HotTop Test

Well the Ambex experiment fell through since the thermocouples weren't K type as advertised. So I managed to get my hands on a used HotTop roaster.
It worked the first time out. This image is from my 4th roast.
I'm not sure if it is common with HT, but I had trouble getting to the 2nd crack in a reasonable time. As you can see, it took a bit over 20 minutes to get there. I have some repairs to do on the HT, which may explain the heat problem, if there is one, due to heat loss. Regardless, I was able to get the information I needed and the profile program did its job.
1st Crack was right on schedule at 390*F, as was the 2nd Crack.
The only problem I may have is the slope is always multiples of 7.2*F. It could be the accurate slope, but I'll have to run a few different tests to see. Also, the graph sometimes seems a bit behind. The data is correct, I just think there is too much information to process for this program, Processing1.0.
Either way, if any one is reading this, enjoy.

Basically when you set the roast time on the roast, and hit start, it begins to preheat to around 300*F, drum/environment temp. The HT does have a temp reading, which read consistantly between my logged temps. I was able to submerse one TC in the beans and the 2nd higher in the drum. They were inserted through the bean drop with the lid over them, which could help explain heat loss. Regardless the temps seemed well place.
After the drum is preheated the roaster beeps and begins the countdown, this is when I dropped in the beans and started the graph. You can see the Bean Temp sank to bottom out at around 183*F. Then a steady climb to finally hitting first crack at 17min.
Next roast I'll try a smaller batch to test. This particular roast was set at 20 minutes with 250 grams of coffee.

Stay Tuned For Ambex 2K Test

Well, I haven't been working on my project lately. I'm preparing for my move back to Philadelphia, where I will be testing my project with a friend's Ambex YM-2. Its already set up with a profiling program, but I will be attaching the existing K-type TC's to my unit. I will be roasting and logging all of my roasts and post them here, mid April.

This should be a good test of its versatility. I can compare my results to the built in profiling system of the Ambex. If it can work with a commercial roaster, than there shouldn't be any problems with a home roaster. Actually if it even works with an oven, it should be good. Considering its just measuring temperature over time.

Fresh Roast test

I had to work out a couple of bugs in the programing, but Its getting better. This is an image saved from the latest roast. Its a FreshRoast which is a small fluid bed roaster, so the roast is short. The probes were placed perfectly to get an accurate read on the temp. This has proven difficult with a fluid bed roaster, which blows hot air at a high speed, thus giving some random temp readings.

As you can see, I was able to record the cooling period as well. I saved the time the heat was shut off, but allowed the graph to continue to monitor the process. It was not until he fan shut off that I clicked the "stop" key, ending the graphing/roasting sequence.

Work In Progress

I've been roasting coffee at home for the past couple of years. The more I've learned about roasting, the more I wanted to know the "science" of the roast. Basically, I wanted more information/data to be able to repeat my roasts or at least give a visual to what I'm doing. This is nothing new, but all the solutions I've found costs hundreds to thousands of dollars, depending on the extent of info or roast capacity.
So, I decided to make something of my own. I was able to build a data logger that transfers temperature data, real time, to my PC or laptop. The next step was to take that data and put it on a graph that also allows you to document different roast stages. Well, as it turns out, the datalogger was the easiest part. The software is a whole other beast. Luckily after much trial and error, I was able to create a very useful program that does just what I want it to.

The goal is to be able to provide my "invention" to anyone for an affordable price. The software was created using an open source programming language, Processing 1.0. Which means that the software program is free and is able to be modified by anyone willing to tackle the language. From there I will be creating a more extensive program to accompany the device, that will be more of what people are used to, as far as user interface. But the bare bones program which I have created, will be more than sufficient to anyone who just wants a window to what they are doing when roasting coffee at home.

As it stands right now, the program graphs the curve of the roast, using 2 temps, the bean temp and environment temp. It uses 2 k-type thermocouples. So if you already have your factory roaster or even a home built roaster, set up with TC's than all you have to do is plug it into my data logger, and voila, you're ready to go. The bean temp and environment temp are displayed, there is a stopwatch that starts, stops and resets the graph. During the roast, the average increase per minute in bean temp is displayed, so you can see the rate of temp change. Also, there are 10 spaces to document the roasting stages. Once each roasting stage is met, you click on the adjacent "button" to record the time and temperature that it was reached and a corresponding dot appears on the graph. After the roast is complete, or when you stop the graph, it can be saved as a PNG image, according to the info typed in the "Roast Title" box.
As I continue to test and modify my project, I will post my progress here. Feel free to contact me to suggest modifications or ask any questions.