DIY Heat Treat Oven Build

Status
Not open for further replies.
Joined
Jan 1, 2018
Messages
862
I was asked to put something together on my heat treat oven so I'l try to provide some of the details and issues I ran into. This will be long so I'll try to break it up into sections. This project isn’t nearly as complicated as you might think and you can build an oven for 1/3 the price of buying one. I will post a you tube video of the oven Here:


Heres some pics of the oven.

https://flic.kr/s/aHskx6ZK6Z

First off, I give all credit to Dan Comeau Custom knives. He has an incredible site with all kinds of builds that are extremely detailed. For this build he even has an excel sheet to calculate you coil design so thanks to Dan for this. I'm a Civil Engineer, not electrical so if I made a mistake anywhere, please feel free to correct me.


I am Canadian but will use imperial measurements as it seems that it's mostly Yanks on here. Prices will be in CAD and Temperatures in F.


I designed this oven to be 240 volt and to run at 3000 watt. With a single element, this draws 12.5 amps (calculated) with a given wire resistance of 19.2 ohms. The internal dimensions are 14.5" by 7.5" by 6" high. I should be able to treat a 16" blade diagonally. This gives an internal area of 0.377 cu ft. At 3000 Watts. this about 8000 watts per cu ft. This is a bit overkill but I didn't want to have to wait an hour for it to heat up. It takes about 15 minutes to reach 1900 degrees F. The total cost of the oven was about $400.00 CAD plus Tax so close to $500.00

I bought the insulating firebrick and Kenthal Wire from the Pottery Supply House in Oakville Ontario.

https://www.psh.ca/.

Here’s an Amazon Link

https://www.amazon.com/Amaco-28035N...=8-2-fkmr0&keywords=insulating+fire+brick+k24

It cost about $200 for 24 bricks and 150ft of 16 ga kenthal wire. This is almost 3 times the amount of wire I needed but it was good that I bought extra. For the Firebricks, I used K23 9x4.5x2.5. This was the most expensive part of the build. Youmay want to buy a couple extra brick because they chip and break easily.

Oven Design First steps:

First you need to decide what your available power is. 120 or 240v. Then decide how many amps you can deliver. 240v uses 1/2 the amperage. I wanted about 3000 watts so I decided on 240v at 12.5 amps. Dan’s site has a chart that shows commercial ovens and compares their wattage to internal volume. These average about 5500 watts per cu ft. I decided on 3000 watts and wanted roughly 8000 watts per cu ft so I needed about 0.377 cu ft. I went high on the wattage as I figured my oven may not seal as well as a commercial unit so I'd compensated with more power. Dan's site has a calculator you can download to figure all this out. Keep in mind that if you use 2 elements in his calculator, it gives you the length of wire for only 1 element.

Also keep in mind that there is a relationship between your wattage and internal volume that will limit your design. I had first intended to use 2-3 elements in parallel in case one burned out, I didn't need to replace them all. However, as resistance is calculated as 1/(1/R1+1/R2+1/R3) every element you add essentially lowers the resistance by a factor. Two elements in parallel have half the resistance of one element. 3 would have a third. Wattage increases as resistance goes down. This may seem counter intuitive but a five foot element would have twice the wattage as a 10 foot element due to the fact that there is less total resistance. Say you decide to use 3 elements. The resistance in each element needs to be 3 times more than a single element to make the same total resistance. Multiply this by 3 elements and you have 9 times the amount of wire. You may run into a problem where your oven simply doesn't have enough room for all the elements. This is why I ultimately decided on a single element of this design.

240v
3000 watts
12.5 amps
19.2 Ohms Resistance

resistance for 16 ga wire is 0.324 Ohms per ft so I needed 59 ft of wire.
If I were to use 2 elements I'd need 118 ft per element or 236 ft of wire.
If I were to use 3 elements I'd need 177 ft per element or 530 ft or wire.

I hope this makes sense. Using a heavier gauge wire means that you need more as well because the resistance per foot drops. Same principal as above.

You may ask, if elements in parallel have to be longer, than what’s the advantage? Having elements in parallel means that each element is using less watts per length than a single element. 3 elements would use the same power but would be 9 times longer so they should last much longer. You may also get more even heating by having more elements in the oven. I didn’t think this was an issue as the elements are fairly easy to make and install.
 
Making the elements.

The pottery supply house shows coils of element wire in 1 lb spools. When I showed up, I found out that they actually sell it by the foot and need to measure it out. This took about 1/2 and hour as the guy who normally does this wasn't in so you want to call ahead if going there directly.

To measure the wire, I clamped 1 end in the vice and ran it out my barn door around a wood post and back in. I took an OHM meter and measured the resistance at certain points until I got 19.2 OHMs. I cut it off a bit longer for connections. You could just check the rated resistance of the wire and measure the length that way if you don't have an OHM meter. Make sure the wire isn't crossed anywhere or your resistance will drop due to the short.


To coil it I took a piece of plywood and drilled a hole that would accept a ½” fiberglass driveway marker. I Drilled a small hole in the end of the marker and stuck the wire in it with about a foot extra. A wood dowel would work as well. I stuck this in my drill and with one hand on the drill and the other feeding the wire as I coiled it around the marker. I did this by myself which was a bit difficult but manageable if I went slow. A helper would make this really easy. Make sure to straighten out the wire first. If you don't, it gets tangled and every time you stop, it unravels a bit. I did see a video where someone made a jig for this.


My way was a bit uneven and I did have a couple spots where the wire jumped over the previous loop but it was manageable. Once the wire was coiled, one end was clamped in the vice and I stretched it out to the length I wanted to fit in the oven. I think coiled it was about 4 feet long and I stretched it out to 15 feet long.

Firebricks

The firebricks are very soft and easy to work. I routered the edges so that they fit together in a tongue and groove fashion. 24 bricks was perfect for this build. Make sure to wear a mask when working with these as the dust is deadly. Routering is not a good idea safety wise. I'm not really sure if it's required as the commercial ovens aren't but I wanted to retain as much heat as I could. I used high heat mortar on the joints. I mortared the bottom and sides together first. Then I made the element grooves with a die grinder followed by mortaring the side and back to the bottom. Make sure you dry fit everything first and if you make tongue and groove, leave enough room for the mortar. Again, please use a mask when ever working with these and clean up all the dust.

Shell

Once the fire bricks were assembled I made the shell. This was simply 20 gauge sheet metal and 1x1 angle iron that I tig welded together. If you don't have a welder, I'm sure you could bolt everything together with threaded rod. Make sure you leave an opening for the element to protrude through. I also welded 4 small bolts onto the side to mount to control box to. I didn’t want to mount it against the sheet metal side because I wanted a small air gap for cooling.

Installing the elements.

You'll want your elements to be a bit shorter than the grooves in the oven. That way you can stretch them to fit. At the ends of the elements it twisted the wire together. By doing this, the resistance drops so I don't get as much heat in the control box. Make sure you secure the elements really well in the oven. I first tried using those u shaped nails for securing fence to posts. I didn't use enough and when I first fired the oven, the elements expanded, softened and drooped and shorted. I needed to make a new element. To secure it the second time, I cut pieces of the extra wire and made u shaped pins that I pushed into the firebrick. I installed these every 4-5 inches to make sure the element stayed in place. I didn’t put elements on the ceiling. I read somewhere that his can cause uneven heating.
 
Controller and box.

For my controller I used an Inkbird PID from Amazon. $50 and it cam with 1 SSR.

https://www.amazon.ca/gp/product/B01LQ8TPA0/ref=oh_aui_detailpage_o02_s00?ie=UTF8&psc=1

I ordered a second SSR as well $13.99. This is important because I used 240v, If I didn't, I wouldn't be able to turn the elements off when the door was open which could lead to a potential shorting of the elements when putting in a knife. If you do a 120v design, you won’t need the second ssr. The SSR link is here:

https://www.amazon.ca/gp/product/B06WD65K6K/ref=oh_aui_detailpage_o01_s00?ie=UTF8&psc=1

I also needed a new thermocouple as the one that came with the inkbird only goes to 400 degrees. I bought this $15.30

https://www.amazon.ca/gp/product/B00XJB4DYQ/ref=oh_aui_detailpage_o02_s00?ie=UTF8&psc=1.

There were cheaper options but I didn't want to wait for Chinese shipping.

For the control box I used 2 plates of .125 aluminum. I bolted these to the oven to some small bolts I welded to the side. The aluminum is important to act as a heat sink for the SSR's as they heat up. I posted a wiring schematic for a 220v oven. Dan has these on his site. I may add a small fan to keep everything cool.

I made the cover for the control box out of sheet metal. I used an on/off toggle switch. Make sure you get the right type for you application. For my 240v design I got a 20amp, DTSP (Dual Throw Single Pole) On Off switch. This switch has 2 poles 1 for each hot wire controlled by a single toggle. It was about $20. Each circuit has its own 15 amp fuse. There is also a 120v LED that shows the power on and one that shows when the elements are on. There’s a bit of wiring involved but it's pretty straight forward. Make sure you have enough room to connect the wires to the PID and can get a screwdriver on the screws. I connected the SSR's to the elements by using the brass inserts from heavy duty mar connectors. Not the twist on type but the kind with the little nut inside. A limit switch was also wired to the door to make sure the elements are turned off automatically when the door is opened. This is critical to avoid shorting out the elements. Dan show’s this on his site.

The door simply has a couple hinges welded on it. It's held in place by a threaded rod with a wood handle that screws into the controller plate. This gives a good firm closure.

One issue I had was with the limit switch. I installed a bolt on the door that contacted the limit switch. When the oven heated up, it expanded and I lost my contact. I thought it was an issue with the PID as it very consistently happened at 1940 degrees. By losing contact, it kept shutting off the elements until the temperature dropped about 200 degrees. I fixed this by mounting the bolt on a spring and allowing it to maintain contact with the button on the limit switch.

Another stupid mistake was messing with my PID. I wanted to confirm the temperate so I kept testing metal for non magnetic to check the temp and by melting salt. I found that it was out 190 degrees. I adjusted the calibration on the PID. What I didn't realize was that my thermal couple probe was not seated all the way down in the sleeve and so it was reading too low. So after calibrating, the oven was running almost 200 degrees too high. I heat treated some knives and coupons of AEB'l and got massive grain growth because I was probably up at about 2150 degrees. I melted my thermocouple and deformed one of my SSR's. Don’t make that mistake.

I replaced the thermocouple and everything seems good now. I'd still like to get some pyro cones to check the temps but the heat treats seem to have gone good pretty good so far although I don’t have a hardness tester to check. That’s the advantage of a commercial unit. You can probably trust the temperature more.

After an hour of running at 2000 degrees, I can still place my hand on the shell without getting burnt. When I open the door to put a blade I, loose about 100 degrees but it recovers in about 45 seconds. I do have a poor fitment on my door where I can see some gaps. I plan to stick a piece of sandpaper to a board to sand the mating surface to get a better fit.

Overall it works very well. You could build one of these in a day if you have all the materials and are well planned out. I could probably make one in about 5 hours from scratch now that I know what I’m doing. For $400-500 it was worth it.

One thing I was thinking about doing was installing a nipple so that I could flood it with Argon from my tig. I didn’t do this as the blade will oxidize anyways as soon as the door is opened.

I hope this helps anybody looking to make one of these. I'll to answer any questions you have. I've learned so much on this forum, that I hope to give back a little.
 
I built my own oven a few months ago. I used some scrap steel and sheet metal from a microwave i found on the side of the road. It actually is pretty awesome and works perfectly, but i do want to remake the exterior now that my welding skills have improved.

I have a question, does your electric not get hot being so close to the where the door opens? Or do you let it cool down with the door closed? Seems like you would have time when you do your heat treat, or else leave a hot oven unattended if you have to run errands or something.
 
I built my own oven a few months ago. I used some scrap steel and sheet metal from a microwave i found on the side of the road. It actually is pretty awesome and works perfectly, but i do want to remake the exterior now that my welding skills have improved.

I have a question, does your electric not get hot being so close to the where the door opens? Or do you let it cool down with the door closed? Seems like you would have time when you do your heat treat, or else leave a hot oven unattended if you have to run errands or something.
It did get hot when I had it calibrated wrong. That's why I was considering installing a small computer fan that runs on 120v. It would run when the oven is on. However, after putting it back to factory it's been fine. The door is open for such a short period it's not an issue. The electronics seem to get more heat through the casing and the element leads.

As for running it unattended, I don't really do that. The heat treating process is so fast I just hang around. Pull one out and plate quench, stick a new one in then so sub zero on the first. 5 knives only takes about 20 minutes.
 
Awesome write up! very much appreciate this. I have a few questions if you can answer them. What diameter rod/dowel did you wrap your coils on? How do you determine the amps? And is the second SSR the only difference between 120v and 240v? Other than the fuses and wiring and such? Thanks man! Really greatful for the help!
 
I used a fiberglass driveway marker. It was 5/16" diamater.

My 220v circuit has a pair of 30 amp fuses. I want to stay well under. You need to figure out what size you want first and get the internal volume of your over. Once you have that, you decide how many watts per cubic feet you want and that gives you your watts. As noted, I was aiming for 8000 watts per cu ft which is overkill but why not.

Power = Voltage x Amps so 3000w = 240v x 12.5A. I could have upped the amps and there for increase the wattage but I was fine with this design.

The second SSR is because of the 240v and having 2 hot leads. If you use 120v you will only need 1 SSR.

Hope that helps.
 
I just wound new elements for my oven too. I made a jig like in the video and they came out real nice and even. I used a 1/4 rod to make mine and it turned out bigger than the first ones I bought ready made. The rod I used was salvaged from an old stand that holds wreaths for a funeral.
 
The second SSR is because of the 240v and having 2 hot leads. If you use 120v you will only need 1 SSR.
Even if using 120V, dual SSRs a good idea, More so if its a plugged appliance. If plugged into a backward wired receptacle, when "off", coils will remain energised above ground.
At my job, I encounter average 8% receptacles wired backwards.
 
Dual SSRs for 120 is a bit overkill i think. Whatever switches you have that ultimately give power to the element, you can get double pole single throw versions and put the hot wire on one pole and neutral wire on the other pole. So when the switch is open it totally cuts off both current carrying wires.

Edit
For me, i have a black wire from power cord to first pole of a manual switch to the ssr to the first pole of a door interlock switch to the element, then off the other end of the element i have a white wire that goes to the second pole of the interlock switch to the second pole of the manual switch to the power cord. All connections listed above are in series, it is just one big loop with lots of connections. So if either the manual switch or the door switch is open, the element is totally disconnected from the power cord, and both switches have to be closed and the ssr activated in order for current to flow through the element.
 
Last edited:
For me, i have a black wire from power cord to first pole of a manual switch to the ssr to the first pole of a door interlock switch to the element, then off the other end of the element i have a white wire that goes to the second pole of the interlock switch to the second pole of the manual switch to the power cord. All connections listed above are in series, it is just one big loop with lots of connections. So if either the manual switch or the door switch is open, the element is totally disconnected from the power cord, and both switches have to be closed and the ssr activated in order for current to flow through the element.
Ok, Sounds like you have it good.
Lots of guys interlock the door directly to SSR control circuit. With that, everything still works, But backward wired receptacle will leave coil hot at reference to ground.
 
Do you mean they put the door interlock in the low voltage circuit between the pid and ssr? I have read ssr can fail closed, and so in the rare case the ssr fails and you open the door without noticing it failed, and didnt manually kill power to the element, you would be vulnerable.

I am collecting materials to run a sub panel (my breaker box is full, i dont understand why contractors feel the need to save $20 on a small box) and when i do i will build a 240 volt oven so i can do stainless steel right. I will just get another element same as the one i have now and run them in parallel. The element i have hasnt hardened, hasnt gotten hot enough to harden.

You know, building ovens is a time consuming thing, anyone that takes it on, i recommend going full bore and doing it once right the first time.
 
you mean they put the door interlock in the low voltage circuit between the pid and ssr?
Right, its easy two wire connection and any dinky ‘ol form “A” switch can work.
As far as SSR failing closed, thats possibility. But any switch potentially fail too. If SSR fails, presumably one notices oven going too hot and something wrong.
 
Last edited:
The door simply has a couple hinges welded on it. It's held in place by a threaded rod with a wood handle that screws into the controller plate. This gives a good firm closure.
First thanks for this thread :thumbsup: About door..............I don t like that screw ! What if you have several knives in oven for HT ? When you pull first knife you need to immediately quench , right ? Who will close door in the
meantime ? Change that for something faster ;)
 
Last edited:
I just give it a quick quarter turn or so quench and then tighten. But your right. A quick latch would be nice
 
Thanks for the info. I'm in the process of building an oven and the info here has been really helpful. I live in the GTA as well,and I found it funny when you said the firebricks were the most expensive part of your build. I'm using 2.5" k23 bricks as well, and it only cost me about $4/brick. I ended up wasting a whole bunch experimenting with how to cut them, so I agree. If you're new doing this, buy extras.
 
I just used a hacksaw blade and router for the coil grooves. I don't recommend power tools as the dust is dangerous. I also made the mistake of not making the groves deep enough the first time. I've lost a couple coils from clips burning out and coils sagging and shorting out. I since used some stainless tig wire for pins. I also read somewhere to put in a small ledge to hold the coils. I made my grooves deeper on the last coil replacement by grinding out a scraper from some left over aebl and it works well.

The screw clamp failed BTW. As it was threaded into aluminum the threads eventually stripped out mostly because it wasn't spring tensioned and the oven heating up expanded things and it got too tight. I'll be modifying with a quick release. Live and learn. I did install a fan and it keeps the electronics very cool now.

I also tried this thermo couple but I burned out after 2 uses so avoid that one.
https://www.amazon.ca/3-3ft-Tempera...9201870&sr=8-4&keywords=K+type+thermal+couple
The first one I suggested lasted way longer. I'd suggested buying 2 so once the sleeve burns up you have a replacement.

I should note that I tend to do batches so I have this running for several hours. My last batch was 28 blades.
 
You don't want a sleeved TC. You want a 8 to 10 gauge bare TC and a TC block on the outside to hold it.. here is what you want. Remove the insulators as needed and shorten the leads to make the tip stick in about 2":
Tc and Block 1.jpg
 
Last edited:
2W1EscO


What I have so far. The back wall is off aligned a bit, but should work fine. I gave up on the hacksaw, which is why I wasted wasted 4 bricks. Took me a while, but finally got all the bricks cut.
 
That is certainly one way of doing things.

It is important that the thermocouple respond fast enough to temperature changes. A 3mm or 1/8” diameter Mineral Insulated thermocouple with a grounded sheath is likely to respond considerably faster than a 10AWG wire-and-ceramic-insulator thermocouple. The speed of response is particularly important when using controllers without ramp-and-soak capability, such as the Inkbird, and running to tempering temperatures, where slow thermocouple response can lead to serious workpiece temperature overshoot.

Up at Austenitizing temperatures, the rate at which the forge temperature changes is slow enough that pretty much any thermocouple will respond fast enough.
 
Status
Not open for further replies.
Back
Top