Digital weigh balance - DIY

[paulca]

Another one of my project ideas.

"How much is left in the keg?"  It has no level gauges or sight glasses.  When it runs out it's horrible.

Easy... weigh it empty.  Weigh it in situ and tare...

Trouble...  Finding a pressure transducer/load cell which will happily take a 25Kg load 24/7 for a month.

Solution?

Mount the keg on elastic feet.  Buy 4 or 8 feet.  Put empty keg on it, measure distance from base.  Fill keg, measure again.

It should be fairly routine to be able to take both measurements when reseting the keg.  So if the elastic "bumpers" degrade it just gets recalibrated.  If you source different ones, as long as they elastically compatible, it's just a recalibration.

Converting a weight of 2kg - 22kg (for example, a Corny keg) into a distance (even if non-linear) could be done by elastic bumpers, maybe even repurposed suspension "bump dampers".  There are sensors which can measure very small sub milimeter changes in length available cheaply.  Bit of noise filtering and averaging over time should give decent enough for at least a "Fuel guage" type display.

There might be other solutions involving inflated feet.  Many options exist for using a counter-balance, but there is no room for such in a fridge.

[paulca]

Long shots.

You could tell where the level is in the keg by measuring either the thermal transfer at many points up the vessel, or measuring the sonic properties.  You could possibly even send an electrical pulse through the stainless keg and somehow detect the level.

I could also drill a hole in the lid and mount a laser distance sensor.  However, that involves modifying a pressure vessel with (from memory) 60PSI releif valve and a 100PSI test pressure.  Not something I want to attempt.  Safety first.  A keg lid rupturing is 100% going to be messy.  It could also 1-in-a-million be fatal.

[Terry Bites]

Gut a set of cheap bathroom scales (from a reli who's past caring) for the guages and the platform metalwork.

Add an hx711 and an arduino. Tweak up the gain to suit.

Ask Chat GPT to write the code for you. I did!

.

Not beer oclock here yet sadly. Lasers, yeah, yeah. yeah............... Ultrasonsic ping, now you're talkin'.

[soldar]

For under $5 you can buy a hand scale for luggage. I use it for all sorts of things like weigh propane bottles to see how much is left.

[paulca]

I should be more specific...  the keg is a highly-hygienic pressure vessel.  We can't have wires and random sensors unfortunately.

A luggage scale is probably the best way to check them.  Obviously it requires lifting the kegs manually.  I was just thinking of something more electronic than that

EDIT: I could save myself money even more and just write down every drink I draw in a notebook and guestimate....  (sarc?)  What this result sin is that last glass you pull of beautiful golden red beer is instantly destroyed when I massive gush of foam and keg trub turns that last fine glass into garbage for the drain.  Knowing when this is about to occur can actually be solved other ways,  (there are pressure differential shutoff valves which detect the loss of 100 liquid and shut the output off. but still.  Unsatisfactory.

Imagine a small chest freezer, on top is a triple beer tap.  In each is a different beer.  On a screen below it shows you how much remains in each of the 3 primary and 1 standby kegs.

[paulca]

For under $5 you can buy a hand scale for luggage. I use it for all sorts of things like weigh propane bottles to see how much is left.

I am buying one.  For exactly that purpose, testing the propane bottles and CO2 cylinders.

Swapped a CO2 (fire extinguisher) with a local fire company for an "out of certification" one for £10 this morning.  The one I gave him back I assumed was empty.  I read the regulator as just under 600psi.  So nearly empty.  Turns out I read wrong.  I was actually reading just shy of 60 bar.  Which means it still had liquid in it.  I only found out rocking it in my hands.  Not much though barely enough liquid to feel by its momentum.

The cooker gas is out of liquid "slosh" and has been for weeks.  I usually just let it run out and go change it, takes 2 minutes. 

A luggage scale would make all this more precise and easier.

[Kleinstein]

For testing a liquid level some ultrasound method should work OK, though is may need experimenting on how it reads at different levels.

For the idea of measuring the displacement for elastic feeds this should be more metal springs and not some elastomeres.  Plastics and rubber tend to creap over time and can also react to humidity.
Commercial springs should be relatively stable, when used well below the rated load / stress.
For a DIY solution this looks a reasonable way to build a scale.

[CatalinaWOW]

 There are all sorts of cheap scales that you could use for this.  The luggage scales that I am familiar with saturate under the weight but that has an easy solution, just hang the four corners of the keg platform from separate scales and add their reports.

All of these should be more stable than the elastomeric feet you propose, and as you say you can recalibrate if necessary.

The hard part is automating the readout.  Video camera to OCR?   Possible but complex.  Tapping into the LCD drive and deciding?  Not easy.   

Do a search for robotic load cells.  I have found these reasonably priced (~$10 the last time I bought).  Or salvage those out of a bathroom scale.  And inexpensive A to D, a micro and some code and you'll be done.

It would even be easy to check for a tap left open by some careless party goer.

[paulca]

So, the only semi-useful options involve industrial grade load cells which will stand up to 50% load constantly and might last a few years.

Other options which popped up was cryptically "A ball and a magnet".   I figure they mean if you put a ball with maybe a ball bearing in it and you slide a magnet around the side of the stainless keg until you get a "Plunk" of the ball.  Sounds tedious and I can't see how to make it automatic.  Am I missing something?

Anyway.  I have found a near perfect substitute which, with some testing could get even better.

One of these (actually a clone):
https://www.ispindel.de/

It's a small sealed bottle containing a Wifi/BT MCU (8266) and an IMU6050.  18650 cell.  You dump it into liquid and the angle it floats at gives you a calibrate-able scale for density/S.G.

How does that help with the keg?

When I am told that somehow, Wifi will get into the stainless steal keg.  Then the fact that this device will read the beer gravity and temp until it hits the bottom, which will occur at a measurable amount of beer before empty.  (downside is, it will disturb some sediment.

To make it better, when I get it back out of my fermentor I want to test what happens if you attach a light chain to it.  As the chain collects on teh bottom of the tank, there should be less pulling on the device and it's angle should decrease?

[johansen]

Hang the keg from a fish scale

[CatalinaWOW]

There has to be something better than shoving something into the keg.  Remember your comments about taste and sterility?  And while WiFi might make it into that crude Faraday cage,  It would also be a PITA when you change the keg.

Bathroom load cells are actually very stable.  I have been using one for forty years with no noticeable drift. 

If you don't trust load cells, how about making a spring platform. Under it place a bladder filled with fluid that is displaced into a sight tube.  The sight tube can be mounted outside the fridge where it is easily read.  Only problem is that you can't read it remotely unless you put a wifi camera on it.  Or something more complex.

Yes springs can age and sag, but if you don't compress them over their entire range the life is measured in human lifetimes.  (Actually the same thing applies to load cells.  And unless you want to know the last handful of cubic centimeters left in the keg there is no need to load them over 50% of their full range.)

[ealex]

I've seen some "propane tank level meters" used by RV guys.
The work with ultrasounds - strap them on the bottom of the tank and then connect via BT.
Maybe you can borrow one to test it ? I don't know how accurate they are :

Also, don't poke the pressure vessel ... I had 2 PET kegs fail - one popped the valve assembly as I was connecting it, and I was lucky my face was not straight above ...
The're also the risk of adding something that can't be properly sterilized ... with "interesting" results

[paulca]

Also, don't poke the pressure vessel ... I had 2 PET kegs fail - one popped the valve assembly as I was connecting it, and I was lucky my face was not straight above ...
The're also the risk of adding something that can't be properly sterilized ... with "interesting" results

True.  Kegs are dangerous.  I use AEB Corny 19L kegs.  They are stainless steel with rubber top and bottom "bumpers".  They are designed to survive being thrown and rolled around when delivered to pubs and restaurants containing either soft drinks (soda) for draught pumps OR with concentrate for a tap mixer system.  CocaCola, Pepsi, etc.

You do still need to have your wits about you.  Especially when using an unmodified CO2 Fire Extinguisher like I am.  On the later, I (no longer) clamp the FE trigger down.  I just "bump" the highside pressure everytime I pull a beer.  So there isn't enough CO2 available to the system to cause an explosion, unless both the FE valve and the keg PRV fail somehow.

EDIT: Fermentation vessels can be far worse.  Yeast actually accelerate under pressure.  So if you pitch yeast into sugar and seal it.  It almost certainly WILL explode violently.  This occurs far more often than you would think.  Just with bottles and priming sugar.  Either too much priming sugar or not letting the primary fermentation complete and it can produce enough pressure to shatter glass bottles if they capped tight enough.

[paulca]

There has to be something better than shoving something into the keg.  Remember your comments about taste and sterility?  And while WiFi might make it into that crude Faraday cage,  It would also be a PITA when you change the keg.

Bathroom load cells are actually very stable.  I have been using one for forty years with no noticeable drift. 

The secret with the "pills" is the bottle itself is just a food grade bottle similar to a cola bottle.  So the electronics can be safely removed and the bottle dumped in it's entirety into steriliser.

Seems quite effective:


I did originally wonder "How long will a load cell (or 4) last under load?", but I couldn't find any concrete information.  There was suggestion that datasheets should mention it, but none of the 3 or 4 I found did.

I expect, for the most part, it will come down to price.  An industrial capacitive load cell costing $500 might last, an AliExpress $30 one might not.

The full keg can't be more than 30Kg.  I only really need like accuracy of 0.5Kg +/- 0.1Kg at best!   So, (haven't checked specs for this), it should be possible to maybe use a 150Kg load cell in it's bottom 20%... if it can be accurate enough.

I have seen photos on Wiki's showing people mounting 400L pressure fermenters on them in micro-brewerys.  Surely that cost a penny and so they will be expecting those to be under near constant load, whle the vessel is in use and an expensive vessel like that, you want to have it in use 100% of the time!

[S. Petrukhin]

As a student, a long time ago I did a course work on determining the level of liquid in a metal closed vessel.
I used the principle of determining the natural resonant frequency.
The knocker from the bell tapped on the barrel and then the frequency of the sound was measured.
It worked very well, I could determine about the difference of 200ml in a 50l barel.
But to ensure accuracy, you will need to additionally measure the temperature, of course.

For stepwise measurement, you can use the usual old way of determining the gas level in a ballon - tap your finger at different levels - the empty part will have a different sound effect.
This can be automated by installing multiple sensors. 

These methods do not interfere with the requirements of safety and cleanliness, but in any case, installation of equipment will be required.
If you cannot approach the balloon with a screwdriver and soldering iron, then only an X-ray remains.

[CatalinaWOW]

I think the real issue with the pill is battery life.  Checking fermentation state while brewing is only a few days even if you toss it in at the start. 

 Every time you open the pill to change batteries you will have to reclean and sterilize and maintain that cleanliness through insertion.  Also how easy is it to remove?

[AnalogTodd]

Fermentation vessels can be far worse.  Yeast actually accelerate under pressure.  So if you pitch yeast into sugar and seal it.  It almost certainly WILL explode violently.  This occurs far more often than you would think.  Just with bottles and priming sugar.  Either too much priming sugar or not letting the primary fermentation complete and it can produce enough pressure to shatter glass bottles if they capped tight enough.

Yeah, I had a fermentation go a bit wild one time in a glass carboy where I had the airlock working fine when I started the ferment but it bubbled enough to push material up that blocked the holes in the airlock. Fortunately, it was in the garage so at six a.m. the next day when the carboy reached its breaking point nobody was near it.

There were still shards of glass lodged in that ceiling when I sold that house years later.

[IanB]

Trouble...  Finding a pressure transducer/load cell which will happily take a 25Kg load 24/7 for a month.

I don't quite understand why the load cells from a bathroom scale would not fit the bill? People weigh far more than 25 kg, so such load cells should not be overstressed the weight of the keg. (It may be that a bathroom scale has four of them, so they could perhaps be under the four feet of the keg.)

You are already talking about constructing something homebrew with the elastic feet idea, so replacing the elastic feet with four load cells is not a significant difference.

[S. Petrukhin]

I don't quite understand why the load cells from a bathroom scale would not fit the bill?

In real life absolutely elasticity is not present.
If you constantly stand on the scales, you will become heavier. 

[IanB]

In real life absolutely elasticity is not present.
If you constantly stand on the scales, you will become heavier. 

Well, yes, but in real life load cells are used to measure the weight of vessels in industrial applications.

I am assuming that if you keep the load down to a fraction of the maximum, the deformation will be minimized. (Human weighs 100 kg, keg weighs 25 kg.)

[S. Petrukhin]

In real life absolutely elasticity is not present.
If you constantly stand on the scales, you will become heavier. 

Well, yes, but in real life load cells are used to measure the weight of vessels in industrial applications.

When we install sensors for legs at industrial facilities, we sell kettlebells in a set so that the staff periodically calibrates.
Or install reset line to zero on the scales according to the void sensor if it possible.

I do not know the exact data, not involved in this process, but according to colleagues, there is information that the readings of the scales lose about 1% per winter storage of grain - they do not return to zerro after released.

[IanB]

When we install sensors for legs at industrial facilities, we sell kettlebells in a set so that the staff periodically calibrates.
Or install reset line to zero on the scales according to the void sensor if it possible.

I do not know the exact data, not involved in this process, but according to colleagues, there is information that the readings of the scales lose about 1% per winter storage of grain - they do not return to zerro after released.

This makes sense. So paulca would have to periodically reset the zero or recalibrate, the same as would be necessary with the elastic bumper idea.

[S. Petrukhin]

When we install sensors for legs at industrial facilities, we sell kettlebells in a set so that the staff periodically calibrates.
Or install reset line to zero on the scales according to the void sensor if it possible.

I do not know the exact data, not involved in this process, but according to colleagues, there is information that the readings of the scales lose about 1% per winter storage of grain - they do not return to zerro after released.

This makes sense. So paulca would have to periodically reset the zero or recalibrate, the same as would be necessary with the elastic bumper idea.

All depends from required accuracy. I suggested a couple of solutions there that do not require calibration. 

[CatalinaWOW]

This application doesn't require absolute accuracy.  Whatever method you use just reset zero every time the keg is empty, and reset full scale when you put the full keg on.   Will compensate for all long term drifts and only requires stability for the duration of a keg.  You can reduce the difficulty of that by having more friends over.

Note that you care only a little about linearity.  Just like the gas gauge on a car  which is often quite non linear but perfectly usable.

[johansen]

If you use a cheap scale that turns off constantly, a switch to short out the load cell when you turn the scale on and the electronics auto zero, will work to re zero the scale without removing the load from the cell.

It will just have an offset due to the load cell not being perfect. But that doesnt matter anyways because you are going to have to remember how much the keg weighs anyways.