MAX31855 problems, unstable readings, Moderators: adafruit_support_bill, adafruit

[adafruit_support_bill]

I am going to do a run with the contactor switching circuit operative (with the power to the heaters switched off)

That will take into account any inductive spikes from the contactor coils. But it won't be totally conclusive - since a large part of the EMI from a contactor is typically due to the arcing across the contacts themselves.

[ColinKing999]

Hi,

I did the run with the contactor switching but nothing went wrong. I tried to limit the current spikes from the contactor by shielding it but that didn't seem to help?

I will now try shielding the 220 V lines from the source.

We live and learn!
Colin

[petespaco]

I have been waiting to see if the thermocouple itself is any good. Shouldn't we first rule out a bad T/C junction before doing all that other stuff? I would think that making sure the T/C actually DOES work reliably above 900° would be item one.

Am I missing something here?

Pete Stanaitis
-------------------

[adafruit_support_bill]

He posted the results of the torch-test above: viewtopic.php?f=31&t=51584&start=15#p728447

[petespaco]

Thank you. Sorry that I didn't pay closer attention.

[ColinKing999]

Hi everyone,

I acquired some braided thermocouple extension cable which I installed. The results are much the same as before.

[code
09:41:44.835 -> ***** Error code: 0
09:41:44.929 -> Temp: 925
09:41:55.007 -> ***** Error code: 0
09:41:55.053 -> Temp: 925
09:42:05.162 -> Temp: 929
09:42:15.240 -> ***** Error code: 0
09:42:15.302 -> Temp: 929
09:42:25.380 -> Temp: 929
09:42:35.442 -> ***** Error code: 0
09:42:35.536 -> Temp: 929
09:42:45.613 -> Temp: 931
09:42:55.691 -> ***** Error code: 0
09:42:55.784 -> Temp: 931
09:43:05.847 -> Temp: 932
09:43:15.924 -> Temp: 933
09:43:25.986 -> Temp: 936
09:43:36.064 -> Temp: 934
09:43:46.188 -> ***** Error code: 2
09:43:46.235 -> Temp: 934
09:43:56.313 -> ***** Error code: 2
09:43:56.407 -> Temp: 934
09:44:06.484 -> ***** Error code: 0
09:44:06.531 -> Temp: 934
09:44:16.609 -> ***** Error code: 2
09:44:16.702 -> Temp: 934
09:44:26.764 -> ***** Error code: 2
09:44:26.811 -> Temp: 934
09:44:36.920 -> ***** Error code: 2
09:44:36.967 -> Temp: 934
09:44:47.044 -> Temp: 940
09:44:57.153 -> ***** Error code: 4
09:44:57.200 -> Temp: 940
09:45:07.278 -> ***** Error code: 0
09:45:07.371 -> Temp: 940
09:45:17.433 -> ***** Error code: 2
09:45:17.527 -> Temp: 940
09:45:27.589 -> ***** Error code: 2
09:45:27.636 -> Temp: 940
09:45:37.729 -> ***** Error code: 2
09:45:37.823 -> Temp: 940
09:45:47.901 -> ***** Error code: 2
09:45:47.947 -> Temp: 940
09:45:58.025 -> Temp: 943
09:46:08.130 -> Temp: 943
code]

I hope the {code] thing works (it doesn't do what I expected it to do.)

I see that the onset temperature is about 900 degrees again.

I am baffled by the Error code 0 that seems to come up around 900. (the run lasted about 90 minutes. ) It seems to me (in ignorance) that the electronics are doing something odd here - The error code 0 is all systems OK, as I understand it, and so shouldn't raise a flag.

Why would the connections seem to change at this temperature, giving the 2 and 4 codes. I touched the thermocouple termination board (where it enters the kiln) and it was barely above room temperature.)

I am baffled!

Colin

[adafruit_support_bill]

Code: Select all | TOGGLE FULL SIZE

I am baffled by the Error code 0 that seems to come up around 900. (the run lasted about 90 minutes. ) It seems to me (in ignorance) that the electronics are doing something odd here - The error code 0 is all systems OK, as I understand it, and so shouldn't raise a flag.

The error code you are printing is not the same error code that caused the NaN.

The NaN is generated during the first read:

Code: Select all | TOGGLE FULL SIZE

   Stage1TempStart = thermocouple.readCelsius();               // get ambient temp before firing
 

Then you go back and do a second read to get the error code:

Code: Select all | TOGGLE FULL SIZE

   if (isnan(Stage1TempStart))
                 {Serial.print("*****   Error code:  ");  Serial.println(thermocouple.readError());
                 Stage1TempStart = 20.;}

But by that time, the transient error condition has passed, so it returns a 0.

[ColinKing999]

Thanks for that explanation!

It is strange that it only pops up at high temperatures or after some hours of running. Why would it bring up the 2 and/or 4 codes? They must also be transient as they seem to go away?

Can I then just ignore any nan readings? Perhaps raise a flag only if a certain number appear in quick succession?

Colin

[adafruit_support_bill]

They do seem like transient errors. Can you see any correlation between the error occurrences and switching events on your contactor?

[ColinKing999]

Nope! I didn't allow the contactor to switch at all. I let it switch on and then let it run.

Colin

[adafruit_support_bill]

I think it's time to put an oscilloscope probe on the system to get an idea of the nature of the interference.

In the mean-time, you can probably ignore transient errors. Only raise the alarm if you get a string of them in a row.

[ColinKing999]

Hi

I have done another run with indicator cones in the kiln. These devices are cone shaped and bend over at a specified temperature. I noticed two things that are odd. We had put a 1000 degree cone and a 1050 degree cone in. When the Tc was indicating 960 or so, it seemed that both cones had fallen over. Could it be that the TC board is starting to misread when the Isnans start to pop up at each reading? Could it be loosing some reference?

The other thing that I have noticed in the last few runs is that the inception temperature for the isnans seems to be getting lower. In this run, they started at 820 degrees.

The last few readings (10 seconds apart) seem to have peaked out at about 920. When I was using my other controller Toho Electronics TTM-J4), we could get up to 1200 degrees at a push. This leads me to wonder if the TC board isn't under-reading once the isnans begin to pop up?

Colin

[adafruit_support_bill]

Pyrometric cones are slow, but reliable indicators of temperature. Though temps can vary depending on the location within the kiln. So it can be misleading to compare measurements from different locations.

Thermocouples in general are not linear at extreme temperatures. And a K-type is probably starting to roll-off a bit by 900 degrees. There is some code here for linearizing the output of the MAX31855:
https://learn.adafruit.com/calibrating- ... earization

[ColinKing999]

Hi Bill,

I tried the linearisation software. The difference between the raw data and the corrected data is about 2 - 5 degrees. The inception temperature for isnans is now down to about 820 degrees.

Once again, when I switched off, the isnans stopped too! But if the cause is electric, why do they not appear throughout the range?

I am contemplating injecting a dc voltage from 0 to about 60 mV directly into the board.

Colin

[petespaco]

Did Whizzzz ever get his problem solved?