Author Options:

How does this furnace work? Answered

I just bought a "custom crucible furnace" from a local university's surplus department and I have absolutely no idea how to get it working. It is roughly 4 feet in diameter and 4 1/2 feet tall, it has 6 'coils' of metal attached in series with thick ribbon-y cable at the top. The side has a 'band' that sticks out but as far as I can tell it is hollow.

From what I can tell it might be an arc furnace that something blew up in and damaged a bunch of the components. I really want to get it working as I've been saving aluminum cans for a long time and scrap metal so my plan now is to just hook up the cables to my arc welder and see what happens :) but if anyone knows anything about it or where I can go to get more info that would be much appreciated!



7 months ago

Can't get the reply button to work, so out of context here again...
If it works with your welder then the only question is for how long and how good.
Amp wise you might have found the sweet spot by accident for the configuration of the elements you used.
Meaning the welder might max out before the elements can produce a short - non linear, remember...
Best option would be to measure the actual voltage and amps going out of your welder.
One hour to get to a usable temp is not bad, around 30 min would be better but without control I would not risk it.

If you have an adjustable welder, like these cheap stick welders then try a single element with the current turned down to the lowest setting.
Check the temp and how the element reacts.
If, after one hour you still see no red increase the current slowly.
You should be able to determine how many elements the welder can handle before maxing out.
Lets say you get three elements nice red hot at max amps:
Try again with three more but keep the three ones parallel and the next three as well but add them in series to the first three.
This way they still get suffient amps and you might be able to run 6.
Go on like this until you might find a series parallel combo that can be powered safely with the welder alone.


8 months ago

This is a very high temperature unit, with Silicon carbide heaters - designed for longservice and high temperature operation. Unlike metal heaters, they won't fail after a time.
They are unfortunately evil buggers to run, because they are non-linear resistors, and have a very low cold resistance which increases rapidly with temperature.

I'll guess at the very least it was designed for three phase power, because that's what you do with them, and you switch the elements around from series to parallel as things heat up.

KANTHAL are one of the big suppliers of these elements, and they have some great books on how to control and use them


Reply 8 months ago

I already suspected that they are not just common heating elements but this would really complicate things.
Not really easy to run these beasts on a hobby level I am afraid...


Reply 7 months ago

Its a great find, but probably impractical to run on a home electric service.


Reply 7 months ago

I did hook the heating elements up, in parallel, to my AC arc welder (~75V 140A). It took about an hour but the heating elements turned red and it felt like a really hot cooking oven. Should I just try that again for a longer amount of time or would that damage the elements? Also, you said that the elements are nonlinear, would that mean that as they heat up the furnace could reach a higher temperature with lower amperage and voltage or is my understanding of resistors just messed up?
I'm not too worried if I don't get it working as it only cost me $35 but I really want to figure out a way to use it :)


Reply 7 months ago

Can you measure your running current and voltage ? I'd try your experiment again, and see what they are doing when they are red hot. You MIGHT be able to run them on a higher voltage, and put them in series. I'll guarantee you won't be delivering 75V at 140A with a buzz box welder.


8 months ago

It is a standard electric heater with heating coils.
Just that yours are not only massive but also most likely too much for your wall plug.
There should be a label somewhere with vital details....
Otherwise you will need a good multimeter to check the resistance of the rods and do some calculations.
Could well be designed for the use with a transformer that provides a low valtage with a few hundred amps.


Reply 8 months ago

Thanks! that makes sense but is there a place I can find the calculations? As I believe it was made in house and it does not have any data plates aside from an identification number for the university


Reply 8 months ago

Makes it a bit harder though.
Start by disconnecting the thick flat cables on the terminals so you can measure the single rods.
I assume they will all be connected in the same way somewhere at the bottom but that should not matter too much.
If you can follow the power supply cable you should be able to figure out how and where the power runs.
Please mark everything you take off or take pics so you get it back together the same way.

Now assuming the rods are all connected in parallel:
If you measure the top of tow rods you should get a resistance value - this by half is for a singlerod and the all should be roughly the same.
Keep in mind the resistance will be quite low, so a dirt cheap tester from the local discounter might not be accurate enough.
In case all are equal and even they should be still fine and in working order.
Problem is the unknown supply voltage and without that it is next to impossible to figure out what current you will need.
The rods are quite thick, so lets assume measured from the top you get 5 Ohm for two.
That means only 2.5 Ohm for a single rod.
I guess you are US, so 120V mains devided by 2.5 Ohm gives us 48 Amps per rod!
For 10 rods you then need close to 500A :(
Per rod close to 6kW...
In this case it would almost impossible to use this without a transformer that has an output of around 48V or less.
How much a rod can take in terms of watt is limited by the max temp it can tolerate.

I suggest this:
Try to figure out how the bottom ends of the rods are connected.
Make a simple drawing where you replace the rods with resistors.
So all parallel rods as resistors in parallel and all rods in series as resistors in series.
Do this from the live part of the power cable to the neutral cable.
With that sorted you can start to measure the rods, works best if you mark them and the resistors in the drawing with numbers ;)
Add the measured values to the dawing.
From there it should be able to figure out a way how you can power this beast up without ending in a dark and powerless neighbourhood ;)