Drum Speed in relation to batch weight?

The drum speed is a variable that I don’t mess with much when roasting but wanted to survey the group to see if anyone else has any discoveries in relation to drum speed and the size of your batch? And as a follow-up, how many of you change the drum speed throughout the batch?

I typically use a higher drum speed for larger batches (1kg-800g), and maybe a 6 or 7 for 500g batches.

What are your go to drum speed settings for batch sizes? Have you had success with slower drum speeds for larger batches?

ALSO - Does anyone start with a slower drum speed to get more contact heat, then switch to a faster drum speed to minimize the drum contact? Haven’t tried this out but have been thinking about this and didn’t’ want to waste beans.

Also… this is my first post. Have been lurking since March and appreciate all of the comments and sharing from everyone.


Hope you get some answers. This is the first roaster that has given me control of drum speed and I haven’t a clue what to do with it. I did run thru all the speeds when I first got my Bullet and couldn’t tell any difference between any of the settings by just listening to it.


Theoretically should have an effect on how the beans are heated…not sure if it makes a difference on a drum this small.

For lower charge weight it does seem to help with temp/ROR stability. For 500-600g I use D7…for 900+g I use D9.

Here is something of interest:


Thanks for this!


I usually roasted 700-1000 gram batch. I found D3 at the beginning and higher at the end gave better result. It is like heart transfer more to the core of the bean. IMHO.

I just remembered this thread from several months ago that speaks to drum speed at each setting. Haven’t seen anyone do this measurement so hope it still applies.

I’ve been leaving the drum speed setting at D9 as I fine tune other parameters first.

Wayne, thanks for sharing the article.


Hah! I’d forgotten this… and I even commented on it about confirming my suspicions. I just let it go.

In fairness (as @luisens states) it was a no-load test (empty drum), so there may be more to this story. Guess I need to do some measuring of my own Bullet.


1 Like

Thanks, everyone!
If there are any other first-hand accounts of the drum speed’s effect on the batch size and bean development, I’d love to hear from you, too.

@kitty264 - Question for you… If you start at D3, do you gradually speed up the drum like you would lower the power as the roast goes on? Or, do you just switch to a D7 or so at a certain point in your roast and ride it out until you drop the beans? Curious to hear alternate techniques to using the drum speed in the Bullet.

I can say that when I compare the Bullet drum speed to my Phoenix roaster, the Bullet is relatively insignificant. The changes in Bullet drum speeds, which have been documented, are so minor it makes very little difference in the results of the roast. I’ve seen very minor temp changes from D4 to D9 unlike my Phoenix which allows me to go full speed or completely stop the drum from spinning.
Take a look at roasts by user @Lav1 who has a lot of experience with the Bullet. He sets the drum at D9 for the entire roast.

At the present I start increase drum speed aftet yellow by 1 at a time. I notice increasing drum speed lower the ror just like increase fan speed. But I did not do any comparison. May be it is the placebo. IMHO

I was a commercial roaster & over 17 years & four machines, learned that every roaster is unique; for me, trial & error was a requirement, not an option. My biggest challenge was converting from slow-roast (13-15 min roasts) methodology to roast cycles on my bullet that range from 8 minutes (naturals) to a max of 11 min (the denser Indonesian coffees) and an occasional 12 for dark roasts. My batches vary from 500-700 grams and I have NOT attempted a full-load roast on my bullet. A long pre-heat is essential for me & I usually throw in a partial batch of old or discard green beens to heat temper my drum. With few exceptions, my bullet is cranking out exquisite profiles at all roast levels for me.

METHODOLOGY: A – extended pre-heat B – Initial soak of 45-55 seconds @ level 1, fan 1; C – With beans soaked, I turn heat to 9 (occasionally 8 for fragile naturals) while reducing drum speed to 3 until yellowing. D. From approx 315f up to first crack, I GRADUALLY increase drum speed to 6 (because I have found that faster drum rotation has a slight temp reducing effect). Here, I carefully monitor that ROR. With some coffees (natural & triple washed), a single notch temp reduction is often necessary. E. From mid first crack to drop, you’re on your own. I say this because I believe bullets are EXTRA sensitive to backdraft (wind, humidity, exterior temperature) and how you vent your exhaust can have major impact on the final stage of roasting. I have stabilized my exhaust with an in-vent exhaust booster fan, but have found it can easily disrupt a roast. Final note: Changing drum speed has had an impact on bean surface color in-roast for me. I’ve found that roast level temp at drop and weight loss % are more reliable factors for batch to batch consistency. IN BRIEF: Initial soak and In-roast drum speed adjustments have had a huge positive effect on in-roast consistency and development. No scorched beans with the bullet!

This works for me. Drop by & I’ll brew you a cup!


Thanks for this @quartzglen. I’ve left drum speed alone because I couldn’t detect any audible change from D9 thru about D5. I’ll measure it next time out, especially since I have no idea if any f/w changes might have related to drum speed.

Couple questions…

Re: “extended pre-heat…”

  • What are you using for pre-heat temp? Must be fairly high if you coast at P1 for almost a minute with cold beans before dialing up the power.

Re: initial soak “… @ level 1 …”

  • Am I right that “level 1” refers to power (P1)?

Re: “… Bullets are extra sensitive to (ambient conditions)…”

  • I can confirm that. I roast outside and have to constantly fiddle with power & fan, but then it’s not unexpected. Surprising to me is that summer seems more fussy than winter.

Thanks for the insight. And I appreciate @habitcoffeeclub asking the question. Definitely going to experiment with this.


1 Like

Bruce: I believe a cold roaster drum always needs pre-roast tempering. When I turn on my cold bullet, it takes about 15 min. to get to “charge,” at approx 330f. Not enough for me. I extend my preheat cycle with a partial roast cycle using approx. 12 oz. of discard green coffee (easy to buy at most local roasters for about $1-1.50 per lb.) that I take just past first crack/385 degrees.

After that partial roast, my bullet will automatically reach a preheat bean temp reading near 370f and a drum temp reading at approx 400 before it gives a “charge” notification. That’s what I consider the perfect drum pre-heat tempering for the bullet. At this point, I’m ready for my first production roast.

My bullet will then automatically announce a “charge” at 370 degrees, which will be consistent for all subsequent production roasts.

SO: At “charge,” I add cold beans (450-600 grams). I reduce my heat to power 1 and, after 15 seconds, reduce drum speed to 3. For approx. 30 seconds, the bean temperature reading drops steadily, as beans “soak-in” the ambient internal drum heat, evening out at approx. 230f between 50-to-60 seconds. At that even-out point, I manually “charge” my roast by turning temp up to 8-9.

SOAKING has always been a bit controversial. I’ve sat through many roasters’ forums at SCA conventions where roasters have debated the value of soaking. I believe the point where soaked beans and external drum are the same temperature is the ideal starting point point for adding a full-roast charge.

Take this for what it’s worth. I’m not out to prove anything. If you try a soak, let your post-roast cupping session be your guide.

P.S. As I wrote earlier, I’ve never done a full-capacity roast. I’m happy sticking with roasts between 450-600 grams. I have no data for you re: full capacity bullet roasts.


Wow. Amazing info and chat happening…

I tried out a soak last night with a full 1KG, @quartzglen

The first round was a little whacky and the ROR was a little hard to maintain… but the THIRD go was great.

I think with a full 1kg load, a power setting of P3 vs P1 might help keep the roast from lagging but still need to play around with that a bit.

The drum speed setting was actually really great at helping manage the ROR, @kitty264

Pleasantly surprised by that.

Starting at P3 then gradually increasing at yellowing through to first crack was great.

Starting from a slower drum speed and then increasing, I found a noticeable shift in the speed of the drum with each increase, @bab.

Give it a try going from slower and speeding up. From P7-P9, I didn’t notice much difference (as noted from @wngsprds earlier “speed tests”)

So… P3 start and gradually increasing to P6 from yellowing through to first crack.

Would bump to P7/8 closer to the end of the roast to help maintain the ROR.

This was my first roast session actively playing with the drum speed and found it super useful.

Tried the beans out this AM and was pretty pleased with the initial results.

Looking forward to playing around with it more.

Thanks everyone!


This is good information and helpful to me. It makes sense that a p3 setting works for 1kg roasts, since the objective of a soak is to balance the bean temp. with drum temp. at heat charge. A pre-roast soak is critical to me, precisely because I chose the conductive-roast bullet over a convective (fluidized bed/air) roaster for many “flavor” reasons. But it’s also very easy for conductive roasting methods to trigger the exterior of the coffee beans to start roast expansion while the bean interior is still green/hard (baking) and resistant to roast expansion. The result in coffee flavor is a kind of chemical acidity that is all too prevalent in coffee from conductive commercial roasters nowadays. Through a pre-soak where I can verify uniform temperature in the bean bed and the drum, I have the ability to more easily control uniform expansiion and ROR, whether I choose a faster roast at higher roasts temps, or a slower roast at moderate temps. Keep me in this loop, please. I’d like to see what you all develop using a pre-roast soak stage. Regards


Good info! Thanks. Looking forward to trying soaking and changing drum speed.

Thanks for the reply Glen. I’ll do some experimenting.

1 kg load is of no interest to me… I’m happy at 1/2 that. :slight_smile:


1 Like

Thanks… noted. Must not have tried low enough drum speed setting to be able to detect the difference.


I am glad to hear you were happy with the roasting.

I do increase Power during drying stage. But I always decrease heat after yellowing by increasing drum or fan, decreasing power. I have never tried increasing Power after yellowing before. Would you mind give more detail about this?