Roasting Washed Coffee from El Salvador


Good day everyone,

Got a new bag of washed coffee from El Salvador that I’m trying to develop a profile for.

This is my current profile

The taste notes are:Dark Chocolate, Lemon, Dark cherry aftertaste, Heavy body.
I’m definitely getting heavy body, dark chocolate, and a tiny bit of lemon.

I want to try and get a bit more sweetness and a bit more acidity from the coffee. How would you suggest approaching the roast with those aims?

Much appreciated <3


No expert here, but if you’re looking for ideas… I’ve been having good luck lately by reducing the RoR to near zero at the end of the roast. This flattens the temperature curve, and allows you to develop the bean without getting too dark. Try doing this by reducing power rather than increasing fan?

Hope this helps. Please post your findings!


Interesting. I will definitely try this. But, what do you mean by ‘the end of the roast’? Like around first crack


I’ve been reluctant to try this- afraid the roast wouldn’t reach target temp if RoR -> 0. But if you can get additional flavor development without having to darken the roast further, that does what I want. Thanks for this… I’ll give it a try.



By “end of the roast” I mean when you drop the beans. If all else is unchanged you certainly would not reach the same final temperature. It’s a trick to hit your target temperature AND get RoR to zero. This one was pretty close (and tasted really good). Maybe should have reduced to P3 a little sooner?

Screen Shot 2020-11-07 at 8.21.39 PM.png

or this one, which really did hit zero and was my best cup yet for this bean:

Screen Shot 2020-11-09 at 2.05.41 PM.png

These were both single roasts (no back-to-back) with exactly 15 minutes preheating.


Fascinating!!! Thank you for sharing. That’s definitely an interesting way of doing it. I will try that tonight…


@bradm Just noticed you used D9 in your roasts profiles above. Have you ever measured drum RPM at D9? I need to quantify that since my drum sounds like there’s a much bigger jump from D8 -> D9 than there is from e.g. D7 -> D8 (it shows up in RoR as a near off-scale jump-to or retreating-from D9). I would never have noticed it if I hadn’t been using drum speed to manage RoR.

Btw, your 2 example profiles above show excellent control of RoR- well done! Yeah, you’re probably thinking… “you should see the others!”, but being able to nail it like this without stalling is something to shoot for.



Here’s my drum speed data, measured on a cold roaster by marking the end of the shaft with a sharpie and counting revolutions for 1 min:

13%20PM Screen Shot 2020-11-12 at 1.41.13 PM.png

BTW, I tried D8 and it didn’t make noticeable difference in the smoothness of the RoR curve.

Thanks for your kind words. Those happened to be good curves :slight_smile: . I have plenty of “ugly” ones too, and many of those roasts cupped very well. Sigh.


Yeah, frustrating- the beans don’t really care what the curve looks like do they? :slight_smile: But I keep trying to be consistent.

Thanks for re-posting your drum speed data. From this sample it appears I definitely have a non-linearity-
D9 76 rpm
D8 63 rpm
D7 60 rpm
D6 57 rpm
D5 52 rpm
D4 48 rpm
D3 45 rpm
D2 41 rpm
D1 37 rpm

My suspicion about D9 being significantly different is correct. It seems to correlate with D9 on your roaster; it’s the settings D1-D8 that appear off.

Well, at least it’s better than Jan 2019 - Jun 2020 when drum speed didn’t change regardless where I set it. But the speed change is so great it makes it impossible to use D9 as part of RoR control. Still that approach in the D4 - D8 range has worked well for me (pretty much). I just avoid D9.

Thanks for this Brad.



Personally, I think you’re undercooking the beans a bit which is causing a lot of the sweetness to be lost. A 10-minute roast that finishes at only 205C is rather slow, and 200C is a low charge for 500g of beans. Experiment with increasing your charge a little; I’d say try for 215 and work from there.

Also, there’s a big jump at first crack that the roast never quite recovered from. Now that this curve is telling you where first crack is happening, make a hard adjustment right at that temperature like increasing fan speed while dropping the power by 1 to mitigate the inevitable 1st crack ROR rise.

Edit: I would not recommend playing with drum speeds at 500g, just keep it at D9 the entire roast. It’s one of the most volatile and difficult parameters to play with and you can get a lot more control out of adjusting airflow and heat application. If you were roasting an extremely small batch (say sub-300g) I would advocate for it but anything above that should be left at D9.


Hey @jmacrae523 , @bab, can you guys give some advice on how drum speed changes can be used? I’m roasting 325g batches most of the time (5# = 7 tries) which is getting pretty small. What effects would drum speed have, when to make changes, … Very curious about this.


Hey @bradm-

Drum speed has become my mainstay for managing RoR. I suspect its effectiveness is very dependent upon batch size. For my usual 550gm batch, drum speed adjustments are very predictable. In overview-

  • raising drum speed will cause RoR to drop a little while
  • lowering drum speed will cause RoR to increase a little.

My take on this is that at higher drum speeds the beans are pressed toward the drum for shorter lengths of time allowing less heat to transfer from the drum. At lower drum speeds the beans stay longer in the bean mass just aft of where the drum gets heated by the induction coil so they gather more heat.

I try to use incremental single digit adjustments to influence RoR. The exception is at 1C when I anticipate a sharp drop in RoR. At 1C I’m usually at D8, so I dial it back to D6 (2 increments) in anticipation of the dive and that usually works out about right. As 1C advances I take it up 1 step to D7. By the time I get back to D8 I usually have to start dropping power and increasing fan speed.

This is a roast that came close to what I wanted to do, but this is still a work in progress with ‘almost’ out numbering ‘got it’.

Also note that I backed off the heat ahead of the 1C up-spike to try to manage that. Without backing off power ahead of the spike it makes RoR management more difficult.

One thing I’ve noticed is that stretching development time at the end of the roast seems to allow hitting 2C at a lower IBTS temp. In this case 2C started at about IBTS = 439°F where I normally don’t hear 2C till ~445°F. I’m saying more than I understand, but that’s my current take on it.



Bruce, good info. I got an error message trying to see your roast, but interesting to hear your approach to drum speed…I’ve left that variable alone so far, but you’ve got me interested.


Odd. I was able to get the link ok. Been some changes here so I’ll leave a note for Derrick. Thanks.



Really appreciate you sharing those profiles and for everyone chiming in… I tried to do something similar with my El Salvador

The difference was noticeable. Sweeter and richer taste. I was super impressed and happy. I kept trying the coffee for a week. However, on the 6th day the coffee turned a bit sour i’m not sure why. The sweetness went away and it started tasting a bit stronger than what I would like it to. I did other roasts today so looking forward to trying those.

When do you see your coffee tasting best and by which day does it start to noticeably degrade?


Excellent to hear your coffee improved! Makes you wonder how good it can really get, eh? Fun to imagine there is an even better cup out there.

Why flavors change with age is a great question. I hope others share some insights. Some roasts are fabulous for the first 3 days, then fall flat. And some others suddenly bloom after 6 days and become great. Must have something to do with development time and temperature.

Another thing to check is your grind setting. I often need to use a finer grind for a fresh roast, gradually more coarse as the roast ages (for brew / pour-over.)


What @bradm said… right on the money.

I can’t give an objective reason for the flavor change, but I can say that for my current favorite from Chaipas, Mexico the change seems to correlate with the appearance of a few droplets of oil after a few days rest in the fridge (3 to 7). Once the beans are at room temp (I store them in an AirScape can once out of the fridge) I’ll see a little more oil appear and that seems to coincide with the best flavor. By the time the bag is finished, flavor is starting to fade a little.

My feeble attempts to replicate what Brad is doing (RoR -> 0 at drop) have produced a little more flavor. I would not have thought that possible, but there it is. And the drop temp is a little lower though bean color is still the same (pretty dark). Must have to do with stretching the roast at the end.