I can very easily slow down the roast by lowering the heater power, and can very easily stay at relatively low temperatures while doing so since the first crack occurs at relatively lower temperatures when someone is going slow. However, no matter how low temperatures or low heater power I used, roasting for more than 6 minutes in total seems to always result in an ashy bean. They look light, but they smell and taste a bit like charcoal, which is the single one thing that I cannot accept in coffee. I almost gave up on this roaster, until I tried roasting in very short total times (going as fast as the highest element power/fan speed permits for a while and then slowing down at the beginning of the first crack). This way was the only way I could find to produce the sort of bean I wanted. According to my experiments, ashy tastes evolve after a certain roasting time no matter what the element power or the temperature at the end of the roast was, it seems that the ashy taste results from depleting the water content in the bean (?), which seems to be happening quite quickly on this roaster. Note that the new roasting chamber I am currently using has more perforation than the one shown in the photo (but is the same fabricated one rather than the stock one).

orwa: I don't have any useful advice or suggestions, but I just wanted to say I've been following this thread and am impressed with what you guys are putting together. Great job!

orwa wrote:However, no matter how low temperatures or low heater power I used, roasting for more than 6 minutes in total seems to always result in an ashy bean. They look light, but they smell and taste a bit like charcoal

Old greens? Dry greens? Bad greens?

What are you roasting? Did you try different, current crop, greens sourced from someone with a good reputation(e.g. SM's)?

The greens I currently have are the Yemeni and Ethiopian coffees generally available to me. I also have green beans from Brazil (Santos), Nicaragua, Nepal (which are the ones I like), and Mexico. I don't think that this is related to the beans as the observations seems to be consistent amongst the different coffees I have. Roasting on this thing proceeds quickly. It seems to me that a good roast is only from 3:00 to 4:40 minutes in total (from the moment I turn the heater on).

How much time is usually needed to properly dry a large bean prior to starting the actual roast at 350 degrees Fahrenheit (with no resulting defects in the final product)? In my case I think that this can be done in 1:30 to 2:30 minutes regardless of the size of the beans (Yemeni's are incredibly small, Nicaraguan's are incredibly large).

Note: Some of the green beans I had for 6 months, some for almost a year kept inside the house. Others (Nicaraguan and Mexican) are recently bought (this summer) from a local roaster. I don't know for how long he has been keeping them.

orwa wrote:..............
It seems to me that a good roast is only from 3:00 to 4:40 minutes in total (from the moment I turn the heater on).

How much time is usually needed to properly dry a large bean prior to starting the actual roast at 350 degrees Fahrenheit (with no resulting defects in the final product)? In my case I think that this can be done in 1:30 to 2:30 minutes regardless of the size of the beans (Yemeni's are incredibly small, Nicaraguan's are incredibly large).
..............................

Orwa, these total roast times are so far from anything I have found workable, I don't have a clue on what the issues are with your roast. (it is completely off my experimental chart!) On the drying times, there may be some issues with your 6 month bean storage (what is the relative humidity?) Maybe they are all dried out from low humidity storage. With high heat drying temps though I have been able to get drying done in 2:30 minutes, but why not run a bit slower and avoid the risk of scorch?

I would suggest you get a 'bare tip' probe and put it in that small space between the heater and your roasting chamber (right below the vents you have cut). It could help to diagnose what is going on in this lightning fast roast. (I'm guessing the temps are 480-500F+ to finish a roast that fast; too hot) If you get the temps right down here, the roast times will fall into place. What are your batch sizes now?

Frost wrote:What are your batch sizes now?

Very small. I think they range from 50 to 90 grams.

Correction: with the extremely small Yemeni beans I have, it seems that I can easily roast 120 grams at once (roasted coffee weight, correspond to around 150 grams of greens), which I think isn't precisely "very small" as I have said before. Larger beans don't at all allow this batch size unless manual mechanical agitation is used, which is something I don't like at all.

The roasting procedure that seems to work well for me is the following (I am still experimenting, but this seems to give better results than the many other procedures I tried):

Conventions:

• - Starting times of different roasting stages are expressed from the moment I turn the heater on, which are enclosed within square brackets.
  - Temperature measurements are all bean temperatures. I have the tip of the thermocouple standing at the centre of the roasting chamber/bowl with a standing length of around 3 millimetres, ensuring contact with the bottom of the circulating coffee mass (away from the hot air blown from the side vanes). This may not be purely indicative of bean temperature, but is the type of measurement I use below.
  - Temperatures in Fahrenheit are expressed right after the temperatures in Celsius, which are separated by a slash (i.e. "Celsius/Fahrenheit"). I use Celsius because my multi meter gives temperatures in this unit.

1. [0:00]:

• - Ramp to 160/320 degrees in 45 seconds with full fan speed.
  - Stick to that temperature with full fan speed for around half a minute to obtain a yellowish appearance of the beans (could hardly extend to a minute). Note: My drying time is not an overkill, but I think is good enough; An optimum drying however wouldn't take more than an additional half a minute.

2. [1:30 to 2:00]:

• - For some reason, at this point I can't resist lowering the fan speed a bit and increasing the element power. This results in slow but very hot air to pass through the beans. This causes the bean colour to go matte reddish and a sweet (I don't know, nutty?) smell to be emitted. This process takes only half a minute and reaches up to 180/356 degrees. I am unsure whether this step has a favourable effect (if any) on the final roast... It's only driven by unfounded hunch.

3. [2:00 to 2:30]:

• - After that I rush to the first crack with a high (not maximum) fan speed, causing the first crack to reach its peak at a temperature reading of 215/419 degrees. Note that it is also possible to use a cautious ramp instead of an aggressive ramp at this stage, by gradually slowing down the ramp after crossing the 200/392 degrees point, which I achieve by lowering the fan speed and not the element power. This supposedly allows the interior of the beans to cope with the exterior thus allowing the bean to crack at a lower temperature (i.e. a lower temperature of the bean's exterior). The utilisation of such a cautious ramp leads to the peek of the first crack to occur at 210/410 degrees. Both ways of reaching the first crack are possible, and in either case, it seems to me that it's only desirable to reach those temperatures (210/410 to 215/419) for as long as it takes for the first crack to pass, but not longer than this. Staying at those temperatures for longer times seems to cause an inevitable ashiness in the final product. Passing to second crack, on the other hand (no matter how fast) seems to result always in charcoal-ness, which is a defect more severe to my taste buds than ashiness (this is only a matter of taste). The ramp to the first crack as well as the first crack itself take around a minute and a half to complete.

4. [2:30 to 3:30]

• - After the first crack finishes, it seems that it's beneficial both to keep the roast going to develop the flavour and in the same time to avoid two things: high temperatures (higher than 205/401 degrees) and high fan speeds. Luckily enough, at this point the first crack is done with and so the beans are soft enough to be ground satisfactorily, and light enough to be circulated using very low fan speeds, which allows me to back off to 205/401 degrees for at least another half a minute with a very low fan speed. This seems to have a very good effect on the resultant flavour/body (Otherwise both are much less than what I am used to with oven roasting).

Roast is finished in 4:00 to 5:00 minutes depending on how long I stay at 205/401 degrees point. I also made some exotic roasting experiments by staying for as long as eight additional minutes at lower temperatures such as 185/365 or even 170/338 degrees with very low fan speeds. The latter case results in a slightly astringent espresso that pretty much has the classical taste of an Italian espresso (Like the ones I would have at Costa cafe for example, with a nutty flavour and a long, slightly astringent aftertaste). All in all, I can say that it seems to me that this final, low-fan-speed curing is necessary to develop flavour/body while avoiding ashiness. Stretching a traditional, monotonically-increasing profile on the other hand doesn't seem to work at all for me (it could a taste issue, a roaster peculiarity, or both).

The procedure described above is not totally arbitrary, but is not the result of systematic experimenting either. It's a mix of hunch-driven choices and choices guided by casual experimenting/tasting. I would like to hear how this procedure compares to that of skilled home roasters, and whether they can relate to it. Is it usual to stay at relatively-low temperatures for extended periods of times with low fan speeds at the end of the roast? All the profiles I have seen in this forum were monotonically-increasing, but this one isn't.

Documenting the last changes and minor additions:

This is how the roaster looks like in the moment with an attached multimeter that serves as a temperature display (can display temperature in both Celsius and Fahrenheit, which was highly beneficial since most of the roasting experience in this website is recorded using Fahrenheit).


This is the tactic I used to insert the thermocouple, it is passed tightly through the joining point of the two ends of an asbestos strip that I used to secure the gap. The joining of the two ends is using regular staples which are directed so that the thermocouple wire can be passed adjacent to the roasting chamber. I still use the o-ring for a better seal which is now shielded from the intense heat by the asbestos sheet.


This how I insert the tip of the thermocouple into the roasting chamber. Note that the standing length of the tip inside the chamber is around 3 millimetres, which I hope is enough to touch the mass of circulating beans, thus giving a reading of the "external bean temperature".