Home roasting again.... finally!

[phaelon56]

I got inspired enough by doing the foodblog this week to dig out my trusty coffee roaster and get back into roasting. In all fairness, the primary reason I have not roasted since last summer is due to being in transitional living quarters from July '03 to January '04, before moving into my new house. The house has a great roasting spot on the counter in an inside corner with two windows. Perfect cross ventilation for that pesky smoke.

I have quite a backlog of beans to work with and will test out some different blends over the upcoming month or two but my current goal is to establish a "half-caf" or even "quarter-caf" blend that wil have reduced caffiene yet still deliver the flavor and crema I seek. I love having a good cappa or latte after dinner in the evening but my need to sleep at night is usually in conflict with the caffiene.

I went home at luchtime today and roasted the first half of a batch of espresso blend for use this weekend. Most beans, when intended for use in espresso coffee, will require about a 24 - 48 hour resting time after roasting. This is best done in a container with the lid cracked slightly open. It can be consumed sooner but really hits its sweet spot in about two days and stays good for about a week after that if stored in an airtight container ina cool or room temperature place, out of direct light.

I have loads of green beans right now - was going to use Jim Schulman's WTC recipe (Way Too Complicated) but forgot to bring it home with me. Jim is legendary over in alt.coffee and Coffeegeek forums. He's a regular coffee guru and a real gentleman.

I decided to wing it and do a "semi-decaf" blend. Decaf beans tend to deliver a bit less flavor and less crema than regular beans when used for espresso. When preparing a blend of regular beans for espresso, most blenders use a "base bean" - something relatively neutral in terms of flavor notes and relatively mild. Many Brazilian beans meet this requirement - it's common to use about 30 - 40% base bean and then a blend of others to achieve different flavor notes. Decaf changes things - its more subdued flavor profile dictates a different matching of beans and a mellower base bean is not as approproiate.

Here's what I'm doing:

1 part Sumatra WP decaf (water process)

1 part African Highland Blend WP decaf

1 part Uganda Nanga Farms Robusta*

2 parts Indian Monsooned Malabar Coelho's Gold

1 part Yemen Moka Haimi

This blend may be a bit too much - five bean types is pushing it but I'll see if it works. I'm testing to see how far I can push up the ratio of decaf before I lose the cream and big flavor notes that I'm seeking. There should be a good balance of fruit and chocolate notes in here.

* Robusta is famous for being the crap commodity coffee that makes its way into canned supermarket coffee all over the world. It tends to grow more abundantly and more easily in greater volume at lower altitudes but by most serious coffee drinkers standards often has undesirable characteristics . It does add a certain type of bitterness that some Italian blenders find desirable as a balance but more important, it produces abundant crema. My hope is to offset the bitterness by including a larger proportion of of the Monosooned Malabar and have the Robusta counteract the reduced crema output caused by use of decaf.

Nanga Farms is one of the few well known high quality robustas available readily here in the US and is close to the same price as Arabica beans.

The beans at weighing time:

The Alpenroast ready to go to work before the lid is closed:

It's a drum roaster, a bit smaller than a bread machine, and sells for about $280. I got mine used for $175 and contrary to reports about many earlier units that were finicky and unreliable, mine has worked great. The manufacturers claim that this is a "set it and forget it" unit like bread machines but to get good results, one must start listening to the sound and pace of the cracking beans starting at about 13 minutes and also make note of how the smoke smells. I usually shoot for a roast best described as Full City + - darker than a medium roast but not to the French Roast level.

Some people use temperature probes and spreadsheet software, carefully documenting their roasts so they can achieve repeatable results with particular beans types and blends. I'm just not that geeky (maybe I am but I'm just too lazy!). I just listen, smell and stop when I think it's done. Seems to work okay for me. Typical roasting time is about 15 - 17 minutes. The beans are close to cooled off but not quite when the roaster spits them into the collection bin. I spread them out on cookie sheets and stir to finish the cooling process.

Fluid bed (hot air) roasters of the consumer variety start at about $100 but many folks use old popcorn poppers. Unregulated air roasting takes beans to the final roast level very quickly - about 4 to 6 minutes. This tends to yield a brighter finish to the roast - good for some people but not to the liking of many, especially for espresso blends. Voltage regulators (known as Variacs) can control this process and allow consumer hot air roasters to roast more slowly by varying temperatures at different points in the roasting cycle, thus achieving results more akin to a drum roasters. Commercial fluid bed roasters such as Sivetz are an entirely different beast and not really relevant to this discussion.

Our own eGulleteer Mike Lloyd is a big proponent of the benefits of low tech roasting with simple tools. Here's a very extensive Coffeegeek discussion on

Heat Gun / Dog Bowl coffee roasting

[phaelon56]

I roasted the second half of the beans when I got home that evening. Should have done a bit more advance research on the roasting times for these bean types - some varieties roast faster than others. Smoke is an inherent problem for those of us who roast indoors (and have to do so during the long cold winters). If you have a good vent hood above your range it might serve the purpose for smoke control but I roast near an open window and use a small box fan to draw the smoke away. When the Alpenroast is put into the "Cool" cycle a huge cloud of smoke emerges - see pic below. Also note the color variation on the roasted beans - this was because I had some Yemeni beans in the mix, which tend to roast unevenly, and some Robusta, which takes a bit longer.

This blend ended producing abundant crema and had not even a faint hint of bitterness. The deep, rich, chocolate undertones were a pleasant surprise but it had no fruity notes - I'll have to work on that.

Latte Art #1 from this batch - if you have young kids you know what this is - a profile picture of Squidward, from Sponge Bob Squarepants (Tom Kenny, the voice of Sponge Bob, is from Syracuse. That means I can watch the show - what's your excuse?)

Latte Art #2 - my hearts are getting much better but I'm still not getting rosettas.

[MGLloyd]

Owen, it is interesting to see the roast color variation from your Alpenrost as compared to some of my multi-bean blends done with the heatgun\dogbowl. Bearing in mind that I really like dry-processed Africans and Indonesians, I think that your roast came out more uneven then a lot of mine. How very interesting.

I don't think we can draw definitive conclusions unless we were roasting exactly the same blends in the same proportions, but I am heartened that I get such even roasts from my technique as compared to the 'set and forget' Alpenrost.

But then again, some blends are notorious for roasting unevenly. I have never in my life gotten a completely even roast on Liquid Amber, despite using a HWG, HWP, popper, Whirly-Pop, Rosto and heatgun\dogbowl.

[phaelon56]

Interesting that you mention Liquid Amber - it is the only blend I've ever roasted that has resulted in multiple hues / roast levels, apart from the one I just did for these pictures. I generally roast blends of beans that have comparable roasting times or do a binary blend - roast the slower roasting beans together and then the faster roasting beans together before doing the final blend. I neglected to check roast times on these before I threw them together and started roasting.

Just curious.... do you get a totally even appearance with Yemeni beans if you go just into second crack? I think the variegated appearance is either from the Robusta or the Yemeni but not sure which.

The "set it and forget it" feature of the Alpenroast is a joke. I always, always, always set it to the max setting and set a countdown timer for 13 minutes. I return when the timer rings and begin listening for the cracks and smelling the smoke.... roasts usually terminate at about the 16 - 17 minute mark - I always terminate manually at the desired crack point. The Alp does do a good job with chaff control - extremely helpful for those of us who must roast indoors in the frigid Northeast (where I live winter starts in late November and continues through late March!).

[melkor]

Interesting that you mention Liquid Amber - it is the only blend I've ever roasted that has resulted in multiple hues / roast levels, apart from the one I just did for these pictures.

I roast a pound of Liquid Amber almost every week in a gas grill drum - how about we all post pics of our next batches of Liquid Amber? Alp, HGDB, and Grill Roaster... anyone roasting Liquid Amber in a HWP or other air roaster?

[phaelon56]

how about we all post pics of our next batches of Liquid Amber? Alp, HGDB, and Grill Roaster... anyone roasting Liquid Amber in a HWP or other air roaster?

It's a deal. What level do you roast to? How far into second crack? I may be doing it before next week. I brought half of my just completed roast into the office for use in our little shared superauto. Amazing - every person who tried the coffee thus far said... "Wow - what's this? It's the best coffee we've had yet from this machine.". Just goes to show the power of home roasting. Previous batches included several varieties from our local microroaster (decent) and some of Wegman's store brand beans (horrible).

I burned through a fair amount of it myself this wekend as it was so tasty I just couldn't seem to stop drinking it.

[melkor]

how about we all post pics of our next batches of Liquid Amber? Alp, HGDB, and Grill Roaster... anyone roasting Liquid Amber in a HWP or other air roaster?

It's a deal. What level do you roast to? How far into second crack? I may be doing it before next week.

I go about 30 seconds or so into second crack

[MGLloyd]

Just curious.... do you get a totally even appearance with Yemeni beans if you go just into second crack? I think the variegated appearance is either from the Robusta or the Yemeni but not sure which.

In my experience, Owen, I think that most dry-processed beans are more prone to unevenly roasting regardless of method. I have gone to roasters here in the Seattle area and bought small amounts of Harrar, Yirg, Mocha, Sumatras and PNG, all of which were dry processed. I was fascinated to see how uneven some of the roasts were, even fresh out of the Probat, Dietrich, San Franciscan or the Sivetz. What was interesting to me was that the Sivetz, being the big fluid bed, was probably the most even of all.

So I am of the belief that some beans, and some blends, are apparently immune to even roasting no matter which way they are roasted.

And after much trial and error, I have found that my wife and I prefer our Liquid Amber taken just into second crack: probably about 20 to 30 seconds. Taking much further than that seems to kill off a lot of the taste, in our view.

I will try to remember to take a picture of my next Liquid Amber roast, which will probably be within the next week.

[phaelon56]

Mike - by the way.... I forgot to PM you about the dog bowl / heat gun pics email that you sent me awhile back. I had an "event" in which everything in my "INbox" was accidentally deleted - I forgot to email you and request another set of the pics to post. I'l still be happy to post those if you like but I also just finished posting a thread on how to post pictures with Imagegullet - for the non-technical user.

It's part of this thread

Attaching a picture to a post or reply

It appears that we all do Liquid Amber to about the same roast level. I'll do mine later in the week and post some pics.

[MGLloyd]

Or as an alternative, the most recent version of the heatgun\dogbowl roasting primer (with pictures) is maintained on Ed Needham's site at http://www.homeroaster.com/heatgun.html

Edited March 29, 2004 by MGLloyd (log)

[melkor]

The pic I took of the beans came out kind of strange, the beans aren't quite as red or as uneven as they look here... but whatever - here's this weeks batch of Liquid Amber:

[Jay Francis]

Re: Hot Air Popcorn Poppers.

The best one I have ever found (second hand store price $5) is the Wear-Ever Popcorn Pumper. This one has ribs that will swirl the coffee beans during the roast. Typical roast volume is three tablespoons of green beans and a 7 minute roast. Warning. I once bought one that produced so much heat that the coffee beans were burned in minutes. It's identifying marking is that it has a metal slide out tray for melting butter.

[phaelon56]

It appears that I won't be able to show pics of Liquid Amber for now - my stash is already gone. Anyone here know of a clone recipe that might approximate it?

In the meantime I threw together something with Brazil Prima Qualita decaf, Mexican Altura Tollan organic, Yemen Moka Haimi, Monsoon Malabar Coehlo's Gold and a few stray ounces of Aged Sumatra. It ollks and smells fantastic but I'll let it degas utnil tomorrow before sampling.

Did not use Robusta this time and I really believe that it's the primary reason that there are distinct color variations in Liquid Amber. The last roast I did (the one picture in this thread) had Nanga Farms robusta. That's one of the only two robusta's that SM's tends to carry and may well be the one in Liquid Amber that needs longer to roast. I did still have a few beans that were on the light side that were culled after roasting - I believe they were the Yemeni but there were only a few vs the many that I found when using Robusta in the blend.