I have an idea to use a scale in a project but I have a question.

If I leave a weight (10-60lbs) on the scale for a long period of time (months) will it loose its accuracy?

Will the metal flex sensor loose its memory and not "zero out?"

Basically I need to monitor weight of a beer keg over time and display how much beer is left in the keg at any given time.

Yes, another beer project:)

Any thoughts?

Don

I guess the only way to know for sure is to do an experiment. I will look for a cheap digital scale tomorrow.

I would like to use the LCD with my project. The Scale tutorial has the excite wires going to PD3 and PD4, but these are used for the LCD. I have looked at the code and at the ATmega168 datasheet and it looks like I could move them to the PB pins. Will this work? If so, can I use 2 scales? One hooked up to PB1 & PB2 and another to PB3 & PB4.

Thoughts?

Hi Don,

Those are good questions, and you are right the best way to figure this out to just try some experiments. My feeling is that at some point your sensor will start to lose its accuracy. Remember how all this sensor really does is measure the change in resistance due to a small expansion in the material, well if it over time it takes less force to bend the metal the same amount your measurements are going to start being off. However, I doubt this will be an observable problem unless you have very exact resolution on your measurement.

You can in fact use PB1 and PB2 as your excite wires, and if you need two scales, you can use PB3 and PB4. However when using two scales you have to be careful about using your ADC to read from multiple inputs.

Good luck! Let us know how your project goes.

Humberto

Let's see if I can help a bit. Over the years, I have used various types of weight sensing devices from capacitive/magnetic levitation to simple strain gauges. The C/M style was great for things like milligrams and lighter but they could not handle more than about 300G at most.

Strain gauges are pervasive throughout the industry these days and quite reliable, provided you do not go beyond their limits. The more robust, the less accurate. So you have to determine the loading beforehand. They do tend to drift a little after operating for longer periods of time but the circuitry in the systems that use them generally compensate for that.

Strain gauges can be used in many configurations to counter these effects as well and also improve their own accuracy. The most common configuration is in a Wheatstone bridge, driving a floating input or instrument amplifier. Mounting methods also need to be taken into consideration too.

If you are using a manufactured setup, it is more likely a load cell, which contains at least one and possibly up to 5 strain gauges. These load cells are very linear in their ranges but mechanical protection is recommended to prevent overloading them.

If your setup does not exceed the limits of the device being used, there should be extremely little or no variation after being loaded for long periods of time. Dave

Ok I bought 2 different scales. One that looked like the scale on the tutorial and another that was totally different.

After opening the scale with a drill (damn triangle security screws. 10 of them!) I found this;

There are 4 sensors. One in each corner with 3 wires coming out each. The extra blue wires on the bottom senors are the "On" switch for the scale. I removed the sensor cover and found this;

There are only 3 wires going to the scale. The sensor is isolated from the metal frame so it is not grounded.

The board looks like this;

The blue wires a grouped left side and right side. The white wires are grouped by opposite corners. The red wires each have there own solder point.

Does anyone have any idea how this is working and how I could make these work with my nerd kit? Ideally I would like to make 2 different scales using 2 sensors in each if this is possible.

I need your help. My beer supply is in jeopardy!

Thanks,

Don

Don, if beers at stake, then I'll help all I can (but I've been in the pub all night)...

To put it bluntly, I don't really know how this works. But my work is in fluoropolymers and due to this I do know that there is a polymer called polyvinylidene difluoride (PVdF), that can be used in electric scales. It has piezoelectric properties, so is better suited than most plastics.

There is a chance that these sensors are using this (or some similar technology), or using a resistance based method.

Whilst I can't offer any useful advise, I'm pretty certain you could use these sensors for your needs (depending on measured weight required).

My first check would be, whilst the scales are in some form of working order, apply a mass to them and measure readings from the sensors (is it voltage, resistance...) once you have established how thay are working you can then design the method for your project.

Keep us informed of your findings.

BTW, how did the brew from the FermentBot (phase 1) taste?

I will find out about the beer this weekend!

i will play around with a multimeter (even though I barely know how to use one:) and see what I can find out.

To address the original question in this thread... "If I leave a weight (10-60lbs) on the scale for a long period of time (months) will it loose its accuracy?"

Short answer: No

Long answer (I'm a mechanical engineering student): Strain gauges(used in most digital scales) work by measuring strain in a material. I will start by describing what strain is. When a pressure is applied to any material the material deflects, this pressure is called stress. The equation for stess is [Stress = force/area or σ=P/A]. The units for stress are pressure units, eg PSI. Strain is the elongation of the material caused by this stress compared to the materials original length. Therefore stress is equal to ΔL/L, or the change in length of the material divided by the original length. Technically strain is a dimensionless unit, but for simplication reasons some people assign it the units of Length/Length, eg in/in. So what is the relationship between stress and strain you ask? The formula is [Strain = Stress * E or ε=σ*E], where E is a material property called the modulus of elasticity or Youngs modulus. Now to answer your question, the above equation for strain only holds true in what is called the elastic region of the material. Take a look at a typical stress-strain curve in this image, stress is on the y-axis in this graph and strain is on the x-axis. As you will notice from the picture there is a linear region at the beginning of the stress-strain curve. The slope of this line is equal to the modulus of elasticity for this material. The linear portion of the stress-strain curve is called the elastic region. It is called the elastic region because all strain in the material is recovered when the load is removed. The non-linear region of the curve is called the plastic region. Unlike elastic strain, plastic strain is not completely recovered when the load is removed. Therefore if you strain a material into its plastic region the piece will stay permanantly deformed. So for your particular situation as long as you are not stressing the loading bar into its plastic region all strain in the material will be 100% recovered and no matter how long you keep the load applied for it will always return to its original shape.

P.S. After reviewing your question I beieve you asked this because you are concerned with the phenomenon "fatigue". Fatigue in metals is a function of stress applied and the number of times a material is loaded rather than the duration time of loading. In the plastic stress region fatigue will cause structural damage or failure in the loading bar after only a few loading cycles. However, in the elastic region you can expect exponentially more cycles, and below a certain threshold you can expect infinite loading cycles. Read up on fatigue at, http://en.wikipedia.org/wiki/Fatigue_(material) if you are concerned. It is to be noted that this information fatigue is only applicable to metals, and polymer materials behave funny and have lasting effects when loaded for long durations.

Best of luck to you. -kcen

BTW, Hi I'm new.

how did you ever make out with this project?

i'd like to hear more.