Heat treat oven PID controller

[John Katt]

Well my oldest son and I finally got back to finishing up and programming my heat treat oven, Thank god for youth and computer knowledge, The PID instructions were 12 pages long, and crazy complicated.

Anyway I was hoping to get some input on ramp curves and the potential of over heating blades, my oven is 7 x 7 x 18 and I am running a single 14 gauge kanthal heating element 220 volts @ 4590 watts, I'll be heat treating 440C, So from room temperature to 1400F takes just short of 5 minutes and from 1400F to 1850F takes just 3 minutes

so what I'm wondering does anyone feel that I should lengthen the ramp times to give the steel a little more time to equalize so as not to over heat the thin edges of the blade or is 5 and 3 minutes good enough

http://www.auberins.com/?main_page=index&cPath=1

 

[Stacy E. Apelt - Bladesmith]

Your times are almost surely not accurate. They are probably the air temp readings, or the radiated temp on the TC. Allow enough time for the entire chamber, to fully soak before putting the blade in. It takes a little time for the refractory to heat up.

I would allow 15 minutes for the oven to stabilize before putting the blades in. The ramp rate can be programmed at 9999. Once the oven is at 1400F the rise to 1900F should take about 10 minutes. You can program that in, or just ramp at 9999. The soak time counter should not start until the blade and chamber are both stabilized at 1900F.

 

[John Katt]

Thanks Stacy, I had a feeling that giving the bricks a little more time to heat up would be a good idea, by lengthening the ramp time that will force the pid to lower the power curve, even with the short 5 minute ramp time that we are using now the power output is only around 60%

During the design phase I read in numerous places to minimize inaccurate temperature readings the probe should be at the bottom of the oven close to the sidewall, so I placed it parallel to the bricks in the corner coming in from the back at 1/4 inch, after putting the pid in learning mode the temps stayed surprisingly stable, vairying no more then two degrees

 

[Stacy E. Apelt - Bladesmith]

Try laying a 2" by 1/4" or thicker bar of steel over the TC, covering it from direct radiation of the coils. See if the heat up time lengthens a good bit.

Once a PID learns how to regulate an oven, the accuracy is quite good. It is a good idea if possible to use another TC and reader (PID or gauge) to check how accurate the installed TC is. The two should read within 10 degrees, which is 1/2% at 1500F.