To quick to FC and roasts are not even

Hi there,
So after roasting on a Kaldi for a year I bought a bullet 2 weeks ago. I did some test roast. Some are ok, some are not.
Yesterday my first roast session was inside under my new ventilation system.
I followed Rob’s Hoos’s (very) small sample roast recepie but got a lot of error’s.

  • I go way to quickly to yellowing stage and to quick to FC.
  • Beans are not even in color. Some are dark, some are light.
  • Some beans did a FC already after 3/4 minutes. Others just after 6 minutes.

What did i do wrong? I just did the recepie from Rob Hoos with 4 different beans. Each 150gram.
Also tried 300g… this was more even… but long from perfect.
If anyone is willing you can check my last roasts.

Hopefully any suggestions or tips.


I’m sure someone much more knowledgable than me will chime in but the first thing I noticed in all your roasts is the extremely low first crack temp. No matter what I’m roasting, first crack consistently starts at around 195 C. Depending on the bean, 1C can sometimes be very hard to hear.
After 2 + years on the Bullet I still get tricked occasionally and always fall back to checking the ITBS temp to verify that I’m close to 1C. The fact that your beans appear light colored supports that conclusion. I know there is a tendency initially to roast small batches in case you screw up but I think you will find that larger batches (500 g or larger) are much easier to control on the Bullet. Consider trying at least one larger roast a try.


Hello. Mellek,

Welcome to the Aillio world, the answer is related to batch size, the energy is too high for the small batch you are charging, I am not quite what is the minimum batch size to be roasted on the bullet, but I believe I read somewhere is 350 grams to be able detected by the the IBTS. For me, I am more into (%80) of the roaster’s capacity. I charge the beans on 240 C, Ethiopian Dried Process, With Total roast time around 8 minutes, due to the batch size and bean’s density, in my coffee, I aim for a clear fruity and floral notes.

More, the 3 batches you roasted already, are for different beans with different density, if. You want to do tests and understand the roaster, I advise you hold on one variety and see how the profile goes as a response to the changes you make.

The differences in beans color sometimes due to the nature of the beans, but also may be a sign of heat transfer method used mostly is conduction. I can’t tell the fan speed in your profiles, but I think they are low and low drum speed as well.

Tell me what do want to achieve from your roasts, then we can work on designing a profile together if you want.

Happy Roasts To You.

1 Like

Yes exactly. So low, but I’m sure I hear them pop. Also, some beans are VERY dark at the end of my roast. Still, I’m under 200 Celcius… How…??
I will try larger batches later today. Hopefully with better results.
Thanks for helping!

1 Like

Hi alghamdi,
Thanks for your reply.
I will try larger batch sizes later today. Maybe this will do the job. And stick with 1 bean.
Earlier I tried a 1kg test. How does that looks like you think?

The profile I’m looking for is light/medium roast. I have Brazil and Kenya beans and would like to have a constant roast. 500-800g each time.

This profile more into what would expect from medium roast, Heavy body, chocolate notes, sweet with clear but not annoying bitterness I guess.

About Roasting and profiling, there is no one correct way to do things, my way in roasting may suites may not, but even if did, don’t stop there keep on absorbing and seeking different techniques, that eventually will lead you to have your own style and methods in roasting.

The advice @alghamdi.majid1b1s gave lines up with my experience too.

On the Malabar roast, I forget what power level Rob Hoos was using in his Bullet sample roasting article, but it seems that P7 is a pretty high setting for a 150g roast, especially for a variety that tends to roast quickly.

For my typical batch size (455g) I preheat to 235C, and use P6 (sometimes with a short soak) to a little before 1C, the P5 until a little while into 1C then finish at P4.

Notice how the IBTS is lower than the BT ?

That’s opposite of what I’m used to seeing. I wonder if you have a dirty IBTS and perhaps the roaster is preheating to a higher temp than you realize.

(edit: Oh, I see you’re doing 150gram. I dont have experience with that, but I know the probes behave differently than with larger batches. So probably ignore my comment, thanks.)

Why don’t you follow the standard light roast recipe provided in documentation. I had been roasting for some years on a behmor prior to the bullet, but found that sample recipe an ideal starting point. I have simply modified it for a medium roast ie 20-25% development time.

the challenge there is that 150g is not a standard batch size, so a standard roast profile likely will require some adjustment

1 Like

Linking Rob’s small batch roasting article on Aillio’s site in case folks haven’t seen it.

1 Like

It seems after seasoning my IBTS sensor was already dirty. Did a clean and after that 150g was more even. I tried 250/300g. Seems way better results. so now I know what was wrong.
Thanks everybody!

1 Like

I just looked at one of your very early roasts (6 Jan). That’s one of the most extreme examples of IR sensor cleaning I’ve seen. Maybe bits of chaff blocking the IR path. If you see IBTS and B-Temp swapping position it’s very likely there’s maintenance needed on the IR sensor. Good job figuring it out & cleaning the sensor!


Hi Bruce,
Thanks for your reply. Indeed, this was a sign things were wrong.

Here is one of my new roasts, after cleaning. Better? What do you think of it?
Thanks in advance.

1 Like

That profile is what I would expect including nice even decent of both B-/I-RoR. But pretty pictures don’t mean much of what ends up in the cup isn’t good. If you like the coffee that’s all that matters.


1 Like

Hi ,I’m having the issue with the uneven roast color, and I found your explanation kinda make sense to me. I’ve been roasting with 200g batch size, with D9 drum speed and F6 fan speed, 210c ish CT. Been having inconsistent color with 3-4 diff beans.

What would you suggest to overcome the issue ? I found it affect a lot in term of the taste

Def clean the IBTS infrared sensor. Did the job for me. Also try a 500/600 batch one time. See if that is more consistent.