Jump to content

Recommended Posts

Posted (edited)

OK, here's the idea, I have an older ceramic kiln that is a usable size for HT.

It has analog controls that the needle assembly is screwed up so no can get proper temp reading.

Iv'e found a digital control system at bang good for a reasonable price and could use some input as to

feasibility of this more than likely harebrained scheme.

As to temps it is capable of I don't know at this time, but would be willing to upgrade the elements if needed, the box is very well insulated

and I think worth the effort to upgrade.

Anyone have experience with this that you would share ?? 

Edited by Clifford Brewer
ba speiun

Share this post


Link to post
Share on other sites

Any decent ceramic kiln should be more than adequate for a heat treat oven for "normal" carbon steel.  You will need to ensure that your power supply is correctly sized for the amperage draw the kiln will have (likely in the 30-60 amp range) and will need a contactor or SCR to interface between the controller and the kiln elements.  I don't know anything about selecting SCR units other than they will usually need some kind of variable signal from the controller.  The temperature controls I've wired are usually configured to control mercury contactors (which stand up better to the sometimes rapid cycling you can get with a temperature controller than conventional magnetic contactors) with "dry contacts" from the controller energizing the contactor's coil.  You will also want a thermocouple and thermocouple wire to use as a sensor.  I recommend type K thermocouples for normal temperature ranges involved.

Good luck.

Share this post


Link to post
Share on other sites
Posted (edited)

Thanks Dan;

The kiln is insulated much better than my propane forge is. that is why I looked at it for this conversion,

The unit is 120v powered so I'm not sure if that will get to the temps I need.

I don't know if the elements can be used with 220v or not, still studying this part, the control unit I have found

is designed for a type K thermocouple so I'm thinking that it should work just tryin to get all the ducks in the right row here.

I can wire up a house or a shop, but stuff like this is not something I am comfortable with, and don't want any bad issues to

jump up at me, so I'm using the only dumb question is the one you don't ask before you hurt yourself theory here.......

Again thanks for your input.......................................B)

Edited by Clifford Brewer
pukshasun anspelun

Share this post


Link to post
Share on other sites

The kiln is almost certain to be able to achieve Austenitizing temperature for Carbon steels at its design Voltage. There's a good chance it'll be able to handle most of the less-exotic stainless steels too. 

I don't know which of the controllers from Banggood you are looking at. They list several and I don't think they are all suitable for a HT setup. Some, particularly the Rex C100, seem to have factory-configured input ranges which cannot be changed by the end user. A 400 degC maximum on a typeK thermocouple seems to  be the most common input and this is not much use to us. 

There is an XMT612 controller listed and this "may" be the same controller as a tet612 that I used a few years ago. The tet612 had fully-user-configurable input ranges and worked pretty well for me. 

As Dan says, the controller needs to have some means of modulating the heat input to the oven. This is usually done with a Solid State Relay switched to provide time-proportioning control. I have found a 2-second output cycle time gives about the best results.  The controller needs to have a DC pulse output to trigger the SSR. On US 220V supplies, I understand 2 SSRs are normally used, one for each "hot". Most controllers with pulse DC outputs can trigger 2 SSRs in parallel.

Whatever controller you get, make sure you have downloaded, read and (mostly) understood the manual BEFORE you buy. The download needs to be from a site accessible without opening some sort of account or logging in with anything that could be interpreted as a "feel free to spam me" permission. I would expect the potential supplier to be able to provide a link to the manual

As you say, this is not really in your comfort zone and you may need to ask for help. If you ask for help with a link to the manual, someone with a general understanding of PID controllers can probably help. If they cannot access the manual, you are on your own. 

One thing worth mentioning is that it is wise to use a 48mm x 48mm (1/16 DIN) controller and to leave a bit of extra length on the wiring: maybe 2". 1/16th DIN is the most common size for controllers and a later upgrade to a ramp/soak controller is easy later on. Any 1/16 DIN controller will fit the mounting hole, but some are longer than others and the terminal layout may be different, Having the wires long enough to reach can save a lot of cussing if you ever change the controller.

Share this post


Link to post
Share on other sites
On 3/2/2019 at 5:10 PM, timgunn said:

(1/16 DIN) controller

My systems are so old that I use 1/4 DIN controllers and contactors rather than solid state relays.  They are fairly robust though, got them from a liquidator, used, over 20 years ago and most are still functioning.

Share this post


Link to post
Share on other sites

1/4 DIN is definitely better for my middle-aged eyes, but 1/16 DIN is cheaper.

My first homebuilt HT oven used a relay-output controller and contactor. It was noisy and when I did some testing with a borrowed high-end controller, several thermocouples and a datalogger, it was clear that shorter cycle times gave better stability. It was also clear that radiative (over)heating was a potential issue, particularly at tempering temperatures, and that ramp/soak capability would be worth having. 

The controllers I have been using for my more recent homebuilt ovens have been either Omega CN7823 or Automation Direct SL4848VR. They seem to be the same controller with different badges and I buy whichever is cheapest at the time. They have ramp/soak capability.

The graphs were from a series of different-diameter thermocouples, intended to (loosely) approximate the heating effect at different thicknesses on the bevels.

The thermocouples were out of my box-of-bits and I knew nothing about them except that they were Mineral Insulated typeK and of different diameters. Exp had an exposed junction. I do not know whether the others had grounded or ungrounded junctions, though ungrounded seems most likely. 

GraphPIDto250deg.GIFGraphRampto250deg (1).GIF

 

The setpoint in both cases was 250 degC, 482 degF. The control thermocouple was a 6mm Mineral Insulated typeN with an insulated (ungrounded) junction and Nicrobel sheath.

The 30-min (-ish) slow ramp-to-temperature gave a peak radiative overshoot of 19 degC, 34 degF. Without the ramp, the peak radiative overshoot was 119 degC,  214 degF. 

Further testing showed that slower ramping helped and there were small errors due to thermocouple tolerances which may account for as much as 10 degC, 18 degF. 

There was also a saw-tooth variation about the setpoint once it was reached, with the pitch and amplitude directly proportional to the output cycle time. There were clear improvements from reducing the cycle time down to 5 seconds. I "think" there was an improvement between 5 seconds and 2 seconds and that I could tell which was which from the charts. Unfortunately I lost the data to a hard-drive failure and these 2 charts were all that I had backed up. 

I use a 2-second output cycle and fit an LED indicator to the control panel that is lit when the element is powered. By watching it for 2 seconds, you get a surprisingly good idea of how much of the output cycle is powered, even from 20-30 ft away. 

 

Share this post


Link to post
Share on other sites
Posted (edited)

You have dug deeper into this than I have.  Use of dataloggers is certainly helpful.  I end up sitting and watching the output parameters on the controller as it goes through ramp cycles. 

I'm not completely sure I have my PID constants optimized, but I have gone through the process of tuning to try to get them close.  As I'm sure you are aware, the thermal mass, insulating value and heating coil characteristics have a large effect on the system dynamics for a heat treat oven, and some tuning of the constants should be expected to get accurate, or even stable, operation.  My old controllers are West units (I think 2050 series controls) that are fully programmable for multiple steps of ramp and soak as well as  PID constants, but don't have auto-tune.  I've also misplaced the manual, so struggle with setting up programs.  Can't recall what cycle time I have set, but expect 2 seconds would be too short for effective operation, even with the mercury contactors I use.  I do have the system choked down somewhat, so I don't overheat and burn out the coils (as I did the first time I futzed with it after repurposing the unit from a glassblowing hot pickup oven running at a maximum of 960 deg. F to a heat treat oven that reaches 1575 deg. F when necessary.

I don't worry about initial overshoot since I typically wait till the oven achieves stability at the desired setpoint before I put the blades in.  I have enough thermal mass in the oven that the quick opening and closing of the door doesn't effect that too much, though my top access is less effective than a side door would be at higher  austenizing temperatures.

Edited by Dan Hertzson

Share this post


Link to post
Share on other sites

Sounds like I'll have a steep learnin curve on this project.....................

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×
×
  • Create New...