Geeks only - Raspberry PI controlled kiln

[MarkTilles]

I know I keep going around in circles and am not directly heeding your advice Bill, but please don’t get me wrong, that’s not out of disrespect;

Yet even without the switch I have exactly the same values, so:  between the new pi and its directly-coupled max31856, versus the Bentrup … what are the likelihood one versus the other is 50°+ off? What are the most likely causes large measurement discrepancies like this would show up in this case?  The controller and very likely the oven as well was manufactured in the beginning of the 1990s. 

[Bill Kielb]

Good meter readings and the s thermocouple chart are likely the most accurate thing you have. It’s too big of a spread but starting out, checking the difference with ice water and boiling water might have given you some idea of which is more correct at low temperatures or a magnitude of the difference. The Bentrup has functioned well for firing and is the conservative reading thus far which means I would err towards its reading for offsets given no other choice.

Both require setup offsets, neither are likely certified tolerance items.

Edited April 26, 2021 by Bill Kielb

[MarkTilles]

Bill, I don 't have acess to a known good and accurate multimeter just now, but does the information below lend favor to the Pi being more accurate, excepting the possibility the S-thermocouple or compensation cable is bad/incorrect? (I described the cable to you and you replied was positive), The cables are under 15 feet long.

I have two different PID controllers I bought online from Banggood. One BERM, one REX. Both set to S-type. 
I checked the temperature of the kiln with all four devices within 2 minutes of eachother,:

Kiln Bentrup:           491 C
Pi w/max31856     551 C
Berm                            557 C
Rex C100                  560 C (but when temp was read its SC was set to -1, I forgot to set to 0 first)

This gives me the feeling it's the OEM Bentrup that is not displaying correctly, and maybe we are overfiring just so much our work is not being obviously damaged?
 

Edited April 26, 2021 by MarkTilles

[Hulk]

Correlation of pyrometer readings to cones - that may be important for any/every control system?

I'm still watching my one pyrometer and flipping three zone switches - low, medium, high; hence, I am the "controller."

Earlier on, was donning kiln glasses and looking at cones through the peeps. Now that I have some firings under me belt, I look at cones afterward, as I'm confident in the pyrometer, wall clock, and notes to guide the "controller."

[Bill Kielb]

1 hour ago, MarkTilles said:

Kiln Bentrup:           491 C
Pi w/max31856     551 C
Berm                            557 C
Rex C100                  560 C (but when temp was read its SC was set to -1, I forgot to set to 0 first)

Funny, see this all the time actually. Why not take these up to 1000c (just asking). Hard to understand these from your previous data as your PI differential seemed to be growing non linearly:

 

Time (min)/Kiln (C)/Pi (C)/deltaT (Pi-Kiln)

0/14/11/-3

75/90/103/+13

210/223/265/+42

285/300/350/+50

My thinking is the Berm and  Rex C are the newest so their values stand a chance of being correct and they are close enough to the PI board so an offset in this range will be a conservative choice.. so going off the PI, +10 - bentrup, (Hi limit)  -6 Berm,  -9 Rex C 100 (something on that order to get them all tracking similarly)

They arę all fairly close I would test them to a higher temperature to make sure no surprises and confirm with cones. Confirming with cones by the numbers would be important so in the last 100c of the firing you need to go the prescribed rate to be meaningful. Check that the kiln did actually go that rate, many kilns for various reasons lose the ability to do so, so your firing record becomes important.

As time goes on if you always fire your last segment at the appropriate rate and use cones you might slide these up or down a few degrees. If you don’t fire with the correct rate in the last segment then it will be impossible to get a feel for the actual heatwork.

Sorry for your difficulty, but this is real world, kiln controllers are accurate enough for their use and firing clay is fairly tolerant, but real world accuracy, not so much when processes cannot tolerate minor differences in temperature.

Using the PI value seems reasonably conservative after offsetting your high limit to match.

Now is a real good time to see how your SSR cooling is working BTW.

Edited April 26, 2021 by Bill Kielb

[MarkTilles]

Yes we are cooking up to 960 on the kiln Bentrup controller right now, those numbers were just during the early process, so I will have more numbers soon.

[Bill Kielb]

Maybe dial in those offsets now and see how they hold up

[MarkTilles]

50 minutes ago, Bill Kielb said:

Maybe dial in those offsets now and see how they hold up

Will do! Offset of Pi versus Bentrup at 960 was about 30C (990 pi versus 960 Bentrup). So the behavior of the Bentrup isn’t as consistent whereas the other devices are much more so with eachother. 
 

Will run a 1260 eventually and see where that lies. 

Edited April 26, 2021 by MarkTilles

[newps]

Its interesting to hear your comments about thermocouple accuracy, even if the Pi results are a few degrees out they still allow me better control.

With my original controller, you set the max temperature with the dial switch that only gave 20 degree centigrade divisions, and regulated the heating with a 4 position energy regulator. There was no way of finding the current temperature other than backing the controller off until you heard the controller click off. 

Yes I could have got a new controller, but the Pi route allowed me to cheaply and quickly work out if the kiln was worth keeping.

[MarkTilles]

Yes without this new Pi-based control system I have to practically sit next to my large Chemetex kiln's controller (VERY old, analog with tons of rectangular LEDs indicating the current temp) and continually adjust its %-effect-dials to keep the heating rate in the range I want. Pain in the ass!   And my small kiln controller, the Bentrup, well it works so-so but only has a pair of settings one can change.

Edited April 28, 2021 by MarkTilles

[leof]

On 4/23/2021 at 2:57 PM, MarkTilles said:

Can any of you out there point me in the right direction to be able to look at (and maybe even tweak) the web interface? I can't figure out where it is coming from and how it interfaces with oven.py. First off I want to see what the other status symbols in the menu represent ...  Thanks in advance

As far as I can tell from the documentation the icons top-right are driven by specific GPIO's that are used in the original project picoReflow. If you don't use these GPIO's, the icons will not show.

From the picoReflow website:
 

[MarkTilles]

1 minute ago, leof said:

As far as I can tell from the documentation the icons top-right are driven by specific GPIO's that are used in the original project picoReflow. If you don't use these GPIO's, the icons will not show.

From the picoReflow website:
 

Yes thanks, I posed that question to Jason (in a conversation off-line), who manages the code for this branch, and he said the same thing. He has also just updated the codebase with my corrections for max31856, though not all were needed.

[leof]

Last week I fired my kiln for the first time with this kiln-controller software. It was a bit of a disappointment. At first things went smooth. But later on I discovered i had set the last ramp to steep: my kiln couldn't heat up fast enough. This didn't worry me much because the config.py sais:

# If the kiln cannot heat or cool fast enough and is off by more than 
# kiln_must_catch_up_max_error  the entire schedule is shifted until
# the desired temperature is reached. If your kiln cannot attain the 
# wanted temperature, the schedule will run forever.

However, it didn't work that way: the firing stopped after some time without waiting for the oven. I then used my own Windows IoT program to finish the job. I don't know what went wrong.

[MarkTilles]

5 minutes ago, leof said:

Last week I fired my kiln for the first time with this kiln-controller software. It was a bit of a disappointment. At first things went smooth. But later on I discovered i had set the last ramp to steep: my kiln couldn't heat up fast enough. This didn't worry me much because the config.py sais:

# If the kiln cannot heat or cool fast enough and is off by more than 
# kiln_must_catch_up_max_error  the entire schedule is shifted until
# the desired temperature is reached. If your kiln cannot attain the 
# wanted temperature, the schedule will run forever.

However, it didn't work that way: the firing stopped after some time without waiting for the oven. I then used my own Windows IoT program to finish the job. I don't know what went wrong.

That’s odd, I haven’t used it live yet but when I have created a test curve and the oven isn’t tied in but the thermocouple is, the countdown timer in the GUI does pause when this occurs. It paused forever for me in this “test“ scenario, as stated in the config file.

Were  you monitoring the GUI during the process?

@jbruce Jason, might you have any suggestions for Leof?

Edited April 29, 2021 by MarkTilles

[leof]

3 minutes ago, MarkTilles said:

Were  you monitoring the GUI during the process?

I use a Pi with original 7" display and monitor the process from a local browser.

[MarkTilles]

Just now, leof said:

I use a Pi with original 7" display and monitor the process from a local browser.

So for you the countdown timer that shows how much time is estimated remaining in the process continued moving even though the oven couldn’t keep up with the curve?

[leof]

I'm not sure about the count down timer. Because the firing stopped at about 900C (while the target temp was set to 940C) I was a bit of in a panic and wanted to resume as soon as possible. I had to switch the Pi off, change SD-card and get all things going again.

[MarkTilles]

1 hour ago, leof said:

I'm not sure about the count down timer. Because the firing stopped at about 900C (while the target temp was set to 940C) I was a bit of in a panic and wanted to resume as soon as possible. I had to switch the Pi off, change SD-card and get all things going again.

You know, I was thinking, if somewhere in the config file you might have a typographical error, it might be that the kiln engine has miss-read that variable setting of yours. And maybe others?  I would first take a close look for this.

Next you might want to test again without the heater relay/SSRs attached, and see if you can re-create the problem.

You might then copy in an unmodified version of the config file, and then make your changes, and test again and see if the problem is gone.

[leof]

As far as I can tell there are no parameters in config.py apart from the PID parameters. I tested these as described here. I left them at these values as they tested OK.

[MarkTilles]

My first Raspberry Pi/max31856/S-thermocoupe/1260C stoneware burn is in progress, and it's going flawlessly!  

 

[newps]

Can we have some feedback on how your 1260C stoneware burn has gone

[MarkTilles]

Yes! I haven’t opened the oven yet, but will do that tonight.

The system operated very well, pretty much exactly as it is designed.  I had the catch-up setting turned on to 10°, and my oven was going very slowly toward the end only about 80 to 90 C per hour, so because the catch up was on I didn’t get 1260 in the oven I only got 1250 because the system shuts off at the target temperature nevertheless, it apparently doesn’t wait for the sensor temperature. So now I know I’m going to change that catch up setting to 1° instead of 10. I don’t know if it can be changed to zero?

At that point I tried to restart the system with a new configuration to get that last 10°, but for some reason I was unable to start it.  It showed in the GUI Interface that the system restarted, but nothing was showing up in the logs nor did the red heating lamp start to blink.  I had to shut down and  restart the service get the system to restart, I don’t know why. But by that point I had lost so much heat I decided it was good enough at 1250 for the stoneware firing without glaze. 

[Bill Kielb]

17 minutes ago, MarkTilles said:

But by that point I had lost so much heat I decided it was good enough at 1250 for the stoneware firing without glaze. 

Did you add cones to the firing and  you using a final segment firing speed that corresponds to the Orton chart to hit a particular amount of heat work?

Edited May 8, 2021 by Bill Kielb

[MarkTilles]

No, we are newbie amateurs, but we’ve been getting satisfactory results with our work nevertheless.  We have fired many times in the past with this oven with the same clay type, and I’m confident we reach the same temperatures because I had compared top temperature is between the system earlier. But heat work, that’s new for us and something we will be looking into, especially to be able to create good, effective heating curves with the oven. I guess that’s the idea right? Making sure that the proper heat work is being applied to the material.

[Bill Kielb]

43 minutes ago, MarkTilles said:

I guess that’s the idea right? Making sure that the proper heat work is being applied to the material.

Yes, it’s fairly simple actually. Generally glazes and clay contain silica and alumina of which do not melt below about 3000f (1650c) so fluxes are added to get them to melt at a lower temperature often referred to as a eutectic point. Anyway this reaction is more dependent on time and temperature, sort of like baking a cake where the recipe says 200c for 15 minutes.

Time is an important element to get it baked correctly. Cones are made of glaze, so when a cone bends a certain amount of work has been done and we are sure it has received the right amount of heat work or has been baked correctly if you will.

It sounds complicated, but not really. For controllers to hit the right cone (which ensures the glaze and clay has received the right amount of heatwork) the last 100c of the firing is completed at a certain rate.

So in the chart below if I want to fire to cone 6 and my kiln will do 60 c per hour at the end, then I would pick 1222c from the chart, subtract 100c and my final segment would be programmed: 1122c @ 60c per hour until 1222c..

This is the easiest way to approximate a proper amount of heat work with a controller, peak temperature does not really have a bunch of meaning actually. Sort of like baking our cake, just because I hit  200c, it may or may not be finished to perfection.

Anyway, it’s easy to do and helps ensure the right amount of heat work has been done which is important to fully bake the clay and glaze. Unlike our cake, soaking at one temperature does not have the same effect, we use final rate in the last 100c to get a specific amount of heatwork.

Most kilns struggle to do 100c per hour at high temp, so 60c is usual and in the center column. Also going this moderate speed helps even out the firing through the kiln.

@MarkTilles sorry, just got all my rates corrected above, final edit.

Edited May 8, 2021 by Bill Kielb