Geeks only - Raspberry PI controlled kiln

[neilestrick]

10 hours ago, Bill Kielb said:

Yes, the screws are brass, SS, You name it. The block terminal is the correct metal per the marking in the block. Brass screws seize up less than most especially in chromel and alumel. 

These are just brass blocks with stainless (sometimes plated) steel screws. No TC specific metals. You can use them with type K, J, N, etc. Standard on most kilns in the US.

[neilestrick]

10 hours ago, Bill Kielb said:

I didn’t know you got an SSR control. How do you like it?

It's pretty cool. I replaced my big 21 cu/ft L&L  DaVinci a couple of months ago with an L&L eQ2827-3, and I ordered it without a control panel. I then rebuilt my DaVinci freestanding panel with SSRs and a Genesis and a lot of the DaVinci parts. Came out cheaper than the stock control box. I love how quiet it is compared to the DaVinci, which had 50 amp relays and pilot relays slamming on and off. It's very strange to look in and see the elements glowing at 70% or whatever. Between the quad elements and the SSRs, it should have amazing element life.

[MarkTilles]

Bill, my kilns here in Sweden as well as my accessory thermocouples I am using with my PID controllers are S-type. I just bought 25 meters of cable from RS components, “IEC PVC insulated flat pair extension and compensating cable” ...  “type U (RCA) 7/0.2mm for R/S thermocouples”. My thermocouples look to have stainless steel screw terminal mounts. Hopefully this was the right material for me to purchase!?

[Bill Kielb]

1 hour ago, neilestrick said:

These are just brass blocks with stainless (sometimes plated) steel screws. No TC specific metals. You can use them with type K, J, N, etc. Standard on most kilns in the US.

Yuck! Send it back it’s for an  RTD mine are all k type, never have seen that but maybe never took much notice.  I have had some with stripped or seized screws though that were definitely not brass.

[Bill Kielb]

1 hour ago, neilestrick said:

It's pretty cool. I replaced my big 21 cu/ft L&L  DaVinci a couple of months ago with an L&L eQ2827-3, and I ordered it without a control panel. I then rebuilt my DaVinci freestanding panel with SSRs and a Genesis and a lot of the DaVinci parts. Came out cheaper than the stock control box. I love how quiet it is compared to the DaVinci, which had 50 amp relays and pilot relays slamming on and off. It's very strange to look in and see the elements glowing at 70% or whatever. Between the quad elements and the SSRs, it should have amazing element life.

Fantastic! You should be able to drop the controller relay cycle time  to SSR levels which ought to extend the life of the elements a bit more and smooth the control out  as well.

You may already know but Skutt actually has a design using the old 12v relays upstream of the SSRs instead of the monster contactor. Pretty cool economical idea I think.

Edited April 15, 2021 by Bill Kielb

[neilestrick]

First one was from an older Skutt. This is from a new L&L. Terminals appear to be stainless, or possibly plated. I scratched the surface and didn't see any evidence of plating, though. Non-magnetic. No specific markings for TC type. I've definitely had some old one strip out. Also seen a lot of old ones that had plated screws that were totally rusted. I always replace those, but they continue to work even with the rust, since the screws aren't carrying the signal. When they're really, really old, the brass oxidizes and needs to be cleaned up, but the kiln is usually dead by then. I just run a drill bit through the holes to get fresh metal, and hole onto those blocks as emergency replacements, although it's rare that I need to replace them.

 

[neilestrick]

4 minutes ago, Bill Kielb said:

Fantastic! You should be able to drop the controller relay cycle time  to SSR levels which ought to extend the life of the elements a bit more and smooth the control out  as well.

The Genesis has an 'SSR Mode' that cycles the relays at 500 milliseconds. Pretty cool.

[Bill Kielb]

1 hour ago, MarkTilles said:

Bill, my kilns here in Sweden as well as my accessory thermocouples I am using with my PID controllers are S-type. I just bought 25 meters of cable from RS components, “IEC PVC insulated flat pair extension and compensating cable” ...  “type U (RCA) 7/0.2mm for R/S thermocouples”. My thermocouples look to have stainless steel screw terminal mounts. Hopefully this was the right material for me to purchase!?

Looks like you bought all compatible components made to do what they should. 75 meters is a long way for  millivolt signal so if your tcouple board is noisy, worst case convert it to current loop.

[MarkTilles]

Thank you for the confirmation! It feels good to know that I’m starting to know what I’m doing :-)

And I only bought 25 m because that was the smallest rol they sold

[Bill Kielb]

2 minutes ago, neilestrick said:

The Genesis has an 'SSR Mode' that cycles the relays at 500 milliseconds. Pretty cool.

I meant to add since it is still a switch it can only go as fast as a half cycle  which you guessed it is .5 seconds. They are zero crossing as well so they always turn on at zero, rather than turn on at peak. Another cool  longevity thing.

[neilestrick]

2 minutes ago, Bill Kielb said:

I meant to add since it is still a switch it can only go as fast as a half cycle  which you guessed it is .5 seconds. They are zero crossing as well so they always turn on at zero, rather than turn on at peak. Another cool  longevity thing.

I'm excited to see what sort of element life I get. I've got customers with the same model kiln that are getting 300 firings or more from their quad elements with standard relays. Sucks that I'll have to wait a couple years to find out!

[MarkTilles]

24 minutes ago, Bill Kielb said:

Looks like you bought all compatible components made to do what they should. 75 meters is a long way for  millivolt signal so if your tcouple board is noisy, worst case convert it to current loop.

Quick question, the terminal block of the thermocouple is of course outside the kiln in a much cooler environment. Previously when I was using the wrong cable I was getting measurement drift at high temperatures when I compared the temperature on my PID  to the temperature on control panel of the kiln itself. You see, I have two thermal couples in this Kiln, the built-in and one I added. I want to have a secondary system for overheat protection and temperature comparison.

Are these discrepancies likely because I simply had the wrong cable in place?

[Bill Kielb]

30 minutes ago, MarkTilles said:

Quick question, the terminal block of the thermocouple is of course outside the kiln in a much cooler environment. Previously when I was using the wrong cable I was getting measurement drift at high temperatures when I compared the temperature on my PID  to the temperature on control panel of the kiln itself. You see, I have two thermal couples in this Kiln, the built-in and one I added. I want to have a secondary system for overheat protection and temperature comparison.

Are these discrepancies likely because I simply had the wrong cable in place?

Any temp difference can be an issue. The thermocouple itself will generate a very specific voltage at a temperature but couples formed by splices are always a concern and less predictable.. K type thermocouples depend on cold junction compensation for their accuracy so any additional couples in the line (even very small voltages)  can be an issue. Even when we design with current loop or conditioning transmitters we like to keep the conditioner as thermally isolated as practical for accuracy reasons.

The discrepancy is likely due to the wrong cable and the additional couple effect from the wrong cable. As the individual couples by the splice heat up, the ‘splice’ couples are less likely to be equal and opposite. So yes, your theory matches some of the known problems with mixing materials.

As an interesting example, for kilns with a top and bottom thermocouple , I will mount both transmitters at the same height and in a cool location to avoid minor differences in the ambient temperature between the two transmitters.

Edited April 15, 2021 by Bill Kielb

[MarkTilles]

On 4/15/2021 at 6:59 PM, Bill Kielb said:

Any temp difference can be an issue. The thermocouple itself will generate a very specific voltage at a temperature but couples formed by splices are always a concern and less predictable.. K type thermocouples depend on cold junction compensation for their accuracy so any additional couples in the line (even very small voltages)  can be an issue. Even when we design with current loop or conditioning transmitters we like to keep the conditioner as thermally isolated as practical for accuracy reasons.

The discrepancy is likely due to the wrong cable and the additional couple effect from the wrong cable. As the individual couples by the splice heat up, the ‘splice’ couples are less likely to be equal and opposite. So yes, your theory matches some of the known problems with mixing materials.

As an interesting example, for kilns with a top and bottom thermocouple , I will mount both transmitters at the same height and in a cool location to avoid minor differences in the ambient temperature between the two transmitters.

Then here’s a nitty-gritty question, should I solder in the screw terminal block, or should I solder the new compensation cable directly to the board?

[newps]

For what its worth I soldered the terminal block to the Max31856 board.

The purists will say I am working totally wrong because I have the Max31856 wired in parallel with the original analogue controller which I am using as my over temperature protection.  But I am taking the view that the modern electronics has a much higher input impedance than the analogue unit so it will hardly affect it. In practice that seemed to be the case. 

My current issue is more with the basic Pi operating system having lost some of the background files (Gevent) that I need. Possibly a corrupt memory card, so when I click on the start button the Pi doesn't recognise I have told it to start. However with house moves fixing this is low on my list of priorities

[Bill Kielb]

31 minutes ago, MarkTilles said:

Then here’s a nitty-gritty question, should I solder in the screw terminal block, or should I solder the new compensation cable directly to the board?

Solder in the screw terminal block, the cable probably won’t solder well anyway. This final junction will be fine.

Edited April 16, 2021 by Bill Kielb

[Bill Kielb]

19 minutes ago, newps said:

But I am taking the view that the modern electronics has a much higher input impedance than the analogue unit so it will hardly affect it. In practice that seemed to be the case. 

Generally works but does often have it’s issues and doubling on an input is a definite not allowed but I have seen many get away with it. You have other issues for sure to work through.

[MarkTilles]

6 minutes ago, Bill Kielb said:

Generally works but does often have it’s issues and doubling on an input is a definite not allowed but I have seen many get away with it. You have other issues for sure to work through.

However, I want to put this double pole double throw switch in the loop, since I want to have one raspberry controller to control either of my two kilns. I guess I can do some testing and see how this affects things, but this will be within the residential room where we work, not where the space is getting heated. Compensation cable all the way to the raspberry card. Problem?

[neilestrick]

As long as you're using the proper TC wire the whole way it should work.

[Bill Kielb]

2 hours ago, MarkTilles said:

However, I want to put this double pole double throw switch in the loop, since I want to have one raspberry controller to control either of my two kilns. I guess I can do some testing and see how this affects things, but this will be within the residential room where we work, not where the space is getting heated. Compensation cable all the way to the raspberry card. Problem?

 

It’s not skookum as they say but it likely will work. In other words you are not supposed to as standard practice for thermocouples.  Alpine used to do it, I don’t believe they do anymore and even in the day flipping that switch often cause full scale deflection of the giant (very high quality) analog high limit / operate switch to trip momentarily.

If you have  two thermocouples and one input and this is not a make before break switch you may need to add a bleed resistor to each to avoid the open circuit spike. In the end for the Alpine gas kilns if it influenced the temp by a degree  or  even a few degrees it made little difference since these kilns were shut down by observing firing cones.  Late in the firing used to be the rule, don’t touch that switch as we would operate as a high limit and if it spiked to set-point the kiln would drop offline.

I mention because the spike may or may not affect your high limit or the switching noise may or may not affect your stuff ........ even later on as the switch wears it will generate more or less noise. Hard to know for sure until you try it, but easy to fix should you need to.

Edited April 16, 2021 by Bill Kielb

[MarkTilles]

22 minutes ago, Bill Kielb said:

It’s not skookum as they say but it likely will work. In other words you are not supposed to as standard practice for thermocouples.  Alpine used to do it, I don’t believe they do anymore and even in the day flipping that switch often cause full scale deflection of the giant (very high quality) analog high limit / operate switch to trip momentarily.

If you have  two thermocouples and one input and this is not a make before break switch you may need to add a bleed resistor to each to avoid the open circuit spike. In the end for the Alpine gas kilns if it influenced the temp by a degree  or  even a few degrees it made little difference since these kilns were shut down by observing firing cones.  Late in the firing used to be the rule, don’t touch that switch as we would operate as a high limit and if it spiked to set-point the kiln would drop offline.

I mention because the spike may or may not affect your high limit or the switching noise may or may not affect your stuff ........ even later on as the switch wears it will generate more or less noise. Hard to know for sure until you try it, but easy to fix should you need to.

Maybe I should clarify, I have two kilns but can only run one at a time. Just one 3-phase 16A circuit in  my garage! Not planning on flipping over live to check each kiln...

[Bill Kielb]

Just now, MarkTilles said:

Maybe I should clarify, I have two kilns but can only run one at a time. Just one 3-phase 16A circuit in  my garage! Not planning on flipping over live to check each kiln...

No worries, you are aware of switch noise now and issues it can present. As to calibration, you will need to see if it affects it and by how much. If you are asking how would I have done it, probably installed k type connectors but your solution is likely fine.

 

[MarkTilles]

OK, @jbruceI FINALLY got my max31856 card and have plugged it in now. All cables connected. I set  Simulation=False and enabled 31856, but get this error on startup. Seems to be a software and not hardware issue. Any ideas anyone?

(venv) pi@Raspberry-1:~/kiln-controller $ source venv/bin/activate; ./kiln-controller.py
2021-04-17 00:05:14,369 INFO kiln-controller: Starting kiln controller
2021-04-17 00:05:14,391 INFO kiln-controller: this is a real kiln
2021-04-17 00:05:14,393 INFO oven: import MAX31856 
2021-04-17 00:05:14,396 INFO oven: init MAX31856
Traceback (most recent call last):
  File "./kiln-controller.py", line 44, in <module>
    oven = RealOven()
  File "/home/pi/kiln-controller/lib/oven.py", line 311, in __init__
    self.board = Board()
  File "/home/pi/kiln-controller/lib/oven.py", line 46, in __init__
    self.create_temp_sensor()
  File "/home/pi/kiln-controller/lib/oven.py", line 74, in create_temp_sensor
    self.temp_sensor = TempSensorReal()
  File "/home/pi/kiln-controller/lib/oven.py", line 107, in __init__
    from max31856 import MAX31856, MAX31856Error
ImportError: cannot import name 'MAX31856Error' from 'max31856' (/home/pi/kiln-controller/lib/max31856.py)
(venv) pi@Raspberry-1:~/kiln-controller $ 

Edited April 16, 2021 by MarkTilles

[Bill Kielb]

9 minutes ago, MarkTilles said:

OK, I got my Max31856 crd and have plugged it in now. I set  Simulation=False and enabled 31856, but get this error on startup. Any ideas anyone?

(venv) pi@Raspberry-1:~/kiln-controller $ source venv/bin/activate; ./kiln-controller.py
2021-04-17 00:05:14,369 INFO kiln-controller: Starting kiln controller
2021-04-17 00:05:14,391 INFO kiln-controller: this is a real kiln
2021-04-17 00:05:14,393 INFO oven: import MAX31856 
2021-04-17 00:05:14,396 INFO oven: init MAX31856
Traceback (most recent call last):
  File "./kiln-controller.py", line 44, in <module>
    oven = RealOven()
  File "/home/pi/kiln-controller/lib/oven.py", line 311, in __init__
    self.board = Board()
  File "/home/pi/kiln-controller/lib/oven.py", line 46, in __init__
    self.create_temp_sensor()
  File "/home/pi/kiln-controller/lib/oven.py", line 74, in create_temp_sensor
    self.temp_sensor = TempSensorReal()
  File "/home/pi/kiln-controller/lib/oven.py", line 107, in __init__
    from max31856 import MAX31856, MAX31856Error
ImportError: cannot import name 'MAX31856Error' from 'max31856' (/home/pi/kiln-controller/lib/max31856.py)
(venv) pi@Raspberry-1:~/kiln-controller $ 
 

I’m not a pi guy - vba, apl, FORTRAN .......... PLC ....... lots of DDC, import error fails on non capitalized file name ? Maybe as simple as that.

Edited April 16, 2021 by Bill Kielb

[MarkTilles]

28 minutes ago, Bill Kielb said:

I’m not a pi guy - vba, apl, FORTRAN .......... PLC ....... lots of DDC, import error fails on non capitalized file name ? Maybe as simple as that.

Forgot a few important wires ...  nope, no change fully wired. Don't think it's a filename problem. Will see if any of our Python friends can debug.

Edited April 16, 2021 by MarkTilles