Typica introduction [Coffee & Tech!]

Posted: March 27, 2016 in coffee roasting, linux, postgres, tech, Typica

It’s been a while

Things have been super busy for the last year and not much noteworthy or mentionable to blog about in either my coffee or tech world. It’s been too boring to write about.

OK that is totally not true. There has been the excitement of fresh lots of coffee from Sweet Maria’s, and then there was the night when I had several people over for a roasting demo (will definitely do that again soon).  AND then there was our visit to the Diria Coffee Processedora in Hojancha, Guanacaste, Costa Rica. I definitely should have blogged about that. That was awesome.

But that was months ago, I don’t think I could capture all the excitement and joy of that visit. I’m sorry…

So here we are, and in the words of Buckaroo Banzai, “no matter where you go, there you are.”

So let’s start over. I’m Kevin, I’m a little nerdy and techy and love coffee. So the next several blogs (I’m envisioning several) will combine both Coffee AND Tech as I explore the Typica program for roast logging. This will be a win-win of instead of just writing about Coffee OR Tech.

Roast Logging and Profiling

I’ve always wanted to have more data from my roasts. I believe that I need to track a roast through temperature and activities in order to understand it. This is the beginning of making a roast profile. A roast profile is simply a guide to turning the beans brown in the best way possible. A profile could be simple, like a flat even climb in temperature until end of roast; or a profile can be complex with a couple of cycles of heating and reducing heat. The different approaches to  cooking the beans can change the beans flavor and improve enjoyment.

The reason is simple– the process of cooking any food create chemical reactions. The chemical reactions break down or create chemicals in the food items, coffee included. Some chemicals have a taste that may be desired and others perhaps we want to avoid. I’m far from an expert on this process but I believe that data logging can assist me in my hobby anyway.

My goal is to capture the data from roasting so that I repeat a great roast or avoid making a mistake again. This type of logging is extremely important to commercial operations who have a lot at stake every time they put some of their investment into the fire. They want to make sure they don’t ruin a batch, and always if possible, improve on a previous batch.

Home roasters can benefit from this as well. We may not have the same investment quantity of beans and as much at risk, but we can benefit from logging.

I have filled half a dozen notebooks with handwritten notes for the last nine years. I’ve also tried several methods of computer or tablet logging. They are ok, but they are difficult to refer to with any precision when it comes to repeatability. And the real challenge with any method of roast-logging or note-taking is also tracking cupping and tasting notes.

I first found Typica a number of years ago while googling. I was interested in the software, it was opensource and used a real relational database– postgres! These are exciting things to a nerdy tech-minded homeroaster! At the time I investigated it a little and even conversed with Neal the author, but decided not to invest in the expensive device it required. Neal required any hardware to be open-source friendly and at that time that was pretty much only the National Instruments expensive data logger device. I couldn’t justify the expensive hardware for hobby operations.

I tried a DMM (digital multimeter) with temp input and optical RS-232 for connecting to the computer. Logging this data was something, but cleaning it up to make a graph in Excel or Google Spreadsheets proved to be very time consuming and difficult. The paper logs continued until I began using an Android tablet and an app called Coffee Roaster.

The Android app was almost good enough for recording data manually and making graphs. I configured it for two temperature trackers and used one to track my burner levels and the second for actual temperatures seen at the DMM. I lost data several times due to upgrades and corrupted database. Through it all I kept at least basic written logs to at least keep track of my numbering ( I did roast #396 in the RK+Grill last week).

Typica Again

Fast forward to late 2015 and I looked at Typica again because of my recent ennui with paper and the android app. Imagine my surprise and joy to find that it can now use a Phidgets device which costs a fifth less than the NI devices. This is hobbiest territory.

I will begin using this but it will take a bit to setup the way I want. This is not quite as simple as just install and make a few configuration likes the Android app. This is a little more complex. But it doesn’t have to be scary. I hope my next few blogs can make it less complex for someone else.

I am using an old Mac Mini computer for a NAS and database host. I was going to use my new Windows computer to run the application but as I began experimenting and thinking about the risk of fire and heat on it, I think I might use an older laptop and/or netbook to do the logging. The logging computer may run Linux– stand by and read on as I discover the path forward.

I’ve already found a limitation that many home roasters will face and will work out a change in code to fix it. Neal created the software for commercial roasters. Commercial roaster can log the bean temperature which is ideal since they have a non-moving piece of the drum. They can put the temperature probe right into the beans. The RK drum and grill setup has one full drum that rotates completely, no fixed piece to mount a probe in the beans. The best I can do is to log the temperature in the chamber outside the drum. This at least gives me data about how hot it is with the burners going. The limitation I hope to patch sets the graph scale at a top of 500F, plenty of head room for a bean probe. But outside temps? Much too low, it should be double that.

Looking forward

As you can tell, I’ve already done a little bit of setup on this so I have an idea of what I can write about. Neal documents many typical scenarios on his site. However, I see where I can provide some additional information in the next few blog entries.  I will write with a little more detail about setting up a hosted/shared database, including how to do backups of the database. I have my backups being copied to my Google Drive account directly, so from database to cloud copy for offline and offsite. I will also attempt to document and write about setting up the compiling environment and my attempts to provide a more home-roaster friendly scaled graph.

Beyond that I imagine I will start trying to document the tastes of the roasts and see if I can improve my roasts through better profiles. We’ll see where this goes.

  1. Dan says:

    Hey I’ve really been appreciating your posts on Typica use so far. Did you build Typica for use on a Mac mini running Linux Mint? I’ve had problems trying to install Typica on any of my PCs running Linux. I get lost with making and building on qt. If you have any pointers I’d love to hear them. Right now I’m using Typica on an iMac but I’d like to transition to a fully FOSS setup.

    • kevincreason says:

      Dan, I actually have! That was going to be next blog before the Mini croaked. Now that I have my database up and running again in a permanent location (on a qnap NAS) I hope to capture those steps in my next blog.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s