The test procedure hasn't been confirmed, but the most demanding possibility is a 50A draw on each of the 5 rectifiers simultaneously.

Heater elements are the way to go, cooling can be added to the enclosure holding them. (Fans nothing fancy) We currently use 400W 16ohm elements in our production lines. Not hot water, used for heating air in HVAC units. I was hoping we could manipulate those to handle the requirements above. You'll have to give me a bit of slack, I'm a computer engineer...my background is a bit sore with power calculations past what I've given to figure it out...
:-/

One of the "theory" questions I guess that would help me...I'm not sure how watts are dissipated by these resistances. Is it cumulative how much wattage the load could handle? (Two 400W resistors could continually handle say 600 or 800W vs. 400W regardless of configuration?)

 

[itsmoked]

A 400W resistor can handle 400W. You are required to hook them up such that you never feed one so that it draws more than 400W.. Otherwise series/parallel whatever.

Two 400W can dump 800W as long as you don't somehow constrain their ability to dissipate the heat.

I once soldered wires to (100) 100W light bulbs for a load bank [saving on the (100) sockets] I then hung them in bunches like grapes. Fired up the controller which ran about 8 seconds before the solder start melting and the unsoldered wires started springing loose. I lost a lot of semiconductor fuses before I could hit the power switch.

Took a large fan and some separation then they worked fine.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[tinfoil]

A heating element designed for 480V will only consume its rated W value of you put 470V across it.

You should connect these elements in parallel go get multiples of this output

So , for example, if you need 30kW (30000W) of power, and you have 5000W 480V elements, then you need 6 of them in parallel to make the 30kW.

If you need balanced 3phase, then connect 2 of these in parallel in each of your 3 phases.

CAVEAT: You are playing with some fairly substantial energy levels here. Ensure that you have appropriate measures in place to deal with the heat levels you will be making, and with the current levels you will be producing!

 

[tinfoil]

In previous post, 470V should actually be 480V.

(I which this forum allowed for edits after submission )

 

[MikeHalloran]

I which it, too.


Mike Halloran
Pembroke Pines, FL, USA

 

[davidbeach]

Actually, what you do is post your correction, such as what tinfoil did. Then you Red Flag your post with the error and ask to have the correction made and to have the following post, with the correction, deleted. I've done that occasionally and never had any trouble; now if someone were asking for multiple posts to be revised every day the gnomes that do all that behind the scenes might get a bit testy.

 

[Warpspeed]

If I was doing this I would not try to rate my resistors for continuous operation, that would not be cost effective.

I would be thinking more along the lies of a giant electric jug or electric kettle. These can pour maybe 2.5Kw into a couple of litres of water for considerably more than 15 seconds. The elements are fairly frail too.

So how about using some 240V 1.2Kw standard radiator bars submerged in sufficient water (or oil) to absorb the heat? These bars can dissipate 1.2Kw in air, at 480v they should easily dissipate four times that power if submerged in water.

Perhaps seven 240v 1.2Kw radiator bars in a couple of gallons of water should be well up to the job for a once only 15 second "shot". 4.8Kw per bar x7 = 33.6Kw

As they will certainly not be glowing red hot, the resistance will be lower, so maybe eight bars will be required? A little experimentation should get you fairly close to your requirements, and you can always strip off a few turns of resistance wire to lower the resistance.