LCD temp readings not correlating with log

Tell us about your experience, ask if you're unsure of something, let us know if you have a problem.
Post Reply
lukehk
Posts: 14
Joined: Wed 05 Aug, 2020 4:02 pm

LCD temp readings not correlating with log

#1664

Post by lukehk »

Twice now i have followed the roast on the LCD screen and thought great. Its within 1 or 2 degree of the profile. I have the computer by its side and click on the profile every 30 seconds to track where it should be. When i uploaded the actual log there was a 4-7 degree difference to what i recorded from the screen and what was on the log. Any ideas? I'm worried im adjusting to an incorrect log?
lukehk
Posts: 14
Joined: Wed 05 Aug, 2020 4:02 pm

Re: LCD temp readings not correlating with log

#1665

Post by lukehk »

So I checked it just now by photographing the LCD and comparing it to the log.
LCD at 30 sec = 97.9 log = 109.5
LCD at 1min = 141, log = 145.3
LCD at 1min 130 = 157, log = 158.4
Gets better after this but as I am trying to change the power setting to fit the profile the changes dont seem to be having much of an impact on the log and I could be changing it incorrectly.

Have I misunderstood and the LCD is actually reading what the profile wants not what is actually happening at the time? The LCD is still a touch higher than the profile but would be pretty close.

Thanks
Luke
Attachments
log0067.klog
(62.68 KiB) Downloaded 230 times
User avatar
Wayne
Posts: 103
Joined: Thu 02 Apr, 2020 12:14 pm
x 29

Re: LCD temp readings not correlating with log

#1673

Post by Wayne »

Good day lukehk

Thanks for this .. I just want to say : "It's awesome that you are actually doing the test" .

I use the live time /temp on the front display as a live reference with my laptop open next to me .

Its not uncommon to get a slight difference in temp when looking at the display and comparing to the log.

The Kaffelogic Studio program is constantly converting time to profile time with the PID control parameters tuned to match shape of planned curve and not time . This gives the roast a better - closer to the profile shape without meeting temperature on the screen .

The mean temp is constantly averaging spot temp and temperature .

If you have to open the roast log in excel , the temperature will match up .

Studio will convert all times to profile time and this will cause the log to increase by +- 5-10 degrees C .

The difference will be larger in the beginning as studio aligns itself to the variable delay depending on your ambient temp . If the profile starts at 20 deg and your ambient temp is 24 or 25 - after roast , 35 deg . Studio will compensate for this and the added smoothing of the curve by compensating temp .

As the roast evens out , the temps at the same time , on profile and live screen , will start to meet or be 2-4 deg out from each other . This is seen in your example at 1min . So the algorithm will continue over the roast .

I hope this helps . This is how I understand the correlation between the two .

Thank you again for this .

Wayne
lukehk
Posts: 14
Joined: Wed 05 Aug, 2020 4:02 pm

Re: LCD temp readings not correlating with log

#1682

Post by lukehk »

Thanks Wayne for the detailed reply. Would you say the log is more useful in looking at what to adjust or the LCD. Judging by the reading on the LCD i am following the intended temperature early on so would keep the preheat as is. But on the log it's way over so i would reduce the preheat. This is using steves new profile where a power profile is used in the first 40 seconds.
Thanks again for your advice
Post Reply