sword heat treat oven build

S

Stiode

Joined
Dec 18, 2016
Messages
13
Hi,

I`m planning on building a sword heat treat oven, Long enough to take up to 50" blade .
I want it to be vertical, with 3 sepperated heating zones. to keep the oven from getting to hot in the upper part.

And i need some help :)

Do anyone know of a PID controller that I can handle 3 thermocouples and that can controll three heat zones?
 
You would be better to use three PIDs. That way you could have constant metering and easier adjustments.

My friend Dave Catoe uses three 18" Sugar Creek kilns stacked on each other. Each has its own controller. He sets them all for the same value and they track quite well.
 
To do what you are saying, basically, yes. It will operate in three zones. If you build them with removable front and back panels, they can be used as three separate ovens or combined to make a longer one. If you want a dedicated sword oven, then build it as one unit, but with separate coils in the three zones. Even with this configuration, it will be simplest to build each segment separately and assemble them on each other.
 
Last edited:
Use a type K, 8-10 gauge, bare TC.
Use a TC mount to hold it.
Use type K TC wire to connect to the PID - observe polarity.
Use TC plugs and sockets at the PID - observe polarity.
 
Those controllers (Omega CN7823 and AutomationDirect SL4848VR) are both excellent. They work very well, but are industrial controllers and are not nearly as user-friendly as dedicated kiln controllers when it comes to setting up the programs. They are, as far as I can tell, the same controller. I have used both the AD and Omega versions and buy whichever is cheapest at the time.

They are able to autotune at up to 4 different temperatures and can automatically select the set of PID terms from the autotune temperature closest to the setpoint. This is nice, since it saves having to decide on a single temperature at which to optimize the control.

The "industrial" nature of these controllers means the menu system is not organized to keep the end user away from the settings that can louse things up. Setting the ramp/soak profiles requires you to go past the thermocouple type setting for instance.

I had a maker using an oven I had built inadverdently change the thermocouple type from N to S. The oven only reached tempering temperature but showed Austenitizing temperature as a result. That one was easy to diagnose and fix with a phone call, no harm done, no damage to any workpieces. It could easily have gone the other way though. Since then I've been keeping a book in which all the settings are noted with each oven.

I am not a great fan of type K thermocouples in HT ovens, much preferring type N. Type N was originally developed as an "improved type K" for high temperature applications (above about 1000 degC, 1832 degF) with reduced "drift". Type N tops out at 1300 degC (2372 degF) which is fine for HT ovens, even those intended to handle HSS, though I still use type K for forges, where the higher, 2500 degF, top end is useful for checking forge temperatures.

I have some industrial experience of thermocouples in temperature-controlled combustion plant and type N have proved vastly superior to type K in that application (1000 degC control setpoint, 1832 degF). That said, it's a different application to HT ovens

I tend to use Mineral Insulated type N thermocouple assemblies with either Nicrobell or "Super Omegaclad XL" sheaths and grounded junctions. These give nice fast response and can handle long horizontal insertion depths for accuracy. I usually buy transition joint assemblies with a meter or two of cable and a miniature plug already fitted to eliminate any chance of getting the wiring wrong. I use a miniature panel socket on the control box.

I am currently at the testing stage with a pair of 28" ovens that will bolt together to make one long one. I have stayed with a horizontal arrangement, partly to use a single control box and avoid the problem of vertical temperature gradient.

I'm always interested to know what process other people go through to specify equipment for their needs, so any more information the OP can provide would be appreciated.
 
Thank you for the answers.
I remeber reading that it could be a problem with a grounded thermocouple and reading problems.
If you touched a grounded thermocouple with your workpeace. It would give wrong readings.
so some recomended buying a ungrounded thermocouple. what do you think?

And also that the thermocouple gives the most accurate reading if conected directly to the PID controller, without any joints?

And could you please put up a link to the terrmocouple you are the most happy with?
Have tried to find one with all your recomondation, but it is still many choises.
And how long do it need to be?

Is this a good one?
http://www.omega.com/pptst/KMTXL_NMTXL.html#related
 
I usually buy my thermocouples from a local supplier who we use at work. For the HT ovens, I've been using a 6mm diameter grounded-junction Nicrobell-sheathed Mineral Insulated type N thermocouple with a transition junction (aka pot seal) onto 2 metres of type N extension cable and a miniature plug. These have been 18" long and have been mounted through the back of the oven to give an insertion depth of about 9", placing the tip near the centre of the 18" chamber length. 9" insertion with a 1/4" diameter thermocouple is about 36 diameters.

On my most recent ovens, the back has been removable so I've come into the side. Because my ovens are only 7" wide, I reduced the thermocouple diameter to 1/8": 3.5" of insertion on a .125" thermocouple is about 28 diameters. The general rule of thumb is to use an insertion depth of at least ten diameters, though I prefer to use more when I can.

http://www.asminternational.org/doc...p048.pdf/236107c7-ec38-40e6-8ecc-51a23dfefaa4

I use a grounded junction thermocouple to increase the response rate, but do not ground the sheath. This has not caused me any problems, though YMMV.

The Omega transition junction thermocouple in your link looks fine to me, though I'd suggest you talk to the technical sales guys at Omega and take their advice.
 
Thank you for answering! :)

I have now tried to search for a place to buy the kanthal elements.
but all the links I have found have been dead.
also the ones I found on ebay was too thin.

I`m going to build three ovens about 40-50 cm long ( 16-19")

I live in Norway so my power supply is 16A 230v. so i think 3300W is the maximum on each oven. but 3000W is maybe enough for the size of the oven. i want it to reach 1100 degree Celsius in no more than 20 min.
I will of course connect each oven on different fuses.

Any recommendation on where to buy elements?


Best regards Stian
 
Last edited:
came across an old thread where you Timgunn1962 used to buy elements from jrider12 on ebay. but i could not find him.
Does he still sell them?
 
I think he dropped off of ebay, then resurfaced as pmtoolco or something like it. I've not seen anything from either jrider or pmtoolco for a couple of years now.

I got my last lot of elements from KineticaCNC in the UK: 1.6mm wire and 1500 Watts apiece on 115V. I use 2 connected in series for 3 kW on 230V.

I've just finished building a couple of HT ovens that are each 28" long and will bolt together to make a single 56" oven. I'd tried pretty hard to build them identical so that they could both be run from a single controller when bolted together, but still found they needed to be run as separate zones using a controller per zone to get the sort of temperature control that I'd expect.

On the 3 kW elements, the 28" oven (6 IFBs long plus a 1", 25mm, fibre blanket door seal each end, which compresses down to about 1/2" with the door closed) will reach 1300 degC, 2372 degF, though it takes something close to 2 hours to do so. It will probably go a little higher, but 1300 degC was the limit of the type N thermocouple I was using.

From a cold start, 800 degC (1472 degF)was reached in 22 ½ minutes, 1100 degC (2012 degF) took 54 ½ minutes, the temperature at an hour was 1125 degC (2057 degF) and 1177 degC (2150 degF) took 71 minutes.

I suspect that aiming for 1100 degC (2012 degF) from cold in 20 minutes is being pretty ambitious. It might be possible with an 18", 450mm, oven on 3.3 kW, but I'd be inclined to regard two ovens at around 27", 685mm, each as more realistic.
 
Thank you that helps alot.
It is clear that i was too ambitious about the heat up time.. i underestimated the power needed to reach these temperatures

So you have buildt two 28" ovens that can be connected. is it horicontal or vertical? and how even does it heat up?


I will build so I can connect them vertical to about 48- 57 in lenght.
Since you build two ovens do you think it is overkill to build 3 ovens (three heat zones). Since its going to be vertical when connected I thought it would need three heat zones to make sure the top of the oven dont overheat.
 
I'd do one controller with 3 SSRs operating 3 different zones.
I've done that with 2 controllers and 6 SSRs with good result on some vulcanizing equipment I made for a company.
It was all solidstate and easy to wire up. The Solid State Relays will kick on when the temp controller calls for heat.
Just use one thermocouple in the middle and it will turn all of them on.
I think you can control your heat treat pretty well that way in theory.
I have only done this on a belt cooker for big Goodyear rubber belts in coal mines and quarrys to repair broken sections.
The heater probes were inside big platens and the controller box controlled 6 heating elements with 2 controllers just fine out in the open atmosphere and cooked hot.
In an oven well insulated with one thermocouple in the center would work fine..
I'd look into US Heat to start, I order from them and they have everything you need and custom make stuff very reasonably.
Just call him and tell him what you're trying to accomplish and he will have an answer for you..
 
For most industrial applications, I'd expect the single control loop to provide adequate control. My dual-oven experience suggests that it is unlikely on a Heat-Treat oven even one that is horizontal.







The control thermocouple was the one in the front (master) oven and I lashed up a second SSR to power the slave oven off the master control output.





I attached a handheld readout to the thermocouple in the slave oven, set the temperature to 800 degC (1472 degF) and started it.

The indicated temperature of the slave oven seemed to rise faster than that of the master, showing a difference of around 60 degC at 300 degC and 240 degC respectively. The gap then closed until both readings were within a couple of degC when the master reached 800 degC. Leaving it soaking at 800 degC for another half hour, the gap opened up somewhat until, after half an hour of "soak", the slave was showing 795 degC with the master at the 800 degC setpoint.

I did not open the oven or do anything that might disturb the temperature distribution during this time: I was looking for "best case" values at this stage. 5 degC of difference is 9 degF.

Having gone to the trouble of building the ovens, I would want to see better control than this. Running off a single controller was an attempt at keeping the cost down and I don't feel the cost saving justifies the reduction in performance. In future, any joining of ovens will be done with each of them controlled independently by its own control box. This will give the tight control expected of an HT oven.
 
7" Wide, 6" High internally.

I think I used 27 bricks for each oven, plus the board for the door and the back. The bricks are 9" x 4 1/2" x 3" JM23 IFBs made by Morgan Thermal Ceramics. They are easier to find in the UK than K23s.

The door and back are a single layer of 2" Ceramic Fiber board.



I've discovered that knifemakers as a species are seldom satisfied: at least 2 of my earlier (admittedly only 18") ovens have been modified by the end-user to get longer work in. I therefore decided to make these 2 ovens easily (and reversibly) extendable by simply adding a couple of additional layers of the board to the door and back.





If that's not enough (36"), then it's a case of bolting them together.
 
Thanks for all this great info, guys. I'll have to bookmark this thread for later reference.

I just wanted to add that if you're into European type swords, ya may wanna go a little longer than 50" overall. There are plenty of Longswords that were just a little over that length. Like 51" to 56" or so.
 
Timgunn, thats some solid work! Despite the performance not being perfect I do like your system for the "slave oven" connected with the extra SSR. Great door system too...I Still need to make an actual door for mine, so far it has been stacked up bricks as a door [emoji51][emoji15][emoji848]
 
You have a point there Possum.

Since I'm going to build 3 ovens, I will probably make them close to 20" each, so I can add 2 to get 40" and 3 to get a 60" oven.
 
You must log in or register to reply here.
Back
Top