Homeroasters - Home Roasting Coffee Community » ALL ABOUT ROASTERS » Popcorn Popper roasting
Who is here? 1 guest(s)
 Print Thread
Magnitude of temp change with probe movement
TallPaul
Top  #1 Print Post
Posted on 06/24/2018 7:50 PM
Hello All

I have been roasting for several months, and recently using modified hot air popcorn poppers. I have noticed in many threads that the placement location for the thermocouple is important to the temperature reading that one will get. My question is this ? if you have a thermocouple simply sitting on top of the bean mass when you start the roast, and as the roast increases in volume and covers the probe, can you get a very large rise in apparent temperature? I ask because as my probe becomes more and more submersed it appears to be in the middle of airflow coming up from the bottom of the chamber and my temperature seems to rise above a level consistent with the degree of crack going on. My probe was reading as high as 240?C today, and I?m still fairly certain that I wasn?t all the way through first crack. Either that, or my first crack is rolling right into my second crack. But I barely have any oil on the outside of the beans, so I?m fairly certain I did not get into second crack, nor do I have any smoke during the roast.

I think that I definitely have a problem with placement of the probe and that the temperature is going up substantially simply because it is effectively lower beneath the surface of the beans than it was at the beginning. I was wondering if anyone on the forum had noticed such a thing. I was having a hard time iggnoring the 240 C and letting my temperature continue to rise so that I could hit second crack. I find that my lack of temp consistency continues to make me shut the roast off probably at a time before I want to do so - because the Temp seems too high even though I doubt it is.

Most of the posts I have seen would indicate that a consistent location for the probe is probably more important. Before I invest in drilling through the side of my west bend poppery I, I did wonder if people found that if you were below a certain height of bean before you started, do you experience more overall consistency - as compared to starting with the probe right at the bean surface and becoming submersed through the roast. Thank you. Paul
 
renatoa
Top  #2 Print Post
Posted on 06/25/2018 2:14 AM
It is very important to know how hot air comes into the roast chamber: circular bottom openings that force creation of a vortex near the walls, or perforated center of bottom, with a small funnel.
The hot air path is different for the two scenarios, and central probe placement is good for the first case only. For the second case probe should be placed near the wall, out of central column hot air path.

For both cases, as the beans roast they become lighter, and start to float on top of an increasing height air column, so your probe measure more hot air than beans, whatever is the location you place it.
You can raise the probe, but the real cure, imo, seems to be lowering the airflow to have the same agitation as the beginning, i.e. RC full with beans until bottom.
Edited by renatoa on 06/25/2018 7:22 AM
 
ChicagoJohn
Top  #3 Print Post
Posted on 06/25/2018 5:04 AM
If the air flow in your popper is vertical through a central port in the bottom, I'd suggest drilling a hole through the side of the roasting chamber and placing your probe about 1/3 the distance from the perforated air inlet at the bottom and the top of your green bean mass and also off center with the tip about 6 mm from the edge. Also, use some silicone tubing or silicone RTV to insulate the thermocouple wire from the metal sides of the roasting chamber. Thermocouples generate EMF from the hottest point along the circuit, not necessarily the tip, so since the metal sides can heat up faster than the bean mass, you want to insulate that from the wires to some extent.

In this type of roaster, the beans are lifted in the center and then move out to the edges where they fall back down along the sides, and placing the thermocouple like this will put it more where they are coming back down to re-enter the upward air column.

From that point, the temperature data should not be considered a direct measurement of bean temperature but rather a relative temperature index that can be used to gauge a repeatable process, not to compare with other roaster types and configurations. In my air-roasters, popper mods or the larger one I built, I am typically seeing first crack at around 380F / 190C and for a light - dark range I'm starting cooling at 405 - 425F / 207 - 218C. When you get up around 250C /480F, you're in the auto ignition range for many organic materials. Fire bad.
Edited by ChicagoJohn on 06/25/2018 5:17 AM
So many beans; so little time....
 
JackH
Top  #4 Print Post
Posted on 06/25/2018 3:27 PM
Excellent post ChicagoJohn! Repeatable is the key.
---Jack

KKTO Roaster.
 
TallPaul
Top  #5 Print Post
Posted on 06/25/2018 5:16 PM
thanks for all the comments thus far. I have two roasters, and both are using vortex flow. One is a WBP 1, and the other is a Presto Poplite 1440 W model where I replaced the RC with a Wearever Popcorn Pumper RC (vortex flow with dimple in the bottom). So both have spiral flow.

As the beans roast and get lighter (as pointed out by Renatoa), I've noticed that the airflow does come out the center more with beans popping up through the center. In this way, it does look more like a center airflow model at this point. I have turned down the airflow some, but I keep being concerned that at the lower airflows, I have poorer agitation leading to less homogeneous roasts. That, or I have some of these "quakers" I've read about. Regardless, yes, I could turn down my airflow and try that. But I'll still have my probe at least an inch deeper than the start.

I'm going to bit the bullet and pick a location for the probe and "gasket it in," as suggested by ChicagoJohn. I've seen various placement suggestions on the internet and I'll pick one that seems reasonable unless someone has the definitive placement information here (some of the links on this forum from years past aren't active anymore, so I'm missing some good historical input, likely). I suppose consistency is important for me. And John, yeah, I should have realized it isn't possible for the bean internal T to get that high that fast and I'm just reading the air mass. Looking at the beans a day later, less than 5% of them have any oil on the outside, so they are barely Full City. Definitely were not 250 ?C!!

thanks for all the suggestions so far. I appreciate them and enjoy this forum. Paul
 
TallPaul
Top  #6 Print Post
Posted on 06/27/2018 8:48 PM
Soooo... I added a thermocouple (k type, bead tip, woven fabric covering cheap type) for both ET And BT. I went through the side of the WBP1 roast chamber for the BT. For the ET, I drilled through the heat housing and the ET thermocouple tip is touching the vanes of the RC from the heating side. Pretty much like I?ve seen done in pictures provided here. I used a high temp epoxy to glue them in. I also secured the probe leads to the Bakelite chimney section with a hoseclamp to act as a strain relief. I was rather pleased with it - until I tested it on a roast. BOTH probes give T readings that bounce back and forth 10 to 30 degrees Every couple of seconds. It might read 110, then 92, then 125. Directionally they were correct over 10 minutes of roasting, so the temp did rise. As it got hotter the ET was sometimes lower than the BT, although that could be accounted for by the bouncing. I have 3 theories:

1) the epoxy ruins the probes- seems unlikely
2) the hoseclamp causes some sort of interference that makes the T bounce - I don?t know enough to know if that is plausible, but others have used the hoseclamp, so I doubt it
3) I damaged the thermostat in disaseembly and reassembly and it is cutting on and off, allowing for an overall rise, but not a steady one. I had small bounces before I turned on the heat (like 17 to 26) so I question the thermostat idea as well

Any ideas here on those? Or other theories? Thanks!

Paul
 
ChicagoJohn
Top  #7 Print Post
Posted on 06/28/2018 5:26 PM

Quote

TallPaul wrote:

Soooo... I added a thermocouple (k type, bead tip, woven fabric covering cheap type) for both ET And BT. I went through the side of the WBP1 roast chamber for the BT. For the ET, I drilled through the heat housing and the ET thermocouple tip is touching the vanes of the RC from the heating side. Pretty much like I?ve seen done in pictures provided here. I used a high temp epoxy to glue them in. I also secured the probe leads to the Bakelite chimney section with a hoseclamp to act as a strain relief. I was rather pleased with it - until I tested it on a roast. BOTH probes give T readings that bounce back and forth 10 to 30 degrees Every couple of seconds. It might read 110, then 92, then 125. Directionally they were correct over 10 minutes of roasting, so the temp did rise. As it got hotter the ET was sometimes lower than the BT, although that could be accounted for by the bouncing. I have 3 theories:

1) the epoxy ruins the probes- seems unlikely
2) the hoseclamp causes some sort of interference that makes the T bounce - I don?t know enough to know if that is plausible, but others have used the hoseclamp, so I doubt it
3) I damaged the thermostat in disaseembly and reassembly and it is cutting on and off, allowing for an overall rise, but not a steady one. I had small bounces before I turned on the heat (like 17 to 26) so I question the thermostat idea as well

Any ideas here on those? Or other theories? Thanks!

Paul


If you have a multimeter that does mV, put that on the leads and see if you seeing the voltage jump around. If it's stable, then it could be your pyrometer / temperature device. It does cold junction compensation and maybe a low battery could act like that.

If the voltage isn't stable with the multimeter, see if it is stable with the motor shut off. You could be seeing some induced voltages in the thermocouple wires perhaps caused by the blower motor. If that's what it is, then you could use a shielded T/C but I think you could also try signal conditions by tack soldering a low value ceramic capacitor across the leads.

I've seen erratic T/C readings arising from both types of cause but not when I was working with popcorn poppers. I'd start by verifying the device you're plugging the T/C into first.
So many beans; so little time....
 
TallPaul
Top  #8 Print Post
Posted on 06/28/2018 8:54 PM
ChicagoJohn

Thank you for your reply. What I found was curious. I do have a MM that reads in mV. But I had trouble getting good readings. That didn?t matter though, as the temperature readings under different conditions with different output devices shed light on the matter.

The readout was relatively stable with the motor off. With the motor on, the readings fluctuated. I changed the battery on my T meter just in case. Still fluctuated, but not as much as Sunday evening when this first happened. I was about to assume I was picking up voltage from the motor and would have to remove the probes and insulate. But, based on your statement of not seeing this much in hot air poppers, I went in and got a different T meter to see what happened. The T reading was stable with this meter, even wth the motor on! So, one meter seems susceptible to motor voltage and the other isn?t. Ok, so I can buy another meter. That?s easy. But it got stranger. I put my soup can chimney onto the Bakelite (I tried to attach pictures below). When I do this, the original meter bounces even more, and rises so much that the BT is higher than the ET. When you take the can off, it drops back down, but still bounces. The second meter seemed indifferent to the can presence. Hopefully my annotated pictures come through with the post. Very strange. But, i will buy another of the simple one channel meter and hope my mods are complete!!

Paul
 
TallPaul
Top  #9 Print Post
Posted on 06/28/2018 9:38 PM
I think the files I tried to attach were too large. I?ve tried to attach files all under 250 KB each now. Paul
 
ChicagoJohn
Top  #10 Print Post
Posted on 06/29/2018 11:24 AM
I think you're getting EMI/RFI effect within one of your meters. The can is probably acting as an antenna and that may be worse if it's iron / magnetic. I have some cheap K T/C meters and haven't seen this phenomenon, and I've been using MAX6675 temperature modules for data logging / control purposes and have not see it there either. So I'd guess it's some design issue with the meter in question -- shielding issues, inadequate signal conditioning, or maybe a case ground isn't connected etc. Who knows. But good detective work on your part. Sure sounds like you've identified the source of the issue.
So many beans; so little time....
 
Jump to Forum: