The Golden Bean 2013 - Coffee Roaster Competition

[Barry O'Speedwagon]

I'll try that next weekend, thanks for the suggestion.
Yeah, I've been reading the comments on roasting, especially where it's for a Gene. I found the judging notes a bit confronting, but I guess I'd rather have an honest opinion even if it isn't all that complimentary. I didn't get the baked comments, but the discussions on that have got me wondering about whether to change my roasting. I've been starting at 150 for 5 minutes, then 250 for 6 minutes, then 240 until stop, and when I stop I dump into a colander and then a flat tray, I don't use the cooling cycle. It sounds like I'm roasting a little hotter than others, plus I'm going by time alone until I decide to stop, but others are going by FC - I have to admit it's quite rare that I hear FC at all.

I usually don't have much trouble hearing first crack (may not be certain about the first one or two snaps), I have the large chaff catcher which may help amplify sound a bit? Anyway, one of the good things about the Gene is that you can see the beans in the roasting chamber. If you're having trouble hearing the cracks, you can at least observe changes in bean colour / shape which should help you get a better idea of where the bean is 'at' (Sweet Marias have a reasonable visual guide to the roasting process), and some beans give off quite a bit of smoke just b4 second crack. Also, if you weight the bean before and after roast you can get a fair idea of how dark you have taken them.

[Steve82]

Woot got some feedback today.

Total of 50

1: Great flavour, well roasted, complex toasty dark roast, nice coffee with a smiley face.

2: No upfront notes, pleasant mouth feel and a soft almost toasted lingering after-taste.
Cuts well through milk, overall enjoyable coffee.

Happy days

[trickydicky2]

Excuse me for the lag Tricky,

The blend I formulated can be found on reply #82

Doh! Remember commenting on it now

[trickydicky2]

I would follow the advice of Barry and Chris.
When I had a Gene it was; preheat to 150°C then 230°C until 1st crack then 225°C til finish.
By setting the temp at 220°C from 240°C I think the thermostat would actually fall below 220°C before it started to reheat but your bean mass
might continue to lose even more heat before stabilising. (????)

I would look at that tho' as your possible cause of baked notes.

I followed the general principles of this advice and... SUCCESS!!!

Roasted a nice SO Colombian old vs new profile. I had enjoyed the old version but, wow, the new profile is dramatically better. What instantly impressed was the clarity of both aromas and tastes. I got a nice balance of sweetness acidity and body and a very pleasing roasted flavour. The old one is muffled and generally less "pretty" in comparison. It's sort of like the difference between a slightly corked bottle of wine and an uncorked one

I decided to try to reverse engineer a roast profile of bean temperature and RoR. See attached spreadsheet/graphs

Basically I recorded the exit temp from the Gene every 30sec then used smoke and mirrors to:
- firstly guess the bean temp at each increment
- then calculate the differential of the exit temp and the bean temp
- then work out a factor to calculate the RoR by assuming the instantaneous RoR has a linear relationship with the differential between the exit temp and the bean temp
- oh, and the guesstimated bean temp was anchored against the ambient bean temp at start of roast and the presumed temp at first crack (~8 mins after end of drying phase)

This was an iterative process - I had to tweak the bean temp guesses a bit to make it look vaguely right. Hopefully, it's not too much of a case of wishful thinking

Having found a roast profile that tastes pretty good, I'm not sure I'm inclined to faff around with this model any further. However, it would be interesting to validate this profile by manually probing the bean temp at different points in the roast

Some empirical changes I will explore, however, are:
- running a slightly lower main roast temp to reduce risk of burning the outside of the bean (particularly with softer beans)
- going for a more S-shaped RoR curve during the development phase to stretch this out a bit further (reduce the thermostat temp more steeply at the start of the development phase)
- deciding if the drying phase adds any value (or hurt)

Not sure if I'm fully on the right track here, so some further comments would be great... . Thanks

roast profile 30–nov-13.xls.zip

[trickydicky2]

Final comment on gene roaster profile as it's way off topic for this thread

I seem to have got better tasting results than the last post with the following approach

preheat empty roaster to 220c. Charge with green beans and set to 245c. At first signs of first crack drop to 240c for 1 min then 235c and 230c for 30sec each then 225c until desired roast level achieved. This way I get to start of first crack in 8-8.5 min and get 2-3 mins development time roasting for espresso or 1.5 mins for a filter roast

this profile has no drying phase and is hotter and faster than say Barry's recommendation. The main benefit I have found is a cleaner brighter more intense flavour with no obvious baked character. However body is a bit lower, but still good due to the stretch at the end

I should point out that test roasts were done with harder beans eg El Salvador, Ethiopia, and a 150g charge. Will be interesting to see how it fares with further use, and at 250g when I roast my Christmas supplies next weekend

My theory for this profile was firstly to get closer to commercial roast speeds to see if this reduces baking, and secondly to have a high preheat temp to build a bit more thermal mass in the roast cylinder, knowing that glass is a poor conductor so hoping that a higher temperature would not scorch the beans

[malliemcg]

Just wanted to pop in and say that I did get my notes, and I have some work to do. Received notes 3/Dec, had an ancient tab with a half composed, unposted post.

I'm glad for the feedback, but really have to work a bit harder.