Coffee Forum
frame
› ›
After the old forum software breaking in a way that we were unable to fix, we've migrated the site to a new platform.

Some elements aren't working as we'd hoped - some avatars didn't survive the transition, and we're still having issues with attachments that weren't added as inline images, but we're hoping to have that all sorted out soon.

Temperature Data Loggers

LwowiakLwowiak
edited January 1970 in Roasting - How To do it
Hi all,
I was hoping for some advice and recommendations for a quality Temperature Data Logger. I am roasting on a KKTO, and have decided that measuring the temperature may make life a little easier.
I understand many have gone down this path and was hoping that those with experience can lend some advice.
It seems there have been quite a few innovations lately. I can see that on CS they have a heatsnobs TDL, that looks promising. I would have to drill a hole through both pots of the roaster to install it (at the base of the inner pot), and remember to remove it, every time I go to tip out the beans.
Any advice on choice of logger, software,  installation, words of wisdom or otherwise, are welcome.
I thank you in advance.
Lw.
Share on Facebook

Comments

  • mycuppamycuppa
    LW, The best way to consider solutions for logging a roast is to segment each of the 3 core components needed in order to achieve your objectives. 1. Sensor - thermocouple 2. Bridge - device transducer or I/O conversion appliance 3. Graph software What I suggest you do is research the various options available in each of these components so that you arrive at the solution that best suits your needs - cost, complexity, reliability, quality of end-result. I've been logging roasts for such a long time on numerous platforms - well before it became a trend for coffee roasters. A background in electronics, process control, industrial and IT engineering helped drive the desire from the beginning. Each topic requires research as there are many choices and decisions to be made. Something as simple a probes - type, grounded, shielded, thickness, length (for both probe and cable), response time, etc. all influence the sort of results, accuracy and stability of the solution. As a general rule, stick with thinner, K-type probes that are more responsive, but placement is the most critical design element. Be careful to ensure the positioning will record the appropriate readings from the target source and that it's reliable and consistent so that observed and developed reference points have some form of reliable integrity. Also be careful that probes don't get damaged by agitation or operation. Bridges are a keenly debated topic. You can start with something simple like an advanced multi-meter or take a DIY path with an Arduino. There are a myriad of possibilities, but what you need to look for is making sure the bridge works reliably, does not "hang" or "freeze" at the worst times and that noise and interference (EMI) are managed. The purpose of the bridge is to convert the resistance changes detected by the thermocouple into signals and output via USB ports that can be fed into graphing apps. Some bridges come with multi-ports that enable you to capture the environment temps or even ambient temps as well as bean temps, so this can significantly aid the roasting process as you can do things such as monitor the air temp around the bean mass and detect when it starts to drop as a leading indicator the bean temps (or Rate of Change) will stall or fall. The key to a good bridge design is for the USB connection with the computer to be robust and reliable - a critical weak link in apps many years ago. Graphing software is the business-end of the equation. Here is where it translates into something that gives you all the important insights. What you need to look for in the application is an ability to "smooth" curves and plot lines without hiding real or valid changes ( dumbing it down ). Although nothing is worse than using graphing software that plots jumpy graphs looking like those old-fashioned seismographs with erratic lines everywhere. You need to be able to scale the graph so that the temperature ranges for your probes are drawn easily, there needs to be are features like large displays and averages for Rate of Rise, etc. The best apps overlay the graphs for bean and air temps with the RoR so you don't need to flick back and forth. Artisan is an excellent roast scope application and it's device support is wide. There are also similar apps like RoastLogger, etc.  and more are coming onto the market - most are free. My pro tip....... Don't get too hung up on what temps your solution is recording by comparing it to others. Just because one person might achieve turning point at say 90 deg C and yours it at 100 C, does not mean something is wrong. Each installation is a little different and the probe implementation makes all the difference. Just ensure that you see a good range of temps between turning point and finish temp so that you know your probe is recording as close as possible the actual or likely real bean temps. Obviously, the design of your vessel and the way in which you roast make a big difference, but let's say for example you should see maybe a low of 100 C and typical dump temps of 209 - 215 C for coffees roasted just before 2nd crack. On one of my many roasting plants I use a Cropster bridge feeding Artisan and it works a treat. I prefer to use that solution over the vendors automation package because some roaster manufacturers have no idea. My manual workaround is simple and hard work but I can beat the vendors algorithm every time.....who said humans are becoming obsolete.  On other systems it’s more complex like control of burner, drum, air, destoner, cooling, etc. in configurable steps. In our most advanced platform which is the best available (and seriously expensive), it can roast according to desired curve which is insanely geeky and cool, nothing else on the market can match this functionality and it’s like the next generation technology that very few people are using......which is surprising. In summary, it’s my firm opinion that you cannot roast high quality with consistency by relying upon olfactory senses. I qualify this by stating it gives you a great foundation understanding but its significance is overrated - by the time you have sensed or seen something it’s too late. It’s fundamentally critical to see everything at all times and this is only possible with data logging - the single most important contributing factor why coffee quality has spiked so high in the last decade, roasters can see what’s going on instead of relying upon sight and smell.
    Share on Facebook
  • SimonSimon
    All great advice from Jeff, especially not comparing your temps with others! As long as it's consistent for you, and you have a reference point that you can work with with your particular setup, that's the most important thing. I'm only recently getting into measuring temps, and bought a Heatsnob recently, and it really is awesome, I highly recommend it. Very responsive. And the software is completely free to download, and plenty of info on there with how to use it and any particular mods you wanna do. I like the fact that you can record comments at any particular points on the graph, comes in handy. But super simple to use! I'm soon going to rig up my Behmor so that the probe goes directly into the drum to more accurately get the temps that the beans are going through, I think the bead-type probes are more responsive than the normal type. I'm not sure how any other temp-measuring devices work, I've only had experience (very little) with the Heatsnob, and I'm a fan :)
    Share on Facebook
  • LwowiakLwowiak
    on 1518674808:
    LW, The best way to consider solutions for logging a roast is to segment each of the 3 core components needed in order to achieve your objectives. 1. Sensor - thermocouple 2. Bridge - device transducer or I/O conversion appliance 3. Graph software What I suggest you do is research the various options available in each of these components so that you arrive at the solution that best suits your needs - cost, complexity, reliability, quality of end-result. I've been logging roasts for such a long time on numerous platforms - well before it became a trend for coffee roasters. A background in electronics, process control, industrial and IT engineering helped drive the desire from the beginning. Each topic requires research as there are many choices and decisions to be made. Something as simple a probes - type, grounded, shielded, thickness, length (for both probe and cable), response time, etc. all influence the sort of results, accuracy and stability of the solution. As a general rule, stick with thinner, K-type probes that are more responsive, but placement is the most critical design element. Be careful to ensure the positioning will record the appropriate readings from the target source and that it's reliable and consistent so that observed and developed reference points have some form of reliable integrity. Also be careful that probes don't get damaged by agitation or operation. Bridges are a keenly debated topic. You can start with something simple like an advanced multi-meter or take a DIY path with an Arduino. There are a myriad of possibilities, but what you need to look for is making sure the bridge works reliably, does not "hang" or "freeze" at the worst times and that noise and interference (EMI) are managed. The purpose of the bridge is to convert the resistance changes detected by the thermocouple into signals and output via USB ports that can be fed into graphing apps. Some bridges come with multi-ports that enable you to capture the environment temps or even ambient temps as well as bean temps, so this can significantly aid the roasting process as you can do things such as monitor the air temp around the bean mass and detect when it starts to drop as a leading indicator the bean temps (or Rate of Change) will stall or fall. The key to a good bridge design is for the USB connection with the computer to be robust and reliable - a critical weak link in apps many years ago. Graphing software is the business-end of the equation. Here is where it translates into something that gives you all the important insights. What you need to look for in the application is an ability to "smooth" curves and plot lines without hiding real or valid changes ( dumbing it down ). Although nothing is worse than using graphing software that plots jumpy graphs looking like those old-fashioned seismographs with erratic lines everywhere. You need to be able to scale the graph so that the temperature ranges for your probes are drawn easily, there needs to be are features like large displays and averages for Rate of Rise, etc. The best apps overlay the graphs for bean and air temps with the RoR so you don't need to flick back and forth. Artisan is an excellent roast scope application and it's device support is wide. There are also similar apps like RoastLogger, etc.  and more are coming onto the market - most are free. My pro tip....... Don't get too hung up on what temps your solution is recording by comparing it to others. Just because one person might achieve turning point at say 90 deg C and yours it at 100 C, does not mean something is wrong. Each installation is a little different and the probe implementation makes all the difference. Just ensure that you see a good range of temps between turning point and finish temp so that you know your probe is recording as close as possible the actual or likely real bean temps. Obviously, the design of your vessel and the way in which you roast make a big difference, but let's say for example you should see maybe a low of 100 C and typical dump temps of 209 - 215 C for coffees roasted just before 2nd crack. On one of my many roasting plants I use a Cropster bridge feeding Artisan and it works a treat. I prefer to use that solution over the vendors automation package because some roaster manufacturers have no idea. My manual workaround is simple and hard work but I can beat the vendors algorithm every time.....who said humans are becoming obsolete.  On other systems it’s more complex like control of burner, drum, air, destoner, cooling, etc. in configurable steps. In our most advanced platform which is the best available (and seriously expensive), it can roast according to desired curve which is insanely geeky and cool, nothing else on the market can match this functionality and it’s like the next generation technology that very few people are using......which is surprising. In summary, it’s my firm opinion that you cannot roast high quality with consistency by relying upon olfactory senses. I qualify this by stating it gives you a great foundation understanding but its significance is overrated - by the time you have sensed or seen something it’s too late. It’s fundamentally critical to see everything at all times and this is only possible with data logging - the single most important contributing factor why coffee quality has spiked so high in the last decade, roasters can see what’s going on instead of relying upon sight and smell.
    Jeff, Thank you so much for your detailed response. It is greatly appreciated and a wealth of information. I have to agree with your final paragraph. I have been roasting quite well, using my senses, and this has been a great foundation to understand and learn the basics. To repeat roast profiles and maintain consistency, more information and data is required, hence my current pursuit. While my roasts are 800g to 1kg, and I do not have the challenges of a commercial roaster, it is still important to see what is happening and plan in advance. My attention to detail (sight, sound, smell, time) has resulted in some excellent roasts that I wish to repeat, but cannot do consistently without a TDL. I roast for family and friends, and have no intention of a roasting business as it is purely a hobby and stress relief.
    Share on Facebook
  • LwowiakLwowiak
    on 1518690503:
    All great advice from Jeff, especially not comparing your temps with others! As long as it's consistent for you, and you have a reference point that you can work with with your particular setup, that's the most important thing. I'm only recently getting into measuring temps, and bought a Heatsnob recently, and it really is awesome, I highly recommend it. Very responsive. And the software is completely free to download, and plenty of info on there with how to use it and any particular mods you wanna do. I like the fact that you can record comments at any particular points on the graph, comes in handy. But super simple to use! I'm soon going to rig up my Behmor so that the probe goes directly into the drum to more accurately get the temps that the beans are going through, I think the bead-type probes are more responsive than the normal type. I'm not sure how any other temp-measuring devices work, I've only had experience (very little) with the Heatsnob, and I'm a fan :)
    Hi Simon, Thanks for the reply, and feedback on the Heatsnob. My challenge is to install a probe through both pots of the KKTO, and not interfere with the agitator. The probe also has to be easily removed at the end of the roast. This poses a problem as I believe that unless it can be securely "anchored" in the exact position each time, the readings will vary. I am exploring the installation aspect with an open mind, and am seeking a practical and repeatable solution. Lw.
    Share on Facebook
  • mycuppamycuppa
    LW, There are compression fittings used to secure thermocouples into position. You could maybe use a short piece as a sleeve marker once you have your probe position ideal. Then it’s a case of just sliding it into the compression fitting until the sleeve hits the edge and maybe retrofitting a wingnut to add a bit of tension so the probe won’t move once it comes in contact with beans. Quick release and insertion.
    Share on Facebook
  • LwowiakLwowiak
    Thanks Jeff, Your advice is greatly appreciated. I did quite a bit of research and have found a number of suppliers for the thermocouples and compression fittings. I ended up buying the "heatsnob" as a starting point, and will assess the unit and results going forward. Any mods/improvements will be noted. Due to a mistake (mine), I got the 100mm probe (default); and installed it by attaching it with some copper wire. (will post photos). As a start, I am glad that I did not attempt to drill through the pots and use compression fittings, though may still consider this in the future. Better to start simply and then devise a solution, as one assesses the setup. First few roasts have been easy to record, and I will post some results and feedback in the next few days.
    Share on Facebook
  • SimonSimon
    on 1520473865:
    Thanks Jeff, Your advice is greatly appreciated. I did quite a bit of research and have found a number of suppliers for the thermocouples and compression fittings. I ended up buying the "heatsnob" as a starting point, and will assess the unit and results going forward. Any mods/improvements will be noted. Due to a mistake (mine), I got the 100mm probe (default); and installed it by attaching it with some copper wire. (will post photos). As a start, I am glad that I did not attempt to drill through the pots and use compression fittings, though may still consider this in the future. Better to start simply and then devise a solution, as one assesses the setup. First few roasts have been easy to record, and I will post some results and feedback in the next few days.
    Ah nice! Yeah that's a good way to go about it, start with a simple setup and assess from there. If it does everything you need, no need to change. But yeah can alter it over time if need be. I'm still looking for a place that actually sells copper wire that I need to secure my mod I'm gonna set up (you'd think it would be commonplace!!) haha. Yeah let us know how you go :)
    Share on Facebook
  • LwowiakLwowiak
    Simon, For the copper wire, hop into Bunnings. You can buy it by the metre. You will need to strip the insulation off, which is easy to do with a sharp knife, or pliers, but not compatible with OH&S. Take care not to cut yourself. They have numerous conductor sizes that you can buy, both single and multicore. I used 1.5mm2 TPS as I happened to have some old cable lying around. Lw.
    Share on Facebook
  • SimonSimon
    on 1520476829:
    Simon, For the copper wire, hop into Bunnings. You can buy it by the metre. You will need to strip the insulation off, which is easy to do with a sharp knife, or pliers, but not compatible with OH&S. Take care not to cut yourself. They have numerous conductor sizes that you can buy, both single and multicore. I used 1.5mm2 TPS as I happened to have some old cable lying around. Lw.
    Yeah I went to Bunnings and they were fresh out, will try again soon. Ah really didn't realise had to strip anything off, oh do you mean if there's any of that rubber tubing stuff around it (sorry, I'm not the most knowledgeable handyman :D). I'm assuming so that it's just the copper itself that'll be subject to the high heats
    Share on Facebook
  • mycuppamycuppa
    Just a basic word of caution when it comes to thermocouples. They work on the principle of resistance and tolerances are very narrow/small. What that means is that when you try to add connections or cables, there is an added resistance every time you join or extend cables. Also, it's important to keep cable types the same, e.g. conductor size, length, type, etc. otherwise impedance is changed and the system will produce skewed results. Keep in mind the following golden rules. 1. Shortest length. 2. Fewest joins/connections - if possible, use soldered joints or high quality crimps. 3. Same cables. Thermocouple cable is a special type and will not be available from Bunnings (it looks similar to figure 8 speaker wire or alarm wire but it's generally yellow sheathed with either a green/white pair or a yellow/red pair). You can try to substitute, but bear in mind that runs risk of sub-optimal results. For domestic use it's probably OK, for commercial use......not.
    Share on Facebook
  • LwowiakLwowiak
    Excellent advice Jeff, thank you. To clarify what I stated. I was referring to Simon's statement of requiring copper wire to "secure" his modification. I "secured" my thermocouple to the inside of my roaster with bare copper wire. I had some some flat TPS (white 2 or 3 core cable used to wire houses) lying around. I stripped off the insulation to expose the bare copper wire and twisted it into shape to secure the thermocouple in place. Quick and nasty, but did the job. Not sure what Simon has in mind?  Secret mod business perhaps. As per mycuppa's advice, I would not modify the thermocouple cable. Once damaged, it is better to start afresh. Most of the thermocouples are made to order and to specific tolerances. You order is actually built depending on your specific requirements (length, type, temp etc). Bunnings do not sell Thermocouple wire, only simple building wire in rolls or by the metre (manufactured by Olex). Building wire is made from stranded copper conductor, which is malleable and easy to use to secure a probe in a home roaster. If damaged, it is easy to replace. The stranded copper wire will withstand the heat in the roast chamber as it has a melting point of 1000C. I will post a picture of my setup in due course.
    Share on Facebook
  • LwowiakLwowiak
    My very basic installation of the thermocouple into the roasting chamber:
    Share on Facebook
  • SimonSimon
    on 1520510026:
    Excellent advice Jeff, thank you. To clarify what I stated. I was referring to Simon's statement of requiring copper wire to "secure" his modification. I "secured" my thermocouple to the inside of my roaster with bare copper wire. I had some some flat TPS (white 2 or 3 core cable used to wire houses) lying around. I stripped off the insulation to expose the bare copper wire and twisted it into shape to secure the thermocouple in place. Quick and nasty, but did the job. Not sure what Simon has in mind?  Secret mod business perhaps. As per mycuppa's advice, I would not modify the thermocouple cable. Once damaged, it is better to start afresh. Most of the thermocouples are made to order and to specific tolerances. You order is actually built depending on your specific requirements (length, type, temp etc). Bunnings do not sell Thermocouple wire, only simple building wire in rolls or by the metre (manufactured by Olex). Building wire is made from stranded copper conductor, which is malleable and easy to use to secure a probe in a home roaster. If damaged, it is easy to replace. The stranded copper wire will withstand the heat in the roast chamber as it has a melting point of 1000C. I will post a picture of my setup in due course.
    AH yeah I just meant normal copper wire to twist to secure the probe wire to another attachment, tricky to explain, and involves windshield wiper metal insert XD. Will post when I have it up and ready, but yeah your mod looks great!
    Share on Facebook
  • Koffee_KosmoKoffee_Kosmo
    on 1520510756:
    My very basic installation of the thermocouple into the roasting chamber:
    Via email and a phone conversation I recommended to @Lwoiak This is the easiest way for a thermocouple probe to be added / fitted on a KKTO roaster No drilling needed KK
    Share on Facebook
  • samkonstassamkonstas

    Temperature data loggers are essential for monitoring conditions where accuracy and consistency matter, but the hardware running the data analysis software also deserves attention. A Bottleneck Calculator CPU tool can help identify whether the processor may limit performance when handling large datasets, generating reports, or monitoring multiple loggers at once. Checking system balance can support smoother data processing and more reliable temperature monitoring workflows.

    Share on Facebook
Sign In or Register to comment.

Coffee Forum

@ 2026 The Coffee Forum, All rights reserved.

Policies

Privacy Policy

Social