Effect of Fan Speed


@jboutte88: good to hear! I hope that the Avid discussion between me and Deven wasn’t a too big distraction :smiley:


its not entirely clear whether you’re referring to the roast or the cooling fase (although lowering fan speed during cooling is a rather obvious to have a more developed outcome, som im guessing you’re referring to the roast fase).


@rsegers sorry, yes, the effects were noticed during roasting


@pavel, yea that definitely sounds possible, might I suggest asking the company you get your green’s from already if they would give your their extra samples they either rejected or don’t want, or maybe a low grade bean they carry for circa $2.50/lb. Have you looked into Agtrong?

Yes, this whole time (in this thread) I have been talking about fan during the roasting phase. I don’t think I’ve mentioned the cooling fan. The only thing I’ve mentioned of my belief (somewhere else?) in regards to the cooling fan is to keep it as high as possible without letting the beans escape. Which I usually keep at 78%, and sometimes they do still.

@Geoff_IKAWA I believe of the “Marshal Etheo” profile I have posted two profiles one with the 82% constant fan and one with the 78% constant fan. I accidentally also posted two profiles of my “A profile” with two different fan profiles, The “Mandheling 71A” I think I posted only the one profile. I’m not sure which one you are talking about, sorry for the confusion but the cooling fan is usually at 78%, if not my bad.

Can you post a link for your profile? Even though I’m not a fan of spikes I’d still like to see and give it a try.

Edit: In retrospect, maybe a low(er) fan speed during the cooling phase would help compensate for the low thermal mass, so the beans don’t cool down too quickly, maybe also help extend the cracking more, so I can drop earlier in the crack rather than waiting for the “perfect” time. Maybe part of my coffees dud-ing out could be the cooling? Also if humidity is a factor of conductive heat, maybe a lower fan (70’s) can compensate for a low humidity (50%)



Here ya go! Perfect for my Dumerso Yirg. natural. Left my wenago Yirg. Nat. way underdeveloped tho.


Deven I ran your Marshal etheo profile on the last of my Rocko mountain and it seems a little underdeveloped. Did you pre heat before running? I didn’t get any cracks but the beans are puffed. I ran it with a flat 81% fan speed. I discarded some beans that looked too pale and hard and bagged it.

I’ve brewed it up over the last 2 days. It has a lovely blueberry aroma on grinding and brewing but very little dry aroma from the beans. It’s a really tough grind and takes ages to drain.

It was nutty, slightly bready, yesterday with nice sweetness and blueberry notes. Today the nuttiness is gone, pretty thin juice like body but again nice blueberry ans some lime coming in. Quite pleasant. The natural “funk” is very low in the mix.

It might be that a declining fan speed would have brought on 1C better but I’m out of beans. I think it is a promising profile. Thanks for sharing!


I don’t know what you mean by Rocko, Rawandan(?)for these kind of roasts that green note your getting will fall off after a couple days. I dont know if your sorting after and taking out your quakers but that could be where the nutty is coming from . Natural funk is probably coming from not sorting pre roast and taking out the defective beans. For your pull you should prob grind a little bigger, less pressure and higher heat.

Also glad you liked it, it sounds silly but I was thinking 80% might be the sweet spot. The person who that profile is named after said to keep the fan half way, I assumed that was around 80% for us. Did you happen to taste blueberries in the bean after roast? For my Ethiopian DP it did not have a prob with very noticeable cracks, also , yes I do preheat the roaster at 490 for 8 min.


It’s an Ethiopian natural from Haricha Wpreda, Godeo zone cup score 89. Accompanying roasting tasting notes are “boozy, blueberry, citrus”.
So I expect natural “funk” or “booziness”. It’s not a defect as such but a part of the natural process as mucilage ferments on the bean during drying. All naturals have it to some extent and is instantly recognisable.
I didn’t taste any blueberry immediately post roast and there were not any noticeable quakers just a few that were noticeably paler than the rest. I was really surprised to find the blueberry 24 hours later. Tbh I thought the roast had failed at the time. Glad to be wrong.
Like I say, the roast was slightly underdeveloped (for my taste). It’s a different bean from the one you used so there could be differences in moisture, density etc… and I did no pre heat, all factors that will effect the roast. I immersion brewed the coffee, I don’t make espresso.
I think it is a good profile, just saying that imo it needs a bit of tweaking for this particular bean, but I have run out. I enjoyed the coffee and I will be sure to try the profile on the next Ethiopian natural I get.
I think the higher fan speed delays 1C and makes the beans harder. I might try declining the fan speed after drying next time.


yea I have a feeling that unless we have the same beans the profile will be needed to be tweaked. And even still because of climate/location, age, etc.
Glad it worked for you, How this profile comes recommended if your going to move the fan move the entire thing and keep it flat, but you are the Picasso of your profile so do as you please.


Picasso indeed! Nice one.
I’ve noticed that high fan speed through drying leads to harder beans to grind , seems to increase drain times for brews and pushes out 1C.
I’ve had some nice results on a Columbian with declining the fan speed after drying so I’d like to experiment some more with that.
I’d also like an extra fan speed point to experiment with bringing the fan speed back up at 1C. Currently we can only do a “V” type shape


What do you take to be a high fan speed. For me it’s 85% or higher, I don’t really go above 90 but I could see how spinning the beans faster against the wall in the beginning can help with the dry and getting more heat into the bean initially.


It depends on where you are in the roast. 80% say is not high for the start but could lead to beans jumping out the chamber once they have got lighter.
I haven’t tried much above 80%, so 85% does sound high to me at this point.


A bit late to this topic - but a few random thoughts…

This seems to make sense to me. Higher fan speed would mean more evaporative cooling, which means that with the same temperature probe reading the bean will be cooler. So lower fan speed means a warmer bean = darker roast.

Also lower fan speed will increase conduction from the chamber to the beans (more contact with less movement). I am not sure how significant this is but it would/should increase at lower fan speed (assuming the same temp probe reading).

Also conduction between beans will be increased - but this does not add extra heat/energy to the total bean mass - so it should not affect the roast level (though it may make the roast more even???).

If the evaporative cooling is the (main) reason for the cooler beans - then this effect would be stronger at (mostly earlier?) stages of the roast when there are more volatiles coming off the beans.

One might think that during the exothermic stages that this argument breaks down - but higher fan speed would still cause more cooling - so the bean would still be cooler than at a lower fan speed… (but closer too or hotter than the probe temp for that phase)

Obviously if you turn the fan down too much this idea would/should break down as the charge of heat (amount of heat energy) in the air will not be enough to heat the beans properly. So if you go too slow the beans would not get enough heat to roast properly. However once you get past a certain “equilibrium point” the increase in heating (from more hot air) is less than the increase in evaporative cooling (from more airflow).

Also there would there would/should be an upper “equilibrium point” also where the extra heat (from more hot air) is greater than the increase in evaporative cooling (from more airflow) - beyond this point you should start seeing more airflow creating a darker roast (at the same probe temp). The reason for this is that the evaporative cooling has an upper limit - because the amount of volatiles that can be driven off is finite.


This is just an update or confirmation based on the experiences I’ve had from the last time I talked about this.

What I’ve noticed since putting in my oven thermometer is that YES, a high fan speed at the end increases the temp much faster of the thermometer, BUT does not increase the color of the beans as drastically as a cool fan speed. This I think I a very particular phenomena that I would not feel comfortable enough to explain yet.

This is a yes

Also yes, the counteractive force, is the beans themselves becoming “exothermic” which I don’t think is the issue for us. What the issue is: the conductivity, because the increase in porosity aka air space within the bean from cell to cell. Makes it harder to transfuse energy and sometimes this cooks the inside of the bean and not the outside.


To further my thinking of how the fan mitigates energy during the roast. We also need to think of the way we are cooking the beans and how that is different (then a drum). Because we are not using a lot of conductive energy, the transference of said energy to the center of the bean is largely buffered. What is being buffered? warning: personal beliefs ahead The embryo of the coffee seed (bean), the embryo is where most of the ingredients for creating the flavor are being held. The surrounding cellular matrix composed of cellulose, minerals(etc.) and the result of the processing method contribute to the second portion of flavor that you can get out of the bean. The fastest way to get to the center of a Tootsie Pop is to lick faster, and so by increasing the contact time and speed (friction) of your tongue. And the fan is licking the beans, removing water instead of sugar as in the case of the lollipop. But that is not to say enough heat from the coils can’t be applied in the beginning to get to the bean faster. As well as fast enough fan will wipe away moisture and start cooking the inside faster. So it’s a balancing act of which kind of the 2 energies should be applied. In what ratio will determine the outcome. As an example think of the Anduino profile and it’s high charge to a temp. By applying enough energy via heating coils there can be such a significant increase in temperature that the beans will reach 1C very fast, but is slowed down by the fan speed. Like wise the “S” Curved profiles can use a steady fan because the heat applied is very gentle.

I wouldn’t think of it as cooling, or you could. But maybe more like evaporative slowing. I think the answer for why the beans get darker is a matter of pressure. With a high fans speed there is a negative pressure occurring in the chamber and as the fan slows the pressure of the chamber is increasing and the focal point of pressure aka the point receiving the most energy is the center. Which could explain why when I cut the beans of a profile with a decreasing fan starting around 1C the center is darker and the beans taste “baked out”.

It does seem to add more energy, the beans do cook a little more but I think mostly it is because the metal grate where the fan is, is so hot, as well as the walls are very hot. The conductive energy of the metal at that temperature does have a significant effect. As well as that energy differential between bean to bean that from ones that held more heat and those that held less are able to “even out” so to say. That is at least, what I think is going and I’m not sure it is correct or not.

Because like was also mentioned in the Ikawa SCA video, the fan is tied to the heat and so with an increase in fan, there is an increase of energy given the same amount of heat input from the coils. It would be best to think of these phenomena as which heats the beans more than the other and in what combination. Only because having an actual cooling effect will have negative effects on the roast outcome and so any dip that the physical bean temp takes will have a negative taste effect (I can be wrong, but for this I doubt it).

Lastly towards the end of the roast (and also something I’m going to start another thread to discuss, is that for us air roasters, the main contributor to color is set by the temperature momentum on the profile) there is very little we can do to ultimately change the color (at the end). But also through observation (and maybe I’m remembering wrong) an increase air speed (at the end) will cook the outside more than the inside, but the bean is very resistant at this point. But what is weird is that the oven thermometer I set up to measure the environmental temperature will increase faster no matter when, if the fan speed is faster.


Techinally it is still “evaporative cooling” (at least as I understand it) - even if the beans continue to get hotter.

The RESULT of the cooling is slowing the temperature increase :wink:

And I agree we don’t want to cool the beans (-ROR) - but that was not what I meant :slight_smile:


I am not totally aware of how this all works yet :slight_smile:

But still in this case you are increasing cooling (from air) and increasing heat from coils. But you fine tune it so they balance - or probably continue to heat slightly as you (mostly) want the temp to go up.

But I am sure there is a lot of subtle interactions and balances going on that we have not exactly figured out yet :smile:


Hey sry I wrote this very late and I hope I didn’t sound too much like jerk :sweat_smile: What I should preface is that on the whole I agree with what you have said, but the devil is in the details lol. Yes I agree that evaporation cooling is happening and that is the base for the thermokinetics of the roasting but I wanted to say it that way because that I how I used to think of it, and not how I see it now. Mainly that the roaster by it’s convective nature is overpowering any cooling effect the beans produce. How I think about it now… is not necessarily worrying about any cooling of the beans except in a decreasing ROR or cooling fan speed/time. And a slow fan will only slow the heat being applied, and thus slow the momentum of the bean heating. The most important thing I feel with the temp input and fan speed is getting the right momentum. This is mainly coming from observing S shaped profiles (Do those Curves look like Spirals?) as it that thread I can get a little hot air so please please please do not think that I mean offense or anything and my apologies for sounding like jerk. :smile:

To answer your question: why I like to think of it as evaporative slowing vs evaporative cooling is because the heat being applied is so great that the overall total energy is still increasing. Yes water on the surface or within the bean is breaking into steam and thus loosing energy which can be seen/measured but because the bean in still increasing temp even though this cooling is happening, that is why I call it “slowing”, because it is resisting the temperature increase. Now, to further how this applies during apply this to Scott Rao’s “Crash and Flick” observations. (check out his blog https://www.scottrao.com/blog/development-time-ratio, this isn’t where you will find the complete answer but this read is along the path to finding it, because I am not 100% sure either, just confident)…because we have a negative pressure effect and a high exchange of energy the things that cause the beans in a drum to crash (cooling) and flick (rebound rapid heating) is avoided by keeping a calm steady rate of rise.


Looks like this article got taken down last night…


I have nothing to add about the effects of fan speed, but I would like to know how you set-up an over thermometer to measure ET in the Ikawa Home. Thanks.


The Lengths We Go I purchased a hanging oven thermometer from a large, very large, retail establishment. Broke it (gently?) with a hammer and removed the inside metal plate with the temp reading on it. And just bent it to fit snuggly on top, on the metal part. I was very careful not to put it on the foam part and not scratch/tear the foam either. You can see it in the thread I linked to and made in March.

@jboutte88 also did a better option but I could not find his post.