DIY 1000 deg C Temperature Controlled Oven

Intro

I needed a high-temperature, temperature-controlled oven for various experiments, and the design below is the product.  The finished internal dimensions (including the soft brick and insulation) are 12.25" x 22.125" x 7.5".  

For full disclosure, the oven has only reached 940 deg C, but with more settling time and better insulation, it would easily achieve 1000+.

Oven

Oven parts list:

K-23 Soft Insulating Fire Bricks (10??)
Standard (hard) Fire Bricks  (40??)
Ceramic insulation (~4 ft)
Steel L-stock
Nichrome heating elements (6)
Chimney Cement (several tubs)
Steel plate
L brackets (2)
Threaded rod
Misc. nuts and bolts

The oven is designed such that every part exposed to heat is rated for very high temperatures (the lowest is somewhere around 2000 dec C) in case I decided to put in a second set of heating elements (which would require a second 240 V line) and push the oven hotter.

I started by assembling the base out of steel L-stock and welding the frame together.  I next stacked fire brick around the bottom and sides to make the oven.  I used chimney cement to seal the seams after placing the bricks.  I didn't cement the entire length of the bricks because I didn't need the structural support and only wanted to stop air flow. It also will make disassembling the oven in the future more of a possibility.  Next, I machined grooves in the soft fire brick using a router mounted to a CNC, and placed the firebrick in the oven as shown in the pictures.  I made extra grooves in case I ever want to add additional heating elements.  The soft brick is so soft that hand tools (powered or manual) can do the job as well [1].  I also drilled holes through the hard fire brick to serve as feed troughs for the nichrome.  I stretched the nichrome heating elements so that they were long enough to run two lengths of the chamber, placed them in the soft fire brick, placed and passed the leads through the hard fire brick.  I then used the remaining chimney cement to seal these holes.  Next, I placed the insulation on the bottom surface and the two exposed sides.

The lid is composed of a steel sheet (I don't remember where I bought it) with insulation bolted to the bottom.  The lid handle are two L-brackets with an attached threaded rod.  I realize that the lid isn't the best or slickest idea, but it works.

The nichrome heating elements each produce 1 KW of power at 220V.  This equates to 4.5 Amps per element.  That particular 220V circuit is rated for 30 Amps so I included 6 heating elements for a total of 27 Amps.  This provides a total of around 6000 W.  I realize that I'm pushing it a bit, but I haven't popped the circuit breaker yet.

Machining grooves in soft fire brick with a CNC router.  

Finished fire brick.  The holes in the sides and not finishing the long groove help hold the nichrome wires in place.  

Setting firebricks in steel frame.

Finished oven with lid.  

Finished oven internals.  I drilled a few feedthroughs in the side for various instrumentation (including the thermocouple).

Finished oven internals.  You can see all 6 heating elements.  Each heating element was stretched so that it runs the length twice.

Finished oven working.  I've been using extra fire bricks to shield the thermocouple and whatever I'm heating from direct radiation.  

Control Circuit and Power

Power box parts list:
Plastic container (container store)
220 Volt power source (such as an oven or laundry outlet)
PID controller
Thermocouple
14 gauge wire.
Higher gauge (small diameter) wire  

Fuse holder and fuses
Spade crimps
Small PC fan
12V transformer (240VAC to 12V DC) to power fan.
Relays (20 amp each)
Relay heatsinks
Switches, 240V rated

The relays listed above are limited to 20 amps each so I split the heating elements into two groups of three.  Each relay controls each set of 3.  This provides an added advantage that I don't have to use all 6 heating elements at once to heat the oven and will allow for slower rise times in more sensitive applications.  The relays also put out a LOT of heat, and I melted my first set during the first week of testing it.  Note: when my relays failed, the circuit failed closed which is really scary.  To prevent this, I added two aluminum heatsinks and an old PC fan to the new relays.  I have not had a problem since.

Also, my Rex C100 PID controllers came with an upper temperature limit of 400 C which is insane as K type thermocouples should are capable of 1350+ C.  There is a hidden menu in the Rex C100 that let's you change this limit [2].  At the main screen, hold set and the left arrow.  Cycle through the menu items until you come across  the SLH setting, and change this new value to something higher (like 1250).

I wired each set of three heating elements in parallel to the control box.  I used crimps and silver solder to ensure good conductivity.

Control box internals.

Control box front.

Conclusion

There are certainly improvements that can be made to the oven, but it has worked remarkably well.  I'm quite satisfied with the end result.  

Note:  I am, however, struggling with one issue that has recently appeared.  The temperature reading on the PID controller has shifted quite dramatically from the correct value.  I changed therocouples just in case, but that didn't fix it.  I'm planning on replacing the PID controller next, and I'll post back here with my findings.  

References

[1] https://www.youtube.com/watch?v=en4yhzLuD9A

[2] https://www.youtube.com/watch?v=_R7_NPLeTI8