Kiln Control

[jcdammeyer]

A few years ago a acquired this kiln. The plan was to be able to either melt out hot wax or filament. Also heat treat castings or weldments.
I bought the parts and pieces for building this up but haven't gotten around to it yet.

GitHub - jbruce12000/kiln-controller: Turns a Raspberry Pi into an inexpensive, web-enabled kiln controller.

Turns a Raspberry Pi into an inexpensive, web-enabled kiln controller. - jbruce12000/kiln-controller

github.com

However I did want to play a bit since my kiln is 220VAC with 110VAC coils (4 of them) with two separate controllers and a master switch when the pyramid bends enough to shut off the power. The above project will require modifications to get there.
Now I do have a controller for K type thermocouples good for control to 400C. Not enough for melting aluminum. I bought it originally for my first crack at 3D printing using my JGRO CNC router.

Since I have all sorts of Arduino type boards and this teensy little display I thought I'd take a look at some of the stuff out there for the NodeMCU V3 with ESP8266 Wifi.
So this little project has been progressing over the last few evenings.

The top line is the amount of time since the start button was pressed (from power up at the moment since I haven't wired in a start button).
The bottom line is the time of day from NTP server. The software connects onto my network and goes out and gets the time of day.

This second display shows the set point controlled by a pot, the temperature read by the thermocouple and the LED shows when the heater is requested ON.

Kind of fun even if I could just buy a solution.

 

[jcdammeyer]

Here's how the Skutt C181 kiln is wired for one of the two pairs of heater coils. They put them in series to run them at half power for low. One winding is on for medium. Both windings are on for High. The bottom half runs on one of the phases and the top half runs on the other phase both relative to neutral.

The Kiln-Sitter switch has a cone inserted that will bend at the target temperature (for a kiln). So if I insert one rated for say 1600F I can run with electronic control to 1300F and if an SSR fails ON it will continue to heat until the cone bends and shuts off the kiln.

I guess I could probably put one SSR in series with the neutral wire but I think it's safer to use one for each leg. The kiln is rated at 20A and requires a 30A 220VAC breaker.

 

[Tom O]

Why use the cone when it is simple to run a thermalcoupling?

 

[jcdammeyer]

Why use the cone when it is simple to run a thermalcoupling?

Safety. The cone in the kiln-sitter is what closes the circuit two the 3 way switches on each level of element pairs. I'd have to tear apart the box and remove all that stuff. Rewire etc. If the SSRs short circuit ON then the kiln-sitter with the #6 cone will shut off power regardless of the SSRs which are in the two legs of the 220VAC coming to the kiln.

I can add an alarm with a text+email message that reports request for heat and heat is not happening.

 

[jcdammeyer]

The heat sinks arrived a day early. The mounting surface wasn't the nicest and depending on the anodizing (or paint) are also insulated against heat transfer. So onto the surface plate with sandpaper up to 400 grit. Then heat sink compound. And mount with a couple of screws.

A test with a LED + 560 Ohm resistor and into the control inputs shows that 12V results in 10mA. That's 5V across the SSR, 5.6V across the 560 Ohm resistor and finally about 1.4 across the LED. Implies 5V could drive the SSR directly to make the internal SSR LED show RED.

Next step, put a space heater in series with it and see how warm the heat sink gets.

 

[jcdammeyer]

Put that little ESP8266 module with display beside the kiln and connected one of the long K Type probes. then closed the lid and fired it up. Ran it to 205C and then switched it off. Kiln continued heating until 230C and then the temperature started dropping.

So with that in mind I built up the little SSR Interface circuit and cables so I can plug one end into the bread board to interface to the Raspberry Pi running the Kiln Control program. The other end goes to the two SSRs mounted on the heat sinks and will be wired in series with the two 220VAC legs.

The software works fine sensing the K Type thermocouple wire warmed with my fingers. I can see the PWM indicator change as it gets close to the set point. If I put an ice cube on the sensor the temperature plummets to 0. Then climbs again sitting at room temperature until I stopped the program.

Now for the major problem. The TTL driver is there for the SSRs. The two cables are there. The SSRs are sitting on the bench ready to be screwed to a temporary back plate. What's missing is the tiny little interface board with the 14 pin socket and the connectors to match the cable.

I put it somewhere. No idea where. I'd ask for help but I don't think anyone else will know where I put it.

Project is suddenly on hold since I don't really want to hand wire another one.

Sigh...

 

[Susquatch]

Project is suddenly on hold since I don't really want to hand wire another one.

I put a backhoe repair job on hold today cuz I dropped an Allen key in the grass that I needed to set the pressure relief. Banged my head on a bracket looking for it. Cursed and swore for a bit and then walked away. Tomorrow is another day.

 

[jcdammeyer]

Still haven't found it. At this point I've now laid it out as a single sided PC board. Use toner transfer method to make the PC board, drill the holes and pop in the parts.

 

[jcdammeyer]

Here's a photo of one of my mistakes.
It's been a while since I've done this so it makes sense the first time I'd get it wrong. Luckily the board is long enough that I could flip it over and drill it from the component side instead of wrongly from the copper side.

I'll have to mount a slightly larger drill bit in the Dremel, like 0.042, and enlarge a few holes so that the connectors fit a bit better. The 6 position one won't go all the way down. It's close but not quite.

The board wasn't quite wide enough so the mounting holes (should be 0.125) are too close to the edge. That extra hole in the Top RHS was already there in this small piece of PC board I had on hand. And I just stuffed in an SN7402 instead of the SN7406 to see how it fit.

I can either do toner transfer on the other side or just use a felt tip marker. I did do a video of the CNC router drilling holes. Pretty boring but I'll see if I can transfer it to youtube.

Still haven't found the finished prototype. Here I thought drilling the holes would make it show up. Clearly I'll have to do some etching and soldering first.

EDIT: Here we go. https://youtu.be/5zh-28CHdj4

 

[jcdammeyer]

The Protel99SE CAD software creates a drill file that is appropriate for the PCB fabrication houses but isn't really compatible directly with either MACH3 or LinuxCNC G-Code. So I had to do some editing. Does anyone know of a translation program that does this automatically? The .tap file is the one for MACH3. The .txt file is the original.

 

[jcdammeyer]

Jon Elson who sells PICO stuff for LinuxCNC sent me a Pascal program from the 90's that sort of did the translation. I've updated it so it works with Delphi or Lazarus. I can now automatically convert the Protel 99SE drl files over to proper G-Code for drilling.
The attached G-Code that's generated does not use the G81 drilling operation. Instead it moves, drills, retracts, repeat...
I'm going to play with this program some more and add options like G81.

 

[PaulL]

The Protel99SE CAD software creates a drill file that is appropriate for the PCB fabrication houses but isn't really compatible directly with either MACH3 or LinuxCNC G-Code. So I had to do some editing. Does anyone know of a translation program that does this automatically? The .tap file is the one for MACH3. The .txt file is the original.

That's a pretty straightforward translation. Probably take 10 minutes with AWK to write up the conversion. But now that you have something working...
I might do it just as a little puzzle.

 

[Johnwa]

Flatcam converts Eagle drill files to gcode.

 

[jcdammeyer]

It's likely the Excellon drill files are the same from all PCB tools. Jone ELson remarked that he wrote his first in Turbo Pascal back in 1995. The reason it didn't quite work right is it was meant for a different system. He later rewrote it in C (which he also sent me) but who am I to resist a challenge. Since he's never had a tool changer he didn't use the G-Code change tool command nor the G81 which I used when I manually rewrote the file for MACH3 and my CNC router.

It's possible my Altium Software even has a feature to create proper G-Code instead of Excelon. Never checked. I guess I should.
Since this system doesn't allow non-text file attachments I've added a .txt to the .pas file name. As with all Linux oriented programs it's based on 1970's technology which means it is a command line program. If it turns out Altium doesn't do G-Code I'll rewrite it to be Object Pascal based with screens and dialogs. Like this one which also started as a Jon Elson command line program.\

 

[jcdammeyer]

Oh and although old

shows how it's done for Altium Summer 09. Unfortunately that's for Roland Milling Language and I don't have one of those machines.

But there's also this video which I haven't watched yet.

 

[jcdammeyer]

Here's a photo of one of my mistakes.
It's been a while since I've done this so it makes sense the first time I'd get it wrong. Luckily the board is long enough that I could flip it over and drill it from the component side instead of wrongly from the copper side.

I'll have to mount a slightly larger drill bit in the Dremel, like 0.042, and enlarge a few holes so that the connectors fit a bit better. The 6 position one won't go all the way down. It's close but not quite.

The board wasn't quite wide enough so the mounting holes (should be 0.125) are too close to the edge. That extra hole in the Top RHS was already there in this small piece of PC board I had on hand. And I just stuffed in an SN7402 instead of the SN7406 to see how it fit.

View attachment 27565

I can either do toner transfer on the other side or just use a felt tip marker. I did do a video of the CNC router drilling holes. Pretty boring but I'll see if I can transfer it to youtube.

Still haven't found the finished prototype. Here I thought drilling the holes would make it show up. Clearly I'll have to do some etching and soldering first.

EDIT: Here we go. https://youtu.be/5zh-28CHdj4

The finished prototype showed up when I gave @David_R8 a shop tour. I'd done some minimal cleaning of mess and it had fallen on the floor. So now I have two. Time to restart Project #42 and get the Kiln working with profiles.