Perfect Cheese Every Time With This Temperature Controller

Anyone who is from a background in which cheesemaking is a feature will tell you that it is an exact science in which small differences in parameters can make a huge difference in the resulting cheese, to the extent that entire batches can be rendered inedible. In particular the temperature at which the milk is held can be crucial to the production of individual styles of cheese. A friend of [William Dudley]’s had this problem, as a dairy farmer and artisinal cheesemaker they had to carefully control their vat with a set of profiles depending upon the recipe in use. This was achieved using an Arduino Mega 2650 and a thermocouple to control the heat source for the hot water in the outer wall of the vat.

A cheap K-type thermocouple amplifier proved unsatisfactory, so a Sparkfun item was substituted. A relay, Ethernet adaptor, and LCD display provided power control, access to a web interface, and user feedback respectively. Four buttons to select programs were added, and the whole was neatly boxed up to survive the dairy and put to work. In tests with a saucepan it was configured as a PID controller, but the real vat proved to have a much greater thermal inertia so a simpler bang-bang home thermostat style approach was used. Temperatures are logged in an eeprom for later retrieval via the web interface.

We don’t see the cheeses produced, but we’re sure they must be worth the effort. Blessed may be the cheesemakers, but doubly blessed are they who have a little help from an Arduino.

13 thoughts on “Perfect Cheese Every Time With This Temperature Controller

  1. Interesting project, yet all this technology does make you wonder… How did our ancestors managed to stay alive or even managed to make cheese at all. I guess that in many cases they didn’t… (regarding both).
    Regarding the project, interesting stuff, didn’t expect to learn anything to learn about cheese today.
    So, hackaday, thanks for posting.

    1. I think they are trying to perpetuate a certain taste, which requires the right temperature. People gkt by withiut fancy equioment in the past, but likely the output varied from batch to batch. Being able to measure temperature precisely and constantly likely means tge result is always the same.

      I made tofu for a few years decades ago, giving it up when I decided the effort was too much. It’s a similar process as cheese, start with soymilk, heat it up, add coagulant, and then squeeze hard. Didn’t require constant temperature, but I used a candy thermometer to monitor the soymilk temperature, because it took forever to get to the right temperature, but boiled over in an instant. So watching the temperature gave an indicatiin of how much !onger, and I avoided too much soymilk boiling over tge side if the pot.


    2. I’ve looked into making various cheese myself, and thought the same thing. Some of the recipes have you hold the mixture at a certain temperature for X number of hours, then increase it by 5 degrees and hold for another period of time.

      Suppose if that’s your livelihood it’s a different story, but definitely more work than I was willing to put in for some provolone.

    3. Yes, cheeses were different as different conditions and product were used.

      That is wonderful so cheese eaters will remember a particular batch as good or bad. There was uncertainty that made it more interesting. Less predictable. More natural.

      Today we are far less tolerant and taste & consistency has to be exactly and precisely the same. Or I’ll sue you. Only price changes are tolerated.

  2. nice clean web page. simple and no junk. I like it ;)

    about the thermo- you do know that each time different metals connect, you have a thermo junction. that can be a local effect in addition to the main one you intend, from the end bead, etc. not sure if you were aware of that or not. to be truly measurement quality, you need the right connectors and follow the right rules.

    I’ve also had good luck with lm34 or lm75 (forget which, I think I tried both analog and digital) and it was easier to deal with than t-couples.

    even the ds18b20 is a nice choice for some temp ranges; and each one has its own serial # inside (I think of it like a mac addr). it can be useful for some things (if you have a bunch of onewire temp sensors, knowing the addr of each one and its location lets you create tables of temperature points).

    1. Yes, I’m aware of the dissimilar metals consideration, and my assumption was that the thermocouple interface from Adafruit took that into account, even if that means “ignored the tiny error”. I used a K thermocouple because I could buy one in a 3 foot stainless probe from Omega. I didn’t want to experiment with home-made sensors when food safety is involved, and the Omega probes weren’t terribly expensive. The temps measured by my apparatus agree with the mechanical thermometer they were using for cheese making.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.