MAX31856 Roast Thermometer

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
After playing with the calculations, I am getting down to some experiments and thought it would be best to start a thread on topic.

First project is a model for a solid roast that I can heat at a controlled rate. I wanted a material that is about the density of water, but would not have convection currents. This is my first attempt:

upload_2018-12-28_20-29-7.png

The wood is cedar and the heating pad is a Minco HR5470 that I got a few years back as a freebie. It is about 3" in diameter. My concern is that cedar is pretty light and a good insulator. I will try it anyway. It is easy to put additional TC holes closer to the edge, if needed.

Thinking that theater artists might have developed inexpensive alternatives, I found a link that used aerated plaster of Paris. The artist simply used a high foaming detergent in the water and rapid mixing. Does anyone have experience with models for meat or foamed plaster? I want it to be cheap, so RTV rubbers are probably out of the question.

For this phase of testing, I do not need to go to very high temperatures.

Regards, John
 

DNA Robotics

Joined Jun 13, 2014
647
I'm not sure if this is what you are asking for but I have used it to make light weight concrete. It would probably work with plaster. You can get it at a garden shop. Walmart?

"Vermiculite is a soft,spongy material made from super-heating mica. Perlite is a hard, highly porous material made by super-heating volcanic glass. ...Vermiculite affords slightly less aeration than perlite. For plants that thrive growing in a more-wet soil,vermiculite would do well as a potting soil additive."
 

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
I thought of vermiculite. It used to be used for packaging chemicals and such, but I didn't know where to get it. Garden shop is a good idea. Will check that out today.
 

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
UPDATE
I have made some progress. Measuring temp was done long ago. I can now measure the three functions of interest to me: temperature, rate of change (+ and -), and time to a target temperature.

The cedar pig worked good for trouble shooting. It heats and cools fast and was not very realistic in that respect. The vermiculite plug was more realistic compared to pork, but making that piece of artificial pig was an experience in itself. My prior experience with vermiculite was as packing for chemicals. It was used as an absorbent in case of breakage. None ever broke. I am not a gardener and had no idea of how absorbent (hydrophilic) it was. I mixed up some plaster of Paris according to the recipe, and it was quite liquid. I added a little more to the consistency of thin ketchup, then added vermiculite. What a near disaster. It solidified almost instantly -- I should have known. I was able to stuff about half of it into the mold and get rid of obvious entrained air before it became too hard to work. I ended up with a plug of about 14 oz by volume and 16 oz by weight -- pretty close to target. The vermiculite plug worked a little better, but even after coating with a few layers of brushing lacquer, I still got splinters from the mica particles. In brief, the vermiculite plug worked to the extent that it did not heat as rapidly, but the heating curve was sigmoid shaped as its temperature got close to the heating pad temperature, as expected. That is unlike my experience with meat for which the rate increases near the end of roasting. Another plug has been made with crushed walnut shells and plaster. We will see how that works in a pre-heated oven to 350°F. Or, maybe it's time for a pork roast.

I wanted to share an image from today of the vermiculite plug heating. The target temperature is 135°F. If one goes through the math, the 5:09 is spot on for linear heating curve from that current temperature. Zeros and negative rates have been addressed. I need to add a small routine for the finish.

Here's a photo. The vermiculite plug can be seen in the background.

show 3 functions 01.08.jpg

John
 

wayneh

Joined Sep 9, 2010
17,496
I’m not sure what you hope to accomplish but one factor that matters for meat and food in general is moisture. Internal heating depends heavily on the evaporation of water at the surface and the sizable quantity of heat it takes to make that phase change.
 

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
I didn't want to get too hypothetical and deleted a comment to that effect. Denaturation of proteins (and various reactions), moisture content, and the much higher "ambient" temperature of an oven (e.g., 350°F versus about 150° to 170°F for my heating pad) probably affect the heating curve. My gut tells me that the curve observed with the model was mostly due to the temperature differential.

The display and calculations are working well, so this afternoon, I pulled out a piece of pig from my freezer to test. Dinner at 8, Thursday. :)
 

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
Today's experiment didn't turn out too bad despite some screw ups at the start. Here's the raw data.
Columns are time, temperature (°F), rate (°F/min), and estimated finish (155°F) time:

+5 54.2 2.8 :36
+10 68.1 2.9 :30
+15 81.9 2.7 :27
+20 96.3 2.8 :21
+25 107.7 2.3 :20
+30 119.1 2.1 :17
+35 129.3 1.9 :13
+40 138.5 1.7 :09
+45 146.5 1.7 :05
+50 154.1 1.4 :00
+50:42 155.0 1.4 :00

Removed from oven at 155°F,temperature rose to 169.0 (rate=0.6)@ +10 min additional.

Here's the graph:
upload_2019-1-10_19-48-38.png

I used Xuru.org to do several regressions (linear, log, exponential, polynomial). The best fit was polynomial (r^2 = 0.998). Linear was second with r^2 = 0.988 . Here are those parameters:


upload_2019-1-10_19-47-44.png

And now for the best part, just out of the oven:


upload_2019-1-10_19-52-0.png

And carved:

upload_2019-1-10_19-53-25.png

Biggest mistake was not letting the roast warm to about 40° to 50°F before cooking. It was at 30°F, which tends to give a less juicy result. We had our first plowable snow, and taking care of that delayed my start. Still tasted good.

John
 
Last edited:

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
One of the things I learned yesterday was that a total elapsed time would be helpful. I pulled out a piece of code @NorthGuy posted here: https://forum.allaboutcircuits.com/threads/ds3231-rtc-enhancements.124008/#post-998336 . Made a few changes to fit my needs and decided to share it. I extended it to fit hours too. My seconds count by 2's so I added an "incf min_sec" before the seconds/minutes/hours count routine.

In practice, call "Timer" then call "PrintHMS." Put232 is just a routine to send 1-wire serial data to my display.

Code:
;*******************************************************************************
; TOTAL ELAPSED TIME 01/11/19
;*******************************************************************************
;The following code modified from NorthGuy counts by 1's from zero
;seconds to 60 hours, and then resets (reset not tested).
;Source: https://forum.allaboutcircuits.com/threads/ds3231-rtc-enhancements.124008/#post-998336
;Start
     clrf      min_sec
     clrf      min_sec+1
     clrf      min_sec+2
;For SmartRoast, since seconds increases by two, add 1 additional
;second with each call.
Timer                    ;call "Timer" from main w/ each increment
      
     movlw     min_sec   ;
     movwf     FSR0      ;
     incf      min_sec,f ;fixes counting by 2's seconds
;Following 2 steps used for testing. In use, call and return will
;be from main program.  
;     call      Countup   ;Delete when called from program wanting
;     bra       $-2       ;only to count and increment seconds
Countup                  ;at 2 second increment start this:
     movlw     167       ;0xA7
     addwf     INDF0,f   ;min_sec,f
     movlw     166       ;0xA6
     btfsc     STATUS,1
     movlw     160       ;0xA0
     btfss     STATUS,0
     subwf     INDF0,f   ;min_sec,f
     btfsc     STATUS,2    
     bra       BumpFSR   ;Err
     movlw     min_sec
     movwf     FSR0
     return              ;about 11 Tcy to here from Countup
PrintHMS                 ;call from Main
     movlw     0xD4
     call      Put232    ;set print to line 3,0
     movlw     3
     movwf     count
     addfsr    FSR0,2
_PrintHMS  
     swapf     INDF0,w
     andlw     0x0F
     addlw     0x30
     call      Put232
     moviw     FSR0--
     andlw     0x0F
     addlw     0x30
     call      Put232
     decfsz    count  
     bra       PutColon   ;avoids colon after seconds  
     return
PutColon
     movlw     ':'
     call      Put232
     bra       _PrintHMS

BumpFSR
     incf      FSR0,f
     movf      FSR0,w
     xorlw     min_sec+3
     btfss     STATUS,2
     bra       Countup    ;FSR0 <= min_sec+2
     bra       Restart
Restart                   ;times up, reset                  
     bra       Timer
The code has been tested up though 1:30:00 in real life and though rollover at 59:59:59 in simulation. It has not been fully cleaned up.

Regards, John
 
Last edited:

djsfantasi

Joined Apr 11, 2010
9,156
I have experience in using Vermiculate and plaster. I have used it in theater as well as model railroading.

I know you’re beyond this stage, but did you add water to the plaster or the plaster to the water?

There’s also a way to slow down the setting time. Vinegar!

I prepare for mixing by filling a large, flexible mixing bowl with the water. I then add a shot of vinegar.

In another container, I premix the dry plaster and Vermiculite.

Then, slowly pour the dry mix into the water-vinegar. Mix thoroughly. Get some latex gloves and mix by hand.

You’ll have about 20 minutes. Longer than that, you’re restricted by the chemical process.

Setting time depends on the type of plaster you have, it’s age and how it was stored. YMMV
 

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
I have experience in using Vermiculate and plaster. I have used it in theater as well as model railroading.
I know you’re beyond this stage, but did you add water to the plaster or the plaster to the water?
Plaster to water

There’s also a way to slow down the setting time. Vinegar!
I had read that before, but didn't realize I needed it until too late. My prior experience with plaster was making molds without any fillers. Did use quite cold water. If there's a next time, I will probably pre-mix the plaster and vermiculite. Crushed walnut shells worked pretty good too. They are heavier than vermiculite, and I used a little too much, so the plaster was weaker, but still worked.

John
 
Top