Kegerator monitor

[bheliker]

I've built a simple "kegerator monitor" ... Basically, the electronics weigh the keg every 30 seconds or so using and publish those values to a little web-app on my phone so I can always know the level of my kegerator. I'm working on some analytic math to get average ounces per hour/day consumed, how much the flow rate spikes at halftime versus in the middle of a big play, or beers per month consumed. ‘Estimated time until empty’ I really need to get working...

Here's the monitor running live: http://www.4drillsbrewery.com/kegmon/

And here's a more in-depth writeup: http://www.brettsbeta.com/blog/2015/01/kegmon/

Any suggestions? Should I add anything?

Cheers,
Brett

 

[helterscelter]

Drift of the weight measurement coming from the load cell when under constant weight is the number one issue I've read about for this use case. I'd be quite interested to hear if you figure out a solution. Decent write up.

 

[day_trippr]

What happens when you add another keg...or five?
Extensibility seems to be an issue, but for a single keg I reckon it's a solution.

I'd change the keg reading to ounces remaining instead of percentage...

Cheers!

 

[helterscelter]

What happens when you add another keg...or five?
Extensibility seems to be an issue, but for a single keg I reckon it's a solution.

I'd change the keg reading to ounces remaining instead of percentage...

Cheers!

I assume he would just add another load cell and stand for each keg to be monitored. I suppose you could use one cell and enlarge the stand to hold more kegs, but you would have trouble knowing which keg was pulled from with out some other sensor to tell you.

 

[raysmithtx]

I've built a simple "kegerator monitor" ...

I have a couple of questions if you don't mind....

Can a single Spark Core handle multiple load sensors? My flowmeters are still giving me lots of phantom pours so I would be interested in this.

According to the spec. sheet the sensors are using a new technology to reduce/eliminate the 'weight drift over time'. How long have you had your system up and have you noticed any significant 'drift' during that time?

 

[bheliker]

Yes, the spark could handle a few load cells. I think it has enough pins for 6 total. An arduino mega could handle more. Processing power isn't necessary.

The actual output is in ounces, I just convert that to percentage for my guests to easily view - they don't often know ounces per keg.

As for drift - there is some. But it is small, I'm seeing a few ounces over a month or so. If I were drinking properly that would be reset every couple weeks with a new keg so it is almost negligible.

Thanks for the feedback!
B

 

[raysmithtx]

I'm gonna give this a try. Servepro is currently out of the load sensors. Quoted price is ~$25 but no estimate on when a new shipment will be available.

 

[bheliker]

For experimentation any load-cell will work. The FX1901 from MeasSpec is just the best one I've ever found.

Maybe I should make a complete product and kickstart it or something. I'm sure if there was an order for more than a few hundred they'd fulfill it!

 

[raysmithtx]

I never heard back from servoflo but I did find the FX1901 at Mouser. More money but in stock.

I have no idea about how large/small the market would be for something like this and I don't know anything about kickstart but if it interests you then go for it.