Hello everyone.

I have been an active member on this website for sometime. I was an espresso enthusiast until I decided to take my espresso hobby a bit further and actually work in the coffee business. I am taking a break from my Masters for six months now.

It all happened after I attended the coffee and tea convention which was held in Dubai, based on the suggestion of one of the members here, which was great advice. I attended SCAA courses and became a qualified level one barista. I also got to know an owner of a local roastery in Dubai who was able to secure me a job as a barista in the coffee bar of the Gourmet section of a fancy supermarket named Galeris Lafayette (I know how that sounds like but it is a unique opportunity for me for many reasons).

Fortunately those people are serious about the quality of the coffee in their coffee bar (and fortunately they have little knowledge and a lot of extra equipment such as grinders, I have 4 now). They purchased a used, Toper 1KG Turkish roaster which I ended up using for the past 3 days, and I frankly feel very lucky to be doing this. To be the one who seemed to know most about roasting in a place where the owners want to roast coffee for espresso and produce great coffees. There are also not much customers which means a lot of experimentation both on roasting and on extraction... This to me is coffee heaven!

The most reliable way to control the roasting profile on this roaster is to leave the element turned on all the time (one of the original heating elements is disconnected to prevent scorching the beans and hence this does not make the roaster very hot). Then, the airflow valve at the top can be used easily to control the transfer of heat, very reliably. This eliminates the unreliability resulting from the very slow response of the electric heating element in case switching the element was to be used to control the roasting profile. Doing so, you need to "think ahead all the time" which is very difficult and very unreliable. This was the way the owner of the roastery taught me to use this roaster, apparently he did not to figure out that the profile can be reliably controlled simply by controlling the airflow. He also advised me to ensure a steady rise of 5 degrees celsius per minute throughout the roasting process, with no differentiation between before and after first crack.

I have been experimenting with 250 gram batches with great success in implementing any profile I can think of, yet my basic question is quite basic: What are the rise rates that I should implement for a decent espresso roast? I have been experimenting based on the advice of the roastery owner and all my roasts taste really bad. They are very gassy and taste kinda awful.

Also, is there any advantage of trying to elongate the warming up phase (< 150 celsius) for more than 3 minutes (since the moment I drop the beans)? Cause that does not seem to have a favorable effect.

My goal right now is is quite simple, which is to turn out drinkable espressos. The greens provided to me are a single-origin fair-trade coffee from Peru. First crack starts around 170 celsius on this roaster.

Since my question is very basic I expect that the answers will summarize neatly the lessons you guys have learned from your long roasting experience. Right now I am only at the stage of producing drinkable shots. I really wanna figure this out. I am hard working and dedicated I just need guidance.

Thank you in advance.

I'm not a roasting expert, but congratulations on your new job!

Congrats on your job, also.

I would suggest that you start with a very basic sort of profile which is "events-based," and then learn to correlate the "events" with what you see on your thermometry. This approach is the most likely one to yield good results, in my judgment, when working with a new roaster. I will have the chance myself to test this theory, when I take delivery of a new commercial roaster in 4 to 6 weeks

What are the "events" that will be the most useful to you in getting a good roast product, you might ask? You don't want to "charge" the drum with beans at a temperature that will produce scorching and "divots" and other obvious signs of having loaded a drum when it was too hot. You also do not want to charge the roaster at such a low temperature that you can't get to first crack quickly enough. I don't know what that temperature will be on your thermometry, but it should be easily discovered after using the roaster a few times.

Given a good starting temperature, you should attempt to hit first crack before 9 or 9.5 minutes; anything much beyond this is suboptimal. Hitting first crack before about the 8 minute mark probably indicates a roast that is going too quickly and that risks scorching the beans, and/or roasting the beans unevenly in that the exterior could be more roasted than the interior, and you want to get the beans to roast evenly from surface to the center.

After reaching first crack, you should attempt to finish your roast in under 15 minutes, and preferably earlier (I personally like 13 to 14 minutes as a goal for reaching the end of the roast). Where you end the roast is of course a question of what degree you want to roast your beans to. For espresso, I would attempt to get at least 4 minutes and certainly no more than 5 minutes between the onset of 1st crack and the end of the roast.

If you proceed with this sort of general "events-based roadmap," you can then confirm that you have not scorched or otherwise damaged the beans, and you can take note of the thermometry readings that can guide you on future roasts.

I think it would be a mistake to proceed in the opposite direction, e.g. to start out basing everything on the thermometry and then trying to force the cracks to conform to what you see on your temperature read outs. This latter approach is much more likely to result in frustration than it is to result in good tasting roasts.

Good luck.

ken

Hello Orwa.

Congratulation on the job! This sounds like a very good opportunity for fast growth in the coffee world!
I also just purchased a used Electric Toper Cafemino (1 Kg). I have run a few batches with it with mixed results. I have a lot to learn on using this machine. I am more than willing to share with you and possibly learn from you as well.

First thing that I learned is that the thermometry, as Ken alluded to, is completely different in the Cafemino than it is for other roaster, specially the old Poppery 1 that I was using. And probably from your Quest M3.

Now that you know that I do not know much more than you take what I write with whatever scepticism that you feel appropriate and feel free to correct me as required. This last statement applies to anybody reading this post, not just Orwa.

I still use the P1 for smaller batches for blends. I have pretty much got to a point where I can get whatever roast profile that I want by controlling the heat input through a Variac. Changing the heat input almost instantaneously is easy on a fluid bed roaster. Not so with the Cafemino, and I assume all drum roaster. There is significant lag as you have noticed yourself. And there is even more lag with an electric roaster than with a gas roaster.

One advantage of the Cafemino over other electric roaster is the fact that the element is inside the drum. This reduces the response time of an increase in power effecting the coffee beans.

This internal heating also introduces another variable in the mix. A large portion of the heat energy that the Cafemino uses to roast is Radiant heat. That I am aware of, the Electric Cafemino is unique in this that it has a thick solid drum that provides conductive heat, an effective fan that provides some convective heat and a very large radiant surface inside the drum that radiates directly on the coffee beans.

It is, I believe, this combination of heating attributes that makes the Electric Cafemino a somewhat tricky roaster to learn on. Radiant roasting is very different than conductive or convective roasting. I am no expert but Mr. Boot seem to indicate that radiant roasting are more tolerant of longer roasts without baking taste "defect". This, we will learn and verify together I guess...

I am already preparing follow up posts. I just need to put some order in my thoughts. I was planning on posting this after more experimentation but since you are in a similar situation as me...

To address your current concerns and statement;

orwa wrote:The most reliable way to control the roasting profile on this roaster is to leave the element turned on all the time (one of the original heating elements is disconnected to prevent scorching the beans and hence this does not make the roaster very hot).

First, it is useful to know the power of the elements that are installed in the roaster. You say that one is disconnected. From this, I am guessing that the roaster was fitted with the 28KW elements. Mine is equipped with a pair of, approximately 750 W each, elements that are mounted in parallel giving a total of about 1500 Watt total. It is important for you to know the power of the elements that you have to understand what is happening and to appropriately compare results with other users. For this you simply have to measure the electrical resistance (Ohms) of the element with a meter and calculate the power with it .
Element Power = (Input Voltage^ 2)/ resistance
If the element is too powerful and the roast is controlled with the heat element only than it turns on and off all the time and this can make the roast too long. If the element is very powerful and it is left on all the time, the heat input could indeed be too fast and scorch the beans and make the roast too fast.

orwa wrote:Then, the airflow valve at the top can be used easily to control the transfer of heat, very reliably. This eliminates the unreliability resulting from the very slow response of the electric heating element in case switching the element was to be used to control the roasting profile.

I am assuming that by leaving the elements on all the time, you mean that you set the temperature controller at a high temperature that prevent switching on and off. This could be risky is you set it too high. Or as you seem to have experienced the element is too strong for this.

Opening the air vent does produce a very fast environmental temperature response (drop) but the effect on the beans is not necessarily what you think.
As you know the installed thermocouple is a sort of Environmental temperature sensor. I say "sort of" because it is very near the beans and with a full load of beans it actually comes in contact with the top of the beans.
I have installed a thermocouple in the bean mass and one outside the drum. I have observed with these that opening the air vent reduces the environmental temperature almost instantaneously and increases the bean temperature a little with a little delay. I theorise that the reason for this is that by increasing the air flow a lot of the heat from the drum and the air around the drum is sucked-in through the drum and imparted partially to the roasting beans. Of course after a short while the bean temperature will also decrease along with the environmental temperature. Closing the vent will result in an increase in temperature but is very slow compared with the drop that is achieved with opening it.

orwa wrote:Doing so, you need to "think ahead all the time" which is very difficult and very unreliable. This was the way the owner of the roastery taught me to use this roaster, apparently he did not to figure out that the profile can be reliably controlled simply by controlling the airflow. He also advised me to ensure a steady rise of 5 degrees celsius per minute throughout the roasting process, with no differentiation between before and after first crack.

I think that thinking ahead is something that you, I and anybody roasting coffee must learn to do to achieve good results. There is no way out of that one, I believe.
I do not know what the results of the previous owner were but I do not think that his advices are very useful... A bit too simplistic, very much too slow a rise rate and having tried it results in flat baked coffee taste.

orwa wrote:I have been experimenting with 250 gram batches with great success in implementing any profile I can think of, yet my basic question is quite basic: What are the rise rates that I should implement for a decent espresso roast? I have been experimenting based on the advice of the roastery owner and all my roasts taste really bad. They are very gassy and taste kinda awful.

There is a well documented belief that the Electric Cafemino works better with full 1 kg or near full 750 grams loads. I do not fully understand nor believe that this is true. For my part I have been trying to make it work with 500 grams loads. This is the load size that is most convenient for me at this time and I find that I can get more control over the profile.
My brain has not absorbed the "roast profile - coffee taste" correlation knowledge yet. Other will be better than me at diagnosing your taste results but I believe that the grassy taste could be a result of a too short drying phase.

orwa wrote:Also, is there any advantage of trying to elongate the warming up phase (< 150 celsius) for more than 3 minutes (since the moment I drop the beans)? Cause that does not seem to have a favorable effect.

3 minutes seems very short to me. The point at which the drying phase is complete is not so easy to determine for me and it varies a lot with different coffee origins and drop-in temperature.

More to come, if anybody cares...

Of course we care, I care

To be honest I have been assuming something which is that the temperature reading I am having on the machine is more expressive of the bean temperature than of the environemental temperature. In fact, taking into account the response of that reading to events such as dropping the beans in and modifying the airflow strongly suggests that it is a target temperature or a bean temperature reading, not an environmental emperature reading.

Today my boss informed me that he purchased a 60-KG (!!! I know!!!) bag of the washed, fair-trade Peru beans that I was experimenting with last time. I told him that I think this might be a huge quantity for our coffee bar and that we better experiment with many coffees before purchasing such a large quantity of one bean and he said that he does not need me to teach him how to run an F&B business. I laughed inside at how offended he was by my comment, it was humurous in a way :p

Then I started telling him that "if you want custmers to smell great coffee aroma" then I think that we should bring retail grinders and start selling and grinding beans because without grinding the aroma will not be released and then he said again, "no that is not what I am after, I just wanna roast coffee". So he DOES want to roast coffee regardless of why he wants to do it, and I get to experiment so lucky me, everyone is happy, haha :p

Thank you guys for congratulating me and helping me, you are very nice.

Today I did a single roasting experiment, using a batch weight of 525 grams of the green bean I just mentioned. My goal was to implement the "9-9.5 min till first crack, 4-5 min till end of the roast" plan. I wanted to see the effect of this simple plan on the quality of my roasts, and though I cannot be totally sure before the beans rest properly, I think the result is amazing! I mean, I extracted a decent (aka perfectly drinkable) espresso today using the beans I just roasted. I was very satisfied.

I did some calculations to ease controlling the roast. The first crack on this roaster starts at 180 celsius and rolls at 185, hence I am assuming that the average temperature for the first crack is around 182 celsius.

If I preheated the roaster to 150 celsius with no airflow and then dropped the beans (with no airflow still) the temperature of the beans drop to 115 and then starts to rise by itself (no airflow still) at the end of the second minute. Then I can use the airflow valve like I said to reliably control the transfer of the heat (the bean temperature profile).

Ken said I should reach first crack in 9-9.5 min meaning I should work towards a rate of ([180|185]-115)/([9|9.5]-2) which equals 8.6 to 10 degrees celsius per minute, depending on whether the temperature of the first crac is taken to be 180 or 185 and whether the time to get there is taken to be 9 or 9.5 minutes.

Then from 180/185 to the end of the roast, let's say at 200/205 celsius, if I want to take 4-5 min then I should work towards a rate of ([200|205]-[180|185])/[4-5] which equals 3 to 6 degrees celsius per minute.

As I noticed it is enough to get to 200 celsius on this roaster/bean to give the bean a nice appearance, hence I am tempted not to increase the temperature more cause I kind of panic from charcoal tastes, unless you tell me it is benefecial for the overall flavour of the roast.

I tried to implement the above rates in my roasting experiment today but I was somewhat distracted and also nervous due to me being new in using this roaster. I wrote a small log of my roast to show it to you:

LEGEND: [minute: temperature in celsius]

0.0:150
0.5:130
1.0:120
1.5:115
2.0: 114/115 (at the end of this minute the bean temperature starts rising without applying any airflow hence I concluded that the inside of the bean was warmed and I start the steady rise by applying airflow)
3.0: 125
4.0: 136
5.0: 145
6.0: 154/155
7.0: 163
8.0: 167
9.0: 175
9.5: 180 (first crack starts here)
10.0: 180
11.0: 185 (first crack rolling here)
12.0: 191
13.0: 198

end of roast is at 13 min 40 sec with a bean temperature of 205 celsius and a rolling second crack (I did not notice when the first crack ended and the second started, I guess I did not streatch that time enough but I could not slow the temperature rise even with the valve fully closed).

I would like you to criticise my roast.

The main lesson I learnt from this roast is that I should turn off the elements when the first crack starts, I think the momentum is enough to sustain the roast until the end at 200-205 celsius. Not doing so made slowing the roast impossible at the end in this case. I wanted to have 4 full minutes between the setting of the first crack and the end of the roast.

The previous owner is the one who disconected one of the elements, he said that originally the beans were easily scortched. I like the fact that this way, I can leave the element on at the beginning of the roast and only concentrate on adjusting the airflow to implement the profile I want, so I think I will leave it this way.

Hello again.

150°C shown on the Roaster display is where I also drop-in the beans most of the time these days. I have dropped-in at as high as 180°C without any sign of tipping or scorching.

Clearly the turning point temperature and time depends on the beans being roasted but also the temperature of the drum and the ET. It is also largely influenced by the power of the heating element.

I must reiterate that the display temperature is from a sensor that is not in contact with the beans in my roaster. Is you thermocouple installed at the factory position? If so, than, certainly the thermocouple is well above the bean mass with only 525 grams of beans.

You can use that sensor but be aware that it is not the bean temperature. Rather, as someone else wrote elsewhere, it is an indication of the heat energy being applied to the beans. With a small batch weight and the element always on (no cycling) this temperature seems to follow the bean temperature somewhat. Just be careful with it.

When measuring the bean mass. I have found that the beans temperature goes down much lower than what you see on the display. With 150°C drop-in, 500 grams of Brasil Cerrado, the temperature goes as low as 80°C before turning. Also a temperature sensor installed just above the drum to measure ET shows that it has very little effect on the ET with only about 10°C drop.

All this with the air vent valve fully closed. Or at least as closed as it is possible. If you open the exhaust tube you will see that the valve has an aperture that prevents fully closing it. A trick to reduce the air flow even more is to open the cooling bin valve. This will reduce the air from the drum to almost nothing but still not exactly nothing.

Rolling 2nd crack on this roaster never gave me burned taste or charcoal taste. It may have to do with the radiant heating of the beans which does not seem to concentrate the heat on the surface of the beans as much as conduction does. Roasting with the Cafemino is very different than with more conventional drum roaster.

If the roaster is well heat soaked you can indeed cut the heater off for a while after first crack but be careful because when the temperature starts going down on the display it is too late and even if you turn the element back on it will take too long to heat up again. I lost a few roast that way.

It sounds like you are doing well. Keep on trying. You now have 60 kg of the same type of bean to learn! Roast them.

Hi guys, I don't have much time to write at the moment but a couple of things spring to mind.

1. I find longer roasting times seem to work better for me on the Toper. Around 18 minutes is what I usually shoot for, but even 20 has been ok. Drying phase around 7 or 8 minutes with a 1kg load. Anything faster for me has resulted in a less enjoyable roast. This seems longer than most other roasters, but perhaps it is because of the heating configuration. I believe Diedrich roasts are typically longer than Probat roasts, so this is perhaps plausible. My drop temp with 1 kg is 230C BT probe, which translates to about 180-190C on the stock ET probe.

2. One element disconnected... earlier Toper model had this as selectable, apparently it was for roasting peanuts. Mine still has two lights, for each loop of the element, but they are not selectable. My roaster has been modified with PID, but I still control my roasts manually by setting a power level. Airflow is controllable, but the impeller is pretty gutless and if you disassemble the machine, the damper doesn't look to be very effective really. Unless your element is dramatically different from mine, you will have no chance of being able to roast 1kg successfully without reconnecting it.

3. First crack if you slow the drying phase down will be 190-200C BT, which I think is more in line with expectations than 170C which to me at least says too fast drying phase. 2nd crack for me generally around 220C BT.

4. Unless 250g is what you think you'll be regularly roasting, I suggest a bigger load. The machine behaves VERY differently closer to 1kg. When you do switch to a larger batch size, be aware there is a lot more thermal inertia. I'd need to check my wattages, but generally I think the roaster is a little underpowered for 1kg.

5. I've been roasting on mine for 3 or 4 years now I think. Good little roaster, built like a tank, but it certainly has its quirks and I doubt its in the same league as Probat, Diedrich etc. I also find the bean circulation stalls a bit around the viewing window on larger roasts, so it needs a little bit of manual agitation there to prevent an uneven roast.

6. Off to drop some beans in now...

Congrats on the job, Chris

Hello Chris.

Good to hear from other users.

When you mention BT, where did you install this sensor? Mine is going through the chute door, where there is (was) a bolt. Before I drop-in any beans the temperature of my BT sensor is within a few degrees of the stock sensor. Which I have come to call the Internal ET.

I also think that the Cafemino does better on longer roast (based on my very short time with this machine), with longer than average drying phase and not so aggressive ramp-up to 1C. When speeding through any of these 2 phases made for uneven roasting. 1C would start early and be very long with 2C not long after. In these cases, when the roast went too fast, I leave it go deeper into 2C with the heater off to even it out a bit.

One intriguing statement you make:

gyro wrote: I also find the bean circulation stalls a bit around the viewing window on larger roasts, so it needs a little bit of manual agitation there to prevent an uneven roast.

It is not always that much of a concern but with chaff heavy beans, it does pack up on the window. How do you provide this "manual agitation"?

I have done 9 roasts so far with the same bean and the only thing that seems to be clear to me so far is that very short roasts taste much better. I am talking about something as short as 10:30 to 11 minutes