Problems with Silicon Carbide Heating Elements

Answer 1 = very unlikely

Answer 2 = very likely

Not my area but two comments:

1) Posting manufacturer's name and part number (or other sourcing information) might allow readers to be more helpful.

2) Heating element is probably much cooler when you measure it. To heat the furnace to 1000 C and overcome loss the element will need to be much hotter than 1000 C. When you remove power it will quickly cool down due to the cooler 1000 C ambient around it. This will probably change the resistance value quickly.

Attempting to use on ohm meter on the live element is, as you know, both very dangerous and guaranteed not to work.

Can you do a little Ohm's law to measure what is going on while the elements are powered up? Measuring the voltage across the elements might be possible. If you have a known length of known AWG cable from a nearby junction box then with a resistance lookup and calculation you have a known value series resistor. Mr. Ohm could be useful here. Obviously, live circuits are dangerous and measurements should only be made by qualified people who are confident that they can work safely.

Your setup isn' t working and making better measurements won't make it work better. But, making better measurements might help you better understand what is going on.

In summary:

#1 no

#2 yes

The filament of a tungsten light bulb has a much lower resistance when cold to when hot, Mildred. It isn't as though such a phenomenon were uncommon.

You haven't explained very well why you see it as some sort of problem as, after all, it just is, really.

What is the design voltage on the elements? Three elements in series dropping 230 volts comes to about 76 volts drop each. Voltage specs on most elements are to available voltages. They match the local voltages. Sure they can be placed in series. But to work properly the sum of the design voltage has to add up to the voltage applied.

I hope the following is helpful, but I have never worked with such a furnace myself, but I do have a Wood Pellets heating system that has a similar heating element.

When I bought it, it had a Chinese element, this burnt out after (all dates approximate and from memory) about 12 months. I contacted the company and they sent me two new ones, also of Chinese manufacturer. Each of which lasted about 12 months.

They were difficult to remove as the heating "end" swelled up and I had to use a Dremel to cut it off "in situ"!!

I contacted a German manufacturer and bought two of the same size and wattage, but "Made in Germany" I believe.

That was in 2010 and the first one is still working fine.....which to me shows excellent quality.

The same company make heating elements in all shapes, sizes, wattage, voltage etc...

If you want full manufacturer details, send me a CR4 private email with a normal email link for you where you can receive a catalog in pdf format and I will send you all the details that I have.

If you need some German/English assistance, I can help you there as well.....but I believe there is an English version of the catalog as well....

I would expect the elements to be available also locally to you as well....probably a lot more expensive, like twice the price at a guess.

I paid around US$50 for two units including postage....at least for the ones for Pellets burners, which may of course be cheaper, I have no real idea for your needs, sorry.

I'm aware that the resistance value of SiC elements changes with temperature. The furnace is heavily insulated and has been running on above 1100 celsius for weeks. When I shut off power it takes hours to cool down even 100 degrees. Even if the temperature dropped to room temperature the resistance change isn't that drastic.

The problem is that I would like to know the resistance values of individual elements to help me with troubleshooting.

How would the currents and voltages behave if one element in one phase had a resistance that was way off? Here's a drawing of what I mean. My voltage and current values are balanced across all the phases. Should this be the case if the resistance values of the phases are not balanced? The wye center point is not connected to neutral line.

I don't think there really is such a thing as design voltage for SiC elements. With certain voltage and resistance you get certain power. Then you make sure that surface watt loading doesn't get too high and you won't overload the elements.

Your right with a certain voltage and resistance you get a certain power. There is a resistance curve with Sic elements. Which is based on the temperature. You need to know what that resistance is for the temperature your trying to achieve. To compute the voltage necessary.

Because resistance can differ between elements at the same temperature. Less expensive elements more so. I would check voltage drop across the elements at the temperature your trying to achieve and see where that differs.

I would go back to the manufacture for the temperature resistance curve data.

If you are seeing the 30 Ohm reading, then perhaps there is an issue with contact resistance in the mounts? Damaged elements?? Are you positively able to remove and isolate each element in order to test it?

Compare results with open wye calculations? I really do not know, but you probably need someone with more expertise with SiC from China. Make sure the elements are not cracked...

Hi

I dabble a tad in SiC elements and furnaces.

Your resistance is double the standard off the shelf elements, be it from China or elsewhere. Resistance can only be measured at around 1050C and once power is switched off the temperature drops so quick, measuring it is very difficult and dangerous. I stand corrected but you need to measure it with a True RMS multimeter.

Some Chinese manufacturers supply with a 25% tolerance but yours is around 40-50%-poor quality I believe. Hence the huge variance in your actual measurements. I think they marked them the ordered values just to keep you happy upon delivery, normally they are marked with the actual resistance.

Breakages are expected and the parcel should have had a "shockwatch" sticker on so you can go back to the shipper and claim.

My advice is to stay with a standard type element, its cheaper and if you find the right supplier they will assist in the furnace design and supply the right ratings.

I would take Andy up on his offer.

In our business it is well know that our friends on the other side of the Pacific will paint a sour orange yellow and sell it to you as a lemon.

Good luck!

Here is something you will want to think about. What is going on where you connect your electrical conductor to your heating elements?

Copper or Aluminum conductors will draw a lot of heat off in the direction of the power line. One very, very good idea it to use a transition pieces (little bars) of stainless steel between the power line and the heating elements. You will need one for each conductor.

Because stainless steel is a good electrical conductor but a poor thermal conductor, you will recover a lot of energy going in the wrong direction. Got any pictures?

I measured the voltage drop across the elements. It differed some but didn't really give me any clues.

The interesting part is what happened when I measured the currents afterwards. Now one phase is drawing nearly 20 A, which is exactly what it should be. The two other phases are drawing around 13 A each, which is couple amperes more than before.

I didn't make any changes. Fiddled a bit with the wires and connections so something may have moved. I've been moving them around a whole lot more before and it haven't made any change.

I was cleaning my Pellets Burner yesterday and I had to remove the heating element to clean the air path and it slid out as if it was new, no swelling or other damage.

It was a bit discoloured from 4 years of usage, that was all.....it still ignites perfectly.

"Made in Germany" of course!! Not China.....

Although the current draw is now little closer to what it should be the furnace is still very inefficient. Using close to double the power this sized and this well insulated furnaces normally do.

Do you have any thoughts on what could have caused the sudden rise in current draw and what could cause to furnace to run inefficiently? Bad connections? Electricity leaking into the refractory materials?

Andy Germany: I actually talked with a German manufacturer, but they don't supply element's with the specs I need. I would need to do major changes in the furnace to fit their elements. Might do that in the future, but it's not possible at the moment. Thanks for the offer though.

No Problem?

Was it the firm called Jeka you talked to?

what could cause to furnace to run inefficiently? Bad connections? Electricity leaking into the refractory materials?

Not likely. Most refractory materials are not conductive because they are made of ceramic fiber. Bad connections would show up as conductors whose insulation has shrunken or burned back from the connections.

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

Now one phase is drawing nearly 20 A, which is exactly what it should be. The two other phases are drawing around 13 A each.

This sounds like a broken element if you have the heater elements wired in a delta configuration. The leg that measures 20 A is the vector sum of the other two legs.

I believe that at over 1000 degrees Celsius the refractory materials can conduct electricity. As will glass. However the SiC elements are installed according to instructions and should not make contact with the inner furnace walls. But there's no way to tell for sure.

The elements are wired in wye configuration actually.

Well then, I'm a little bit confused. I would expect "lead glass" to become conductive but "lime glass" might have to reach a higher temperature before it becomes conductive. My experience was with Inconel heater elements. We had ceramic insulators and ceramic fiber insulation in the oven but softened lime glass would pull the heating elements out if they made contact. It was a very significant pain in the neck.

Since I can not see your situation, I am potentially at a loss to help. But assuming the SiC heating elements have the same rating, your current imbalance is your best indicator of a broken connection. Ideally your neutral current should be zero, but with one element broken the circuit will be unbalanced and your neutral current will be the vector sum of the other currents. Big neutral currents is your best alarm in this case. If the wye connection is made inside the oven, it would be helpful to bring it outside. That, of course, assumes that your power feed is also a 4 wire wye. If not, then you heating element is being exposed to over currents when one element is lost. In that your case, an open neutral changes the circuit such that two elements are now in series with a phase to phase connection and the vector sum puts it over the top.

Let us know how it works out!

It is an open neutral configuration so that I only have to control two legs.

I wired it in 4 wire wye to see what that does. It gave me currents of 10 A, 11 A and 32 A. I also connected it directly to the power outlet to eliminate the control system from the equation.

If an element is lost shouldn't the whole phase be lost?

I will be ordering a new set of elements anyway. Unfortunately having to rely on Chinese manufacturers.

Sorry about the weekend delay. If each element is powered separately, it will confirm that they are, in fact, the same design or part number. If you confirm that each element acts as it should at any specific voltage, that is, they each pull the same current plus or minus say 10%, then the problem is in the control system. You could rotate the wires feeding the heating elements to see if the problem moves to another wire.

If the problem is really in the control system, it is because the three phase output has a short that makes it something other than 3 phase. To check the actual output, you will need to verify 120° phase difference under load. That is most easily done with an oscilloscope. Many controllers synthesize the three phase output using time delays. But an internal shout might continue to operate, just not as expected. In that case, the equasions are very, very different.