• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Reading output from load cell with DMM?

Reloadron

Well-Known Member
Most Helpful Member
I have seen plenty of cells that only use 4 or 6 pins of a large multi pin connector. Some use sense lines so the excitation voltage pins can be 4 pins. I would try as kubeek suggest and try doing some measurements with an Ohmmeter. If you find two pairs for example that appear to be "shorted" they are likely excitation.

Ron
 

fastline

Member
https://measurementsensors.honeywell.com/ProductDocuments/Load/Model_53_Datasheet.pdf

I guess we have one of these available to us in the 20K variety. Our expected full load in our device under test is 5K lbs. We expect failure around 10K lbs. I believe 50-100lb increments to be reasonable for our testing purposes but I am trying to determine by the spec sheet how I can arrive at a reasonable resolution or increment? When I look at the nonlinear accuracy, that is full scale, which is about 100lbs. I realize we cannot weigh a bag of apples accurately with this but wondering if this can work for our needs?

I am looking at using a DB25 Omega meter for this load cell.
 

Reloadron

Well-Known Member
Most Helpful Member
OK, the load cell posted is a button type and unless I miss my guess is a compression only load cell. Called a "button" because of the button where the force is applied. As long as you can fixture it for your application they work just fine. Looking at the DP25B (you mentioned DB25 but I believe you have the DP25 in mind), the DP25B replaces the old DP25 meter, just a B suffix. They are good meters and do include the excitation output. The downside is the display is 4 digit . The decimal point is programmed and what you can do is program the meter. The maximum display will be 9999. which could be 9,999 Lbs on a 10,000 load cell. This is the manual for the meter.

OK, your load cell sensitivity is 2 mV/V so with 10 Volt excitation at full scale load you get 20 mV out of the load cell. Note the low end on the meter's programmable range is +/- 50 mV and I wouuld use the bi polar rather than 100 mV range. Even though your cell is compression. Now if you choose the 20K load cell it gets a little tricky in that you set your display so with 20 mV applied (20,000 Lbs) your display reading is 2000 (four digits) and remember to multiply *10 of the reading. So a reading of 2000 * 10 becomes 20,000. The meter manual I linked to explains how to setup one of these. They basically use a "Read" / "Display" system. The meter is similar to the DP 41 I mentioned but only a 4 digit display. It should work fine along with the load cell you mentioned.

Now you mention a failure at.... A nice feature with meters like this is the analog out card which is an option. The analog out can be scaled for current or voltage out (your choice) this can be useful if you want to run the results into a data logger or similar device.

Ron
 

fastline

Member
Ron, sounds like we are getting close. I may try to use the Instron unit I mentioned above if we can find the data we need from it, but otherwise we will have to go with the puck unit I mentioned above.

I think what I am mostly trying to determine is the "resolution" for the meter if we use a 20k cell? All I can find is a 15bit ADC but in some other reading it seems like there should be a uV spec as the maximum step that the meter can resolve? The cell indicates "infinite" but I would still think there is some level of resolution there too? I need to know how well a 20K could resolve or what is reasonable? 10lb honestly would be super great for us. 100lb could work but less ideal.

I think I can live with the 4 digit display for now as we won't push over 5 digits right now but I am hoping we can soon step into DAQ cards where we can leave panel meters behind. I understand what you are saying about scaling the display to knock off one digit for 5 digit readout and resolve it to 100lb increments. Probably no biggy over 10K lbs.

Also, if we use the Instron cell, it is tension/compression so do is that a bipolar bridge?
 

Reloadron

Well-Known Member
Most Helpful Member
Here is what I would think about doing if the budget allows it. You mention a 15 bit ADC but you have not mentioned the ADC range? OK, so 15 bits is going to give you 32,768 quantization levels. 2^15 = 32,768 so if for example the ADC input is 0 to 10 volts then 0 to 10 volts = 0 to 32,768. So what we have is 10 Volts / 32768 = .0003052 or about 305 uV resolution. Now depending on your software for your A to D (ADC) you scale it. This is where using an Analog Out card with the meter you mentioned can be worked into the scheme of things. This depends on the ADC. I can show you an example using a Dataq DI 710 depending on what you want to do with the data. The more you explain the more I can tell you. Being retired I have all the time in the world anymore and it is still to cold in NE Ohio for my outdoor shooting range trips. :)

Ron
 

fastline

Member
Ron, if my numbers are right, per your math, that is only about 67 available steps assuming the total sense 20mV? Do I have that right. that resolution seems very low? Is the 15bit ADC typical of strain meters?
 

Reloadron

Well-Known Member
Most Helpful Member
Ron, if my numbers are right, per your math, that is only about 67 available steps assuming the total sense 20mV? Do I have that right. that resolution seems very low? Is the 15bit ADC typical of strain meters?
I would not do it that way. I would use the meter you mentioned and get the analog out and scale it 0 to 10 volts.

The sensitivity is 2 mV/V and the excitation is 10 Volts. The full scale out of the load cell will be 20 mV, You will have a noisy signal down in the dirt, Use the meter you mentioned (4 digits) and scale it for 2000 and multiply the reading * 10 for 20,000 for your 20,000 load cell. Then use an analog out card and scale it 0 to 10 volts. That 0 to 10 volts goes to the ADC (data logger?). 15 bit makes for 2 ^ 15 or 32768 quantization levels.

We are bouncing around too much. You need to decide exactly what you plan to use and run with it.

Ron
 

fastline

Member
Ron, I might not be understanding you on the output side. I am not really looking to go into DAQ right now, just need to obtain force values with reasonable resolution. I figured the ADC would be calculated off of the sense input back to the meter, not the exc voltage? With that, I could see resolution to maybe 300lbs with a 20k cell?
 

Reloadron

Well-Known Member
Most Helpful Member
Just for a note. The ADC has a reference voltage and a number of digital bits. So an ADC input can be for example zero to ten or zero to five volts. Now as we mentioned earlier if the ADC has a 10 volt maximum input and is a 15 bit device we get 10 Volts / 32,768 and that gives us the resolution (2^15 32768). So the ADC can resolve 305 uV.

If you just want to obtain force values I would buy the little DP-25 meter as that is what it is designed to do. Setting up a load cell can be as simple or complex as you wish. You can manually write down the data which is simple or you can use a data acquisition system and record the data which is more complex. It's all just a matter of budget and what you want to do.

Ron
 

Alegator

New Member
Could have for example a thermocouple so that you can calibrate it wrt temperature, and since you say it is tension/compression, then it could be two load cells and a thermocouple. It should not be that hard to measure what is connected to what.
This would be possible with a precision DMM, also I do not think you will need a filter, you only need a filter if you are processing the output signal of the load cell, in which case the signal processor (ADC) will have an inbuilt filter. However, you can save yourself all the stress of having to always run the voltage-to-force calculation for each measurement you make and also get more accurate results since structural testing actually requires you need reliable results –lives might be involved - by purchasing a good display and amplifier hardware. This hardware will offer great functions like auto-zero tracking, auto-calibration, switching between different metric units. Find some of this incredible hardware here:
 
Last edited:

Latest threads

EE World Online Articles

Loading

 
Top