A Friday Afternoon Experiment

Recently we’ve been talking about replacing the temperature probe on the roaster. The discussion came up about probe placement, and where would be best. Our UG-15 is something of a difficult roaster in this respect, because the fins on the drum pass very close to the front wall meaning you can’t get a probe very deep.

John, a constant source of good ideas, came up with a plan. We could take the front door of the roaster off, replace it with perspex, throw some coffee in and see where the bean mass was most easily measured.

I have to say that being able to do stuff like this, working with people excited to geek out and experiment, and have fun while doing it makes me very happy!

We had some coffee that was going to be trashed as there had been a gas pressure issue early on in the roast so we’d dumped it. (In case you are wondering why the coffee in the roaster is such a funny colour!) We learned a lot from our little experiment – and it was fun too!

I threw together a few clips from video I shot as we were doing it because I thought other people might find it interesting. Not included in the short are several of the trial and error attempts!

The eagle-eyed amongst you will notice that because the perspex didn’t fit super tight there was some massing of the beans where the movement was clearly restricted compared to how the would be moving when the roaster was in normal operation.

(Music: “Do the Astral Plane” by Flying Lotus)

I look forward to more experiments!

14 Comments

  1. Very enlightening to see where the coffee bed is most dense and to see how well the coffee is curtained in the roaster. We are also trying to improve the temp probe accuracy and precision on our 75k Probat where the fins situated similarly to your UG. I’m curious, how much coffee did you use for your experiment? Thank you for sharing the video and the insight I look forward to learning more.

  2. It might be an idea to repeat the experiment with roasted coffee just to check the probe placement is ideal for that too. Or, maybe with green coffee and then roasted coffee – presumably the density difference will affect its behaviour in the drum.

  3. Im kind of surprised to see beans in contact with the drum for such long intervals of time. I suppose with different amounts of coffee in the drum a variable drum speed motor would help keep that contact time consistent from larger to smaller batches.

    I’m also surprised no one has tried this yet. Seems like the most basic question – what happens to coffee inside a spinning drum?

    Great job, looking forward to more.

  4. I’m wondering if anyone has ever tried placing a series of RFID thermal sensors in various positions on the inside of the drum to gauge drum temperature? I’m imagining you could create 3d model to represent the drum temperatures across its structure (kind of like how motion capture nodes work).

  5. This is genius! I love simple, but smart, solutions like this. I think most people have an idea of what’s going on inside the drum, but until you actually see it, what we “know” is really just a best guess. It would be amazing to do this while actually roasting since the density of the beans changes continuously. Transparent aluminum?

    More tests and experiments like this would be great not only for temperature measurements, but for the evolution of roasters in general. Nice work.

  6. It would be interesting to make a portion of the front of the drum out of this http://www.fireglass.com/glass/neoceram/
    I’m sure it’s not all that cost effective compared to the metal, and would need to be cleaned often, but it would be informative to have a visual all the time with different batch sizes.

  7. what is your bean probe telling you? I feel like a bean probe is a great tool for learning the basics of roast development, but that in the long run it becomes a crutch where people are reading off numbers instead of paying attention to actual roast development. Besides, does the reading truly need to be accurate, or just consistent? I know that everybody is going to say “you use your probe along WITH paying attention to development”, but I personally find the probe to be distracting and when the probe came off of a UG 15 that I was roasting on, I didn’t rush to put it back on and it was the most fun that I had had roasting in a long time. The most important reading in my opinion is the temp. at which you charge the drum. Everything else is just reference.

    That all being said, I do truly believe in using gauges and guidelines in order understand what is happing in the roaster and I think that Ian’s idea would be pretty cool in figuring out the heat map of a whole drum, that would be pretty sweet and help you make proper adjustments with different batch sizes. At the end of the day though, I would rather involve myself with the coffee itself rather than numbers. I don’t know, maybe that’s just a can of worms.

    Also, this is a cool experiment as far as understanding how the mass behaves in the drum, good show.

  8. Have you considered infrared temperature measurement? The technology doesn’t require contact with the surface it’s measuring the temperature of, so you may be able to acquire the average surface temperature of the beans while in motion.

    Here are some (technical) details:

  9. where is the probe now and where did you decide to put it?

  10. I looked at infrared probes but they are not cheap, plus it would still be a bit of a risky experiment.
    The probe ended up in the middle of the bean mass on the right hand side of the door, we did the testing with 12.2kg and 5kg
    I also kept the fake door with all the drawing on it.

  11. Reminds me of a scene in Conrad’s “The Heart of Darkness” where a doctor of phrenology (common practice in the 19th Century) is measuring a man’s skull in the hopes that he it will reveal something about his nature.
    I believe precise measurements have their place but I think a more important measurement will be done after the bean probe is moved and the coffee is cupped.
    The music and video were very poetic and inspiring. Overall well done. Keep us posted of the results.

  12. Hey James, great post. We have an old UG30 that we added temp probes to, the first time around our placement was off as well as only having about 1/2 inch of immersion in the bean mass. We were dropping french roast at 380 F. A reference point, but nowhere near accuracy. Upon trying to better place the probe, I discovered that the fins on the front spokes were fairly easy to remove which gave me the extra 1 1/2 inches to get a proper reading. I have not noticed any negative effects from removing the fins, like uneven development or scorching and my temps are spot on, bottom out around 145-155, fc 380-388, 2nd 430-438, etc.

    I also found it helpful to bend the RTD to a 90 and turn it inside the bean mass while roasting coffee so that you can see a variety of different placements. I ended up settling on the placement that gave me the biggest range of temperature.

    If you would like, I can e-mail you a few pictures, just let me know.

    Keep up the good work!

    Caleb Nicholes
    Kickapoo Coffee

Submit a comment