Typical Temps for Cracks and Delta Between Them


I made an earlier post where I was trying to transfer a roast profile I use on my Diedrich IR-1 gas roaster to the Bullet R1; I received a nice PM which was helpful.

One of the most important observations that I make on all of my roasts is to compare the weight of the roasted beans to the input weight of green beans. I find that the weight loss with roasting is typically 16% +/- 1%, and my most typically observed value is -15.5%. Obviously, one can get to a certain percentage weight loss in any number of ways with any duration of roast, so the weight loss is only comparable if one is using a similar profile with similar time-temperature observations.

That having been said, my first two Bullet roasts, both using DP Ethiopians I like for single origin espresso, 1lb of green coffee (454 g) per batch, produced less roasting weight loss than I expect; 14.5% in one case and 15% in the 2nd. The beans look slightly under-roasted to my taste, but it’s too early to try them yet (only roasted last night).

My questions are: (1) Using the bean temp, at what temp do you normally call the onset of 1st crack? And (2), also using bean temp, what is the temperature “delta” between onset 1st crack and the very beginning of 2nd crack for most of your roasts?

On my Diedrich, roasting at altitude, I usually call onset 1st crack as occurring around 375 F; on the Bullet, this seems to happen at a lower temperature, around 355 F (179.4 C) from what I’ve observed. On the Diedrich I have a “delta” between crack onsets of about 44 degrees F (24.4 C). When I used that “delta” to guide my 2 roasts last evening, I ended up with beans that are probably a bit under-roasted.

Do any of you have observations like this, e.g. how many degrees in bean temp do you typically notice as being present between onset 1st and onset of 2nd cracks?

Obviously, you can go up 44 degrees F (24.4 C) in 2 minutes or in 5 minutes, depending on how you control your heat input and your ventilation, so this may be an unanswerable question depending upon what coffees you roast and what you want your coffee result to be, but I thought I would ask.

Thanks for any replies!


Of course these numbers depend on bean type and how you interpret exactly when the cracks occur. That said, my Guatemalan roast tonight cracked at 177C and 208C (bean temps), 16% weight loss after dropping right at the start of 2C.

You can get a lot more data points on Roast.World using Beans > Find Bean, then picking from the list of roasts displayed on the right to see what others have been marking.

Hope this helps. - Brad


Hi Brad,

Thanks very much for your post.

I did another couple of roasts today and I am beginning to feel very comfortable using this roaster, at least with a smallish (454g/1lb) charge weight, and transferring my prior experiences over to the bullet. For one thing, the Bullet, at least with a small charge, is very controllable. The ability to control the power input and the fan speed actually makes the roaster change course after relatively short periods of time, compared to what I am used to with a high thermal mass gas roaster (my Diedrich).

Cracks are useful but they need to be put into perspective. No one in a commercial operation roasting kilograms (or sacks) of coffee is “roasting by ear.” Using the cracks is basically how a home roaster roasts coffee on an air popper. If you have thermometry, it is way more reliable.

There are, of course, ways in which the cracks relate intimately with the thermometry. Once the first crack starts rolling, the temperature increase/unit of time (ROR) is going to accelerate greatly. If the roast profile you want to do requires you to slow things down at this point (mine does) then this roaster affords ample opportunities to do just that, by reducing heat input and increasing air flow (fan speed).

The other variables, assuming that you know what you want your profile to look like, is the timing of these various events (cracks, duration thereof, and total roast time), and the weight loss percentage of the beans, green vs. final roasted weight.

I think that if you know what you want to achieve in roasting coffee on this roaster, as regards profiling and final results, this roaster is pretty easy to control, at least if you control it via a computer with the included software. I haven’t yet tasted any of the coffee I roasted the last 2 days (it’s too early), but at least looking at time-temperature-weight loss parameters, plus physical appearance of the beans at the end of the roast, this roaster is pretty easy to use.

Thanks again for your suggestions and comments.




I have noticed the larger the batch (more weight) the higher the temp for 1st crack. I was totally thrown off by that the first month of roasting when I tried a larger batch size. Beans from different regions also crack at a different temperature. For example, I am roasting (860g) a washed Ethiopian YG and it cracks at 383 F and a natural Ethiopian Sidamo cracks at 389 F.


Thanks very much for your response, AJ.

People come to this roasting process with different skill sets and objectives. I am probably among the least representative coffee roasters out there, when it comes to what I do and what I expect.

I roast only for espresso, and I have settled on single origin Ethiopians, mostly dry processed, as my “bean of interest.” This comes after experimenting with many other types of beans from many other places and simply narrowing down my focus to what I like. I roast for myself and a couple of others, to the tune of maybe 150 lbs per year. So I’m an idiot savant, of sorts.

Of course, not all Ethiopians roast the same; I do think that if you concentrate on a specific type of coffee and an intended usage, you can probably get to a very high level of expertise on that one little thing. Most people will not be satisfied with that, however for me, if I can’t produce the drink out of a commercial espresso machine, I don’t have either time or interest in it. That’s just me. Most people wouldn’t be satisfied with that approach, but I am. I’m someone who would much rather be really good at one specific application and forego all the other things I can be mediocre at. Just me, again.

Excuse my digression :slight_smile: Getting back to your kind post, my first reaction is that a spread of 6 degrees F for determining when first crack begins is not exactly earth shattering. Within the very limited universe of coffees that I roast (Ethiopian DP beans), I have a spread of at least 5 degrees F for when 1st crack commences, depending upon the bean, it’s age, it’s storage, etc. First crack to me is only a marker, in that it tells me about the next 4 to 4.5 minutes of the roast, after which I am going to end it. I never roast into 2nd crack intentionally.

Coffee roasting can be an engaging interest that one experiments with, plays around with different beans, varies the dose, and varies many other parameters. I’m personally at a point with this roasting process that what I want is repeatable results with the sorts of beans that I like, for the usage I intend, which is for making espresso drinks (more straight shots than any other).

What I would advise most people to do is to find what it is that they like, try to find a profile or profiles that work for them, try to become convinced that their thermometry is producing valid and repeatable results, and to weigh their roast product to confirm that they are in the range of their expected results. That’s at least what I do. And this roaster appears to have enough things that the person roasting can control, to allow decent and repeatable results.


In the ideal world same coffee should have same FC temperature no matter the batch size. However, with the newer Bullets there is also the IBTS (IR sensor) data available. Or if you retrofit it. With the IBTS the FC temp is reasonably independent of the batch size, as it is only concerned with the surface temp that it see. The more old school bean sensor has a lot of inherent measurement artifacts like timelag and the fact that it feel only indirect temperatures - why: The steal sheath that covers the sensor together with its placement in the bean mass. The notation Turning Point is unknown with the IBTS as it is not concerned by the air temperature in the roasting chamber prior to charging the “cold” greens.
Sorry for being teacher like here - and you might know all this already?

So back to your questions - (1) with full batches close to 200C. With half batches closer to 185C. Then on the IBTS it is always within the range of 200-204C depending on the coffee type. To your second question, I am not sure as I seldom go that far but have found deltas of 14-15C when browsing through my roasts


Hi Lav,

Thanks for your response; it contains useful information I can play with over the next few roasts. Now that i see how easy it is to control temperatures and the “rate of rise,” at least with a smaller roast quantity (I’ve used 1 lb/454 g batch sizes so far), think this will be a very easy roaster to use to get the results I want. Variables such as the precise preheat temperature don’t seem terribly important, when you can quickly get the roast on track simply using the power and fan inputs.

My sense from having now done 4 actual roasts (not counting the seasoning, during which I also learned some things as the roasting took place) is that if you are not roasting into 2nd crack, the important steps are getting first crack to occur when you want it to occur, and then finishing the roast at a time and temperature of your choice. I think the final temperature (whichever one or both you settle on using) is way more important than straining your ear to hear the subtle sounds of 2nd crack beginning. More importantly, and more repeatedly, you can go by temperatures, appearances of the beans at the end, and weight of the roast product. These finishing parameters can then allow you to construct a repeatable roast profile.

For example, my Ethiopian DP “recipe,” which has evolved over about 20 years, is to hit first crack between about 8:15 and no later than 9:30. I then want to get at least 4 and no more than 5 minutes in between the onset of 1st crack and the end of the roast, with a total roast time of at least 12 minutes 30 seconds and never as much as 15 minutes (which “flattens” the coffee in my experience). You need to be careful, of course, not to “stall” the roast in the process of drawing it out to last 4-5 minutes after the 1st crack. You never want a Rate of Rise of 0 degrees or an actual fall in the temperature.

The roast will end at a point where the beans have no visible oiling (at end of roast), basically a “Full City” visual appearance, and a post-roast weight loss of 15.5%, +/-0.5%.

I think if you were roasting for brewed coffee, you would probably want a faster finish, to get a brighter result. For espresso, my experience is that if you don’t get 4-5 minutes in between the onset of 1st crack and the end of the roast, the taste suffers and becomes mono-tonal. This might be less noticeable with some blends and some other origins, but using single origin Ethiopians, this effect is very evident, at least to my taste.

Thanks for your assitance.


I use IBTS as its suppose to provide more accurate reading. Although when I go back to examine my roasts history in RW, it shows BT temp which is hard to actually know what happened. For example, I drop the beams IBTS at 204c but in RT, it will record End Temp 179.8c. Are there settings I missed to change the way RT record the roasts?

My apology if this has been discussed prior and I hope I didn’t go off topic too much.


You are right about that - all temperatures under My Roasts in RW relates to the BT sensor, not the IBTS. This should be configurable or just added to the view @jacob, @matthew