Hey guys, I just finished installing the HTC-TC4C board with these thermocouples: http://www.omega.com/pptst/XCIB.html, type 3. I hoped the collar would expose enough of the thermocouple wire to get accurate readings, but it looks like the data is all offset by quite a wide margin, compared to all the original Hot Top display readings I'm used to seeing. Could anyone take a glance at my .csv file to see if it's reasonable? Unfortunately I didn't save a picture of the chart, which would likely make for a quicker assessment, so here's a link to the raw data. Thanks for your help!
https://www.dropbox.com/s/0flszobxwo2qo ... a.csv?dl=0

You can easily recreate the chart. Any previously saved log file can be viewed in the chart at any time. Just run RoastLogger, click the Template tab and the Load Roast Log button. This will load the log and recreate the chart. If you wish to use the chart as a template for the next roast use the Load Template button instead of the Load Roast Log button. Once you have reloaded the log you can save the chart in various sizes using the File menu.

I have loaded your log file and created the chart for you (size reduced to suit forum limits):



The Omega XCIB thermocouples do read low as explained in the RoastLogger Thermocouple Installation webpage.

For computer control of your roasts you only need to make one adjustment to the default action table settings to suit the low readings. The adjustment necessary is explained in this RoastLogger webpage.

Please see the RoastLogger manual, written by Randy Glass, for details of how to fine tune the settings to achieve your desired results. In your case I suggest you look at the section on extending the time from first crack start to end of the roast described on page 16.

Wow, thank you so much!
I did know about reloading roast profiles, but I'm at work and can't install unapproved apps. Thanks again for the amazing info, I'll check it out!

Could you explain the bit about the stem effects a little more? I don't quite understand how the readings are lowered. Is it the partial sheathing that conducts heat away from the exposed tip? If so, maybe I could grind some of the sheath away.
Or would I be able to add some sort of heat resistant barrier between the rear wall and the thermocouple so that it no longer conducts heat there?

I will try to clarify this issue by discussing the bean temperature thermocouple installation. The same arguments hold for the environment temperature thermocouple installation.

The temperature in the rear power board compartment on a Hottop is typically about 50 C (120 F) towards the end of a roast. The temperature of the beans at first crack is about 190 C (374 F). Any thermocouple installation through the rear wall will conduct heat from the thermocouple hot junction to the rear wall and the rear compartment.

The thermocouple readings indicate the temperature of the hot junction. If the hot junction was perfectly insulated from the rear wall/compartment then it would quickly reach the surrounding bean temperature and accurately indicate that temperature. Any real world thermocouple installation will conduct heat from the hot junction to the rear wall/compartment. The more heat conducted away from the junction the lower the hot junction temperature and, therefore, readings will be.

If the thermocouple junction is below the bean temperature then heat will be conducted into the junction from the surrounding beans and their immediate environment. At the same time heat will be conducted out of the junction, down the thermocouple leads, through the insulation on the leads, into the thermocouple sheath (or collar on your thermocouples), and into the rear wall and any support fittings you use and the rear compartment. This heat conduction out of the hot junction and down the thermocouple sheath/collar causing an error in the readings is referred to as the stem effect.

To minimise stem effects and, therefore, achieve the most accurate readings you can:

1. Maximise the protrusion of the thermocouple into the roast chamber. On a Hottop this is limited by the need to keep sufficient clearance from the rear drum vane and the rear radial drum support to ensure beans can not become trapped between these items and the thermocouple.

2. Use insulation to minimise the amount of heat conducted from the thermocouple sheath/collar to the rear wall and any fitting you use to support the thermocouple.

3. Use insulation around any exposed section of the thermocouple in the rear compartment to limit heat loss to the rear chamber.

The RoastLogger Thermocouple Installation Guide, linked to in my above post, suggests a number of installation options to achieve the above. This guide makes it clear that installations using your chosen thermocouples will read low but have the advantage of being relatively easy to install and being ungrounded thus avoiding ground loop problems. Low readings can still provide a repeatable indication of the roast stage and can, therefore, provide the same level of control as accurate readings. For those who want accurate readings I recommend the PEEK rod solution as this provides excellent thermal insulation but many people find this solution difficult to make and install.

I would not recommend removing some of the collar to expose more of the junction as you suggest. If this could be done without damaging the thermocouple then it would improve the accuracy of the readings. I would be concerned, however, that the ceramic sealant/insulation used to seal the exposed end of the collar and support the junction would be removed exposing the thermocouple insulation and braiding possibly shorting the thermocouple leads to each other or earth and reducing the robustness of the thermocouple.