Roasting Natural Peru - too dark


Looking for some advice on how to stop my Peru Natural 1800m going too dark too soon.

Statics: Batch size 500g, charge 240c, Fan 3, Drum 9
Aimed for city and 11-13% loss

Here is batch 1:
Profile P7,5,3
got 14% loss and medium-med dark


Here is batch 2:
same temp change points, P7,6,4 to try to speed up roast and reduce % loss
Ended up medium-dark and 14.8% loss


So that went the wrong way - any advice as to how get lighter and less % loss, thinking start either with P6 then P5,P4 else P7,5,4 as next iteration? My only other guess is i could have missed the real FC and one i heard was SC…?
thanks Jamie

1 Like

Hi Jamie, I am not the expert in this group by any measure but for what it is worth something doesn’t add up with your starting temperature. If you are preheating to 240DegC, your starting BT probe reading should be higher than 150 DegC. Yours are below that.
I preheat to 220DegC and my initial BT probe reading is usually between 155 to 160 DegC
If I preheat to 240DegC, I would expect the initial BT probe reading to be around 165DegC or above.
You may not be getting enough thermal heat energy at the start of the roast for the fast and light roast you may be after.
To get around this issue some in this forum have suggested allowing the preheat phase to run for 40 minutes or longer to allow the thermal energy of the bullet to reach equilibrium.
Just a few thoughts. Good luck :crossed_fingers:

1 Like

I don’t know who made this but I’ve found it to be fairly accurate:

I’d drop at 208 for city. I think you aren’t getting to 2C but you could well be missing the beginning of 1C.
You dropped 4C higher on the second batch so had more weight loss even though the roast was quicker.


Great observation i completely missed that! good idea ill def let run longer and aim for consistent start…ive seen others 10c out on Charge temp…sounds like charge time can be inconsistent measure


yes i think your right, FC likely missed start, at my marked FC its very loud and pronounced, other Ethiopian naturals are much more silent…good point on drop ill pay closer attention too

FWIW I usually see FC at about 200-204C. If you’re dropping at 218C one minute after marking FC, and your RoR is about 5C/min, then you marked FC at about 213C, which seems late. Maybe FC for that bean is really quiet at first? Since you describe the roast as too dark, the 218C drop temperature is probably correct, but it might be worth cleaning your IBTS sensor just to make sure you’re getting good data. I usually see SC at about 223C,

For a light roast with 11-13% weight loss I’d expect FC at say 202C and drop at temperature of 207-209C with maybe 75-90 seconds of development post-FC. If you aren’t sure about when FC starts then I’d just drop based on IBTS temperature.

Hope this is helpful. - Brad

1 Like

Thanks so the IBTS was cleaned relatively recently say 20 roasts ago so should be ok.

my guess is FC is really quiet as above, so i went for the drop at 210 - i marked 9.41 as FC as that was when really loud cracks started, guess real FC was much earlier and quiet - looking at data around 8.55 or 203c. Beans came out much lighter and decent, below is the updated profile. The stone fruit / apple type flavours still came out so was pleased with this - ill experiment a bit more - advice has been super helpful as explains why ive gone on too long on other roasts :slight_smile:

If you think that’s the problem, you might want to try lifting the bean chute plug, just a little bit, and listen from there. The 1st crack is much more audible that way. Just don’t lift it all the way and don’t wait there for the entirety of the roast, but do it when you’re around 1st crack temperatures. Also be careful not to get too close and risk burning yourself.

I will take the tryer out to listen for FC around the time I expect FC to happen. Some beans will crack louder than others for FC. SC is harder to hear.

There is another thread here about hacks people have done to listen for FC and SC.