Strain Guage Summing

svn

Member
Join Date
Jul 2004
Posts
74
For hopper weighing, if more than one load cell is mounted then a summing circuit is used and output of this circuit is given to indicator or controller. I am not clear about the working of the summing circuit. How does it exactly work?

SVN
 
svn said:
For hopper weighing, if more than one load cell is mounted then a summing circuit is used and output of this circuit is given to indicator or controller. I am not clear about the working of the summing circuit. How does it exactly work?

SVN

In platform scales that have load cells in all four corners, a summation board is required. The purpose of the summation board is to equalize the amount of signal contributed by each load cell. If you look at the load cell, it will have a mv/V specifaction stamped on it. Typically this will be 3mv/V but in reality each load cell may be, as an example, 2.999mv/V , 3.001mv/V etc. The summation board compensates for the slight difference in load cell ratings.
On a platform scale, the load can be positioned anywhere on the scale, and if the load is placed near a load cell that has a 3.001mv/V rating and then the load is moved closer to a 2.999mv/V rating load cell, two different weight readings will be observed.
Since the load to measured on a platform scale can be placed anywhere on the platform, the summation board is essential to compensate for this.


On a hopper scale, a summation board is usually not necessary as the load is distributed evenly to all 3 or 4 load cells supporting the hopper. Even if the load cells are not exactly matched mv/V , each load cell contributes the same ratio of signal from empty weight to full hopper weight. A summation board on a hopper is usually a convenient method of physically tying all load cells together.

Ian
 
For a hopper weight, I have tied all four load cell outputs together with a wire nut (without a summing board) and run directly to the weight indicator. It worked okay.
 
Replace 'strain gauge' with 'load cell' and Google 'load cell summing', or 'load cell summing box', or 'load cell summing box schematic'.
 
Loadcell Summing

curleyandshemp, thanks for the clarification.

Lancie1, thanks for the reply.

DougP, thanks for the reply. I already tried to google, but I was sure somebody on this forum will be able to give better information. PLCs.net rocks!
 
Lancie1 said:
For a hopper weight, I have tied all four load cell outputs together with a wire nut (without a summing board) and run directly to the weight indicator. It worked okay.

I am amazed! I can't see how that would possibly work!
 
I read this post with interest because I had a project with Load Cells. This application weighs a 55 gallon drum to check level.
I was confused about Summings boards. Searched the Internet for a schematic without success. A schematic for an amplifier that worked on 1 or 2 Load Cells had the cells in parallel, so that's what I did. The results were confusing, so I contacted Omega Engineering for help.
Bernie's answer is correct. 2 cells in parallel average. A summing board does the same thing - average, not add. It has pots to adjust the Excitation output to balance them, but that's all.

A proper setup for a weighing platform would be a cell on each corner, but this was expensive and accuracy wasn't important. The platforms are rollers, slightly tilted back to keep the barrel in place. There wasn't a good way to do it with 1 cell, so we used 2 - one on each of the rear corners. The front corners are pivot points.

A 55 gallon drum of water is 459 pounds. The Load Cells are rated for 1000 pounds. 2 millivolts per volt with 10 volts Excitation. 20 millivolts full scale.

I was hoping I could set this up to have 10 millivolts at 500 pounds, but Load Cells don't work that way. With 500 pounds on the platform, each corner has one fourth the weight - 125 pounds. Quick thinking was a 1000 pound cells gives a 2x safety margin for a 500 pound weight, but results are actually 8x!

Take a platform with four Load Cells, one on each corner. The millivolt output range will be the same if you read 1 cell, or 2, 3 or 4 in parallel. From what I understand, the results will be the same with a Summing box. A platform designed for 1000 pounds should have 250 pound cells on each of the 4 corners. I was only using 2, but the math is the same - 250 on each corner.
In my case, the 1000 pound cells are too large. I should have used 250 or 500 pound units. With 1000 pound cells, a full barrel outputs only 2.3 millivolts. (459 / 4 = 114.75 pounds per corner. 114.75 x 20 millivolts = 2.295). The meter is scaled 0-4000 pounds.
Analog output from the meter is 0-10 volts to the PLC. This is where I lucked out. I can scale the 0-10 volt output in the meter to anything I want. Resolution (millivolts per bit) is the only part that suffers.
I knew weight would be distributed across the 4 corners, but worried what would happen as someone was unloading a full barrel - one corner might see the full weight. As it is, this could never happen, because cells are only at the rear, and unloading is from the front.

During my search for an amplifier, I found meters from Omega are cheaper. CNiS3253 ($240.00) has 10 volt Excitation (60 ma max), 0-10 volt output (for PLC), and a relay that I used for an alarm.
 
Last edited:
It's ok to parallel load cells, a lot of weigh suppliers use this method, nobel, Applied weighing etc, most only use a block of 4 terminals (+ex -ex +out -out).

One thing to note if mounting load cells on vessels it's better to use 3 rather than 4 as this gives more even distribution, 4 needs to be located very accurately or you get what they call stiction i.e. inconsistant weights, think about it a vessel will sit on 3 legs & all will have a load bearing on them, however 4 legs & it's like the preverbial coffee table you tried to level with a saw.
 
keithkyll said:
A platform designed for 1000 pounds should have 250 pound cells on each of the 4 corners. I was only using 2, but the math is the same - 250 on each corner.
QUOTE]

This isn't true, Keith. There are some mechanical details that must be considered when sizing loadcells to a scale. The first and foremost is that the scale has weight. The loadcells must support the scale (this is deadload). Four 250lb loadcells will not be adequate for a 1000lb scale. The other thing is that load cells are not completely linear and have mechanical limits. They are built on a strain gage / wheatstone bridge format. They are far more accurate in the middle of their range than they are at the top, which is to say; they have a sweet spot. The deadload will, on average, put a loadcell up around 25% to 35% of its capacity. The scale capacity will net an output from the loadcell of 75% signal or less. I.E. 1000lb scale, 3.3 millivolt/volt with 10 Volts of excitation will probably not exceed 25 millivolts and will have an offset signal (zero weight applied) of around 10 - 12 millivolts.
 
Last edited:

Similar Topics

I have a 10klbs S shaped four wire load cell that I am going to be using to study dynamic loading on nylon ropes. Basically I am going to drop...
Replies
14
Views
3,922
hi friends , i am working on a new project where the strain guage (4wire) is connected to plc module.the application is to calib the tank which...
Replies
2
Views
1,854
As a friend, Google can sometimes be just a little overwhelming... and as google doesn't use strain guages, and a lot of you guys do, I've given...
Replies
7
Views
4,089
Customer does not want any silo penetrations. Hates level sensors on top of silos. No capacitance sensors either. Would be great if it had...
Replies
6
Views
1,509
Hello, I'm currently working on a project where I need to read values off my 0-3kg Strain Gauge and display them on my HMI display. I've...
Replies
9
Views
1,833
Back
Top Bottom