I could be interested in doing some beta testing! Any rough idea on the cost?
Arduino Keg Sensor for Level, Temperature, and Pressure with Display
- Thread starter UncleD
- Start date
Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:
I would be interested in Beta testing this as well...
Nathan Graen
Well-Known Member
- Joined
- Apr 10, 2020
- Messages
- 59
- Reaction score
- 14
This is really cool. How is the sensor sealed from the beer, it looks like there is a window in the box that the camera shoots through?
- Joined
- Feb 28, 2020
- Messages
- 79
- Reaction score
- 50
Why not poll every hour (or something) with the ability to poll on demand? So for most of the time and users an hour is prob fine, but if you've got all those friends drinking poll manually as often as you wish...just a suggestion.
rsquared
Well-Known Member
In the setup I was describing (no wires, fully contained along with a battery inside the keg), the device sleeps for that hour to save battery then wakes up to read the sensor and push to your server/logging interface. Because it's locked away inside the keg, there's no way to wake it from sleeping, but you could have a setup where you set "party" mode on the server, and the next time the device wakes and sends data the server gives it an updated config with a 5 minute sleep time.
OP
OP
UncleD
Well-Known Member
- Joined
- Dec 1, 2018
- Messages
- 190
- Reaction score
- 190
@Nathan Graen the sensor is sealed around the surface and the surface itself can be exposed to liquids. It does have to be dried off or shaken after cleaning before re-installing or any liquid on the surface will affect the readings... However, because it's facing downward, the problem usually works itself out.
@fphredd The controller isn't "polled" from the cloud server, it "pushes" the data to the cloud. Data usage is a whole different conversation, there are limits to how much data can be transferred. The more data, more often, the higher the hosting cost. Adding an "on-demand" option would be cool, but I worry about potential abuse...
@rsquared this would be really cool to have everything self contained, and was actually one of my first thoughts. The downside, is that these sensors have warmup times before you can trust the data, and the size of battery that could be put in there would be really small. WiFi based controllers consume more power than BLE controllers (like the Tilt for example). I haven't done any calcs on this, but I would imagine needing to regularly charge it up (in a matter of days, not weeks or months). So much that it would just make sense to have a waterproof USB port on the lid, and have it plugged in 24/7. But at that point, you're already running a cable to the lid, why not have that be a cable that's connected to a powered WiFi controller with the luxury of a display. But that's just my opinion.
Just gauging interest here, how many of you folks like the touch screen display? Or Would you rather have no display and have it configured exclusively from a smartphone or PC? I was kinda digging the touch screen display so I don't have to pull up my phone every time I walk by, but maybe i'm old fashioned? haha!
Bigdaddyale
Well-Known Member
Just gauging interest here, how many of you folks like the touch screen display? Or Would you rather have no display and have it configured exclusively from a smartphone or PC? I was kinda digging the touch screen display so I don't have to pull up my phone every time I walk by, but maybe i'm old fashioned? haha!
I'm in the "Both" camp. cheap little 4X20 LCD and phone access
I'm in the "Both" camp. cheap little 4X20 LCD and phone access
rsquared
Well-Known Member
I'm working on some software for an iSpindel (which uses a WiFi ESP8266 based board) right now, and those take an 18650 battery, which is not a whole lot bigger than a AA battery (65mm vs 50 mm tall, and 17 vs 13 mm diameter). They're nominally a 3.7 volt battery, but at full charge get up to around 4.2. The point of all that is, I've had this iSpindel sitting on my bench pushing data every 15 seconds for the past 48+ hours, not bothering to turn it off between coding sessions, and it's dropped from reporting 4.08V to 4.00V in that time. People regularly get several weeks out of an iSpindel with a 15 minute cycle in a fermenter. My thought on 1 hour cycles was to push that uptime from weeks hopefully into the months range.
- Joined
- Nov 30, 2019
- Messages
- 10
- Reaction score
- 14
Me !!! As a software developer I'd love to integrate this into my projects
The phone (mobile) option sounds good to me as I already have an iPad on my kegerator so show what's available on tap, among other things. Next step was to find the "best" way to get volumes...
Dexlor
Active Member
I would just do the display less and have it report in another dashboard. Hook up a cheap monitor or something. I'm sure with Thorrak's interest, it could be integrated into Fermentrack.
Well, one never knows how a developer will react
But from arm's length a serving keg volume indicator doesn't fit into Fermentrack's existing mission.
A closer relationship would be one of the RaspberryPints paradigms...
Cheers!
But from arm's length a serving keg volume indicator doesn't fit into Fermentrack's existing mission.
A closer relationship would be one of the RaspberryPints paradigms...
Cheers!
I want nixie tubes.
Nathan Graen
Well-Known Member
- Joined
- Apr 10, 2020
- Messages
- 59
- Reaction score
- 14
Definitely both.
Well, one never knows how a developer will react
But from arm's length a serving keg volume indicator doesn't fit into Fermentrack's existing mission.
A closer relationship would be one of the RaspberryPints paradigms...
Cheers!
Probably not ideal with Fermentrack, but @LBussy and I have another project in the works this would be perfect with, so long as @UncleD is interested in allowing his work to be integrated with other projects!
OP
OP
UncleD
Well-Known Member
- Joined
- Dec 1, 2018
- Messages
- 190
- Reaction score
- 190
Sounds like there is a mixed bag in terms of interest in the display or not. The good news is that the sensor can function independent of the display for those that want to integrate into their own custom system, or a community supported system. Basically you don't NEED the controller/display for the sensor to function. I'll look into setting up a small batch of sensors and displays to get in the hands of some of the interested parties for some testing and feedback. Stay tuned homebrew community...
I would definitely be interested in what you guys are working on. In the interest of keeping this thread on topic, shoot me a PM or link to another thread with what you guys are working on and I'm happy to chat!
I would definitely be interested in what you guys are working on. In the interest of keeping this thread on topic, shoot me a PM or link to another thread with what you guys are working on and I'm happy to chat!
I can see having a display being mixed - some want to walk by and see it, it's a fun hobby and fun to geek out on. Sometimes super useful info as well. Other times, some don't want it in their living area / bar perhaps, don't want to find a place for it, and / or just won't actually check it very often.
I'd be good with a phone app, I'd just check in once in a while when I start thinking - "Hmm, I've poured a lot of beer out of this keg, wonder if it's almost gone yet?"
I'd be good with a phone app, I'd just check in once in a while when I start thinking - "Hmm, I've poured a lot of beer out of this keg, wonder if it's almost gone yet?"
Bigdaddyale
Well-Known Member
Have you created a Github page yet?
I would be most interested in a version without a display. That's only because I have a five keg coffin keezer and it's a fair amount of work to open it up, so a display would not be of benefit for me.
Right now I weigh my kegs when they go in, and I have a book on the keezer where I log pours. Whenever a new keg goes in, I pull all of them out to re-weigh them while I clean the lines. This works pretty well on a day-to-day basis, not as well when we have a party and the guests track their pours. I've always had my eye out for a better solution!
Right now I weigh my kegs when they go in, and I have a book on the keezer where I log pours. Whenever a new keg goes in, I pull all of them out to re-weigh them while I clean the lines. This works pretty well on a day-to-day basis, not as well when we have a party and the guests track their pours. I've always had my eye out for a better solution!
Which raises the question: how well does this paradigm scale?
I would suggest, for the wired I2C version, providing a method to daisy chain the bus, while ensuring there's only one set of pull-ups active. So each module would need a pair of 4 pin headers for that purpose, and perhaps a termination plug on the end-most node.
Otherwise you're looking at a radial topology which may be unfriendly, SI wise; or burning pairs of host pins for each module...
Cheers!
I would suggest, for the wired I2C version, providing a method to daisy chain the bus, while ensuring there's only one set of pull-ups active. So each module would need a pair of 4 pin headers for that purpose, and perhaps a termination plug on the end-most node.
Otherwise you're looking at a radial topology which may be unfriendly, SI wise; or burning pairs of host pins for each module...
Cheers!
Harleybrew32
Well-Known Member
- Joined
- Nov 29, 2019
- Messages
- 573
- Reaction score
- 1,008
this is interesting, my brotherinlaw built me 4 keg scales.
they are battery operated been working great for over a year. he is working on a new set that works off of wifi instead of RF. this could be an easier solution.
didnt notice any drift on themand they seemed to always nail the weight of the keg when empty. and full. but im not a programmer, he handles all that.
interested in what parts you used.
cheers
they are battery operated been working great for over a year. he is working on a new set that works off of wifi instead of RF. this could be an easier solution.
didnt notice any drift on themand they seemed to always nail the weight of the keg when empty. and full. but im not a programmer, he handles all that.
interested in what parts you used.
cheers
Thorrak said:
Would be interesting to hear more about these scales, but in another thread so that this one doesn't get sidetracked.
Most electronic scales are based on strain gauge load cells, which tend to drift with time, so need to be re-zeroed often. An electronic scale that didn't suffer from drift sounds like a game changer.
Brew on
Bigdaddyale
Well-Known Member
what is used for the pressure sensor and temp sensors?
OP
OP
UncleD
Well-Known Member
- Joined
- Dec 1, 2018
- Messages
- 190
- Reaction score
- 190
Not yet, I'm planning to have some folks test this product out, work out any issues, then I'll be working on posting up any libraries for integration into other controllers. Stay tuned...
There are a couple different sensors I've been using to dial in accuracy vs complexity vs cost... Short answer, it's not official yet, haha
It scales well. Currently this is set up as a 1:1 sensor to controller. The code is already written to support up to 4 (maybe different color enclosures for different I2C addresses). I've also considered multiplexing if needed, so I2C addresses can be shared. First things first, I want to get this product dialed in, if it works well and people like it, the next step is definitely having a single controller/display support multiple sensors.
- Joined
- Feb 28, 2020
- Messages
- 79
- Reaction score
- 50
I'm not gonna be on the real savvy side for wiring or writing code, though my dev teams always tell me I'm a great edge case...maybe that makes me a good choice for beta tester 
That's them saying "you're a pain in the @$$," but nicely.
- Joined
- Feb 28, 2020
- Messages
- 79
- Reaction score
- 50
That's them saying "you're a pain in the @$$," but nicely.
Actually, I make sure they call me those names to my face...and they do!
Stephonovich
Well-Known Member
I would love to buy these and provide beta testing. On the software side, I have experience in Python, Javascript, and some C/C++. I'm pretty good with AWS infra and SQL/NoSQL administration should that be needed.
Personally, I'd like to see it pushing to a locally-hosted DB, with a web dashboard. Other than maybe libraries for whatever cloud DB you're using, it shouldn't be hard to set it up for a local endpoint.
Personally, I'd like to see it pushing to a locally-hosted DB, with a web dashboard. Other than maybe libraries for whatever cloud DB you're using, it shouldn't be hard to set it up for a local endpoint.
OP
OP
UncleD
Well-Known Member
- Joined
- Dec 1, 2018
- Messages
- 190
- Reaction score
- 190
Yeah, there are alot of ways to go with the cloud DB. Right now I am developing an iOS and Android app to allow alarm configuration, push notifications, historical trending, etc...
I am building a set of these and I'll be reaching out to those interested to get a group of testers together. I'll be looking for feedback from testers with programming/integration feedback as well. While this is totally a standalone product right now, I'd love for it to be compatible with other homebrew community platforms and open source controllers.
Stephonovich
Well-Known Member
Excellent. I've built a whopping single app, with React Native, hope your experience is better than mine.
Right now my only brew tech consists of a Tilt getting pushed through an RPi, but I'd be happy to try my hand at other stuff. Also, as I think you alluded to in a previous post, if you open-source the code that'll be a huge help in getting integrations.
rsquared
Well-Known Member
I was just hoping to see a list of part numbers and maybe a wiring diagram, but if you're getting a whole kit together, that'd be awesome as well! Really looking forward to playing with this!
raymosaurus
New Member
- Joined
- Aug 24, 2020
- Messages
- 1
- Reaction score
- 0
Hi mate. I love your solution. Well done!
I have several questions and plenty of suggestions, but first things first;
How accurate and reliable has the ToF sensor been so far? I see your volume graph, but has it been used on various colour beers yet?
Does it accurately and reliably measure even water?
I often force carb my kegs by shaking the hell out of them at high pressure (~35psi). Do you think the sensor would drip clear every time?
Thank you!
I have several questions and plenty of suggestions, but first things first;
How accurate and reliable has the ToF sensor been so far? I see your volume graph, but has it been used on various colour beers yet?
Does it accurately and reliably measure even water?
I often force carb my kegs by shaking the hell out of them at high pressure (~35psi). Do you think the sensor would drip clear every time?
Thank you!
OP
OP
UncleD
Well-Known Member
- Joined
- Dec 1, 2018
- Messages
- 190
- Reaction score
- 190
I am putting options together along with supporting documentation, stay tuned!
The distance sensor appears to be accurate to ~0.1 GAL, occasionally i'll see fluctuations of 0.2 GAL, but it is pretty stable. I have used it on water and a couple different ales without issue. I haven't tried a dark beer like a porter or stout yet, but I'd be surprised if the performance is worse. But hopefully I'll be brewing one of those soon and I can test.
I force carb my kegs as well. I have noticed that the level doesn't read correctly after getting the lid wet (shaking the keg), but later that day is reading correctly (I am assuming the surface drips clear and the foam settles). That being said, only time and testing will tell if this happens EVERY time. I could imagine a scenario where the foam may coat the sensor and dry, but this hasn't happened yet.
I am putting a batch of these together for a group of testers, hopefully we'll get some feedback on items like this.
Cheers!
Yes. Looks great! I'm in for the beta! Ultimately, I'm hoping to attach four of these sensor lids to a Raspberry Pi, so I can view keg status via a simple web page. I'd be happy to buy printed lid enclosures from you, UncleD, and beta test. I also can code for the Pi. Terrific project!
- Joined
- Feb 28, 2020
- Messages
- 79
- Reaction score
- 50
I do hope I'm on that list...I've been good this year!
SalParadise
Active Member
I'd be interested in this for sure - I'm just getting started researching options for monitoring keg levels. Adding temperature, pressure and notifications would incredible!
I'd be happy to beta test as well. I've got a 3D printer, and am comfortable with soldering electronics, so I'd be happy to assemble a kit.
