Continue to Site

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.

  • 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?

Status
Not open for further replies.

fastline

Member
We are trying to conduct a quick structural test in which we will need a load cell. I either need to figure out what basic display I can use with the cell or possibly just drive it with a power supply and read from my precision DMM? Obviously I would just have to convert the mV readings to force but there is some talk about possibly accuracy issues? Need filtering?

All the gadgets I find on ebay and such seem to just be basic PID controllers that seem to to little more than provide a power supply and convert the signal return to a force value?

I am trying to get more info on the load cell we want to use but appears to have 5-6 ranges of measure so I am not sure how they set that up in the device but I assume there is a pin or wire output for each range using either a separate strain gauge in the device or they have separated zones or something.
 
... read from my precision DMM?
Depends how precise your DMM is. If it is a 6.5 digit bench multimeter like the HP (Agilent, keysight, whatever ) 34401A then you should have no trouble getting your results. A cheapie 3.5 digit handheld DMM will be quite a different thing.
 
Looks like my fluke 867 can read down to .01mV with an accuracy of .025%. 5 digit. What is the typical response of load cells? That does not seem like quite enough.
 
OK, I did a little checking and see what others think. I will have to check specs on our load cell but looks like most run 10V excitation voltage with a 1mV/V output so for a 10K cell, that is .001mV/lb. So my meter could only read that to the nearest 100lb. However, we really only need to run a 5k cell which would get me down to around 10lb of accuracy.

However, I am not accounting for any oddities such as stray voltage, which could skew the result.

We could probably get a load cell display but we want to later just connect to a DAQ system for PC display. I am curious if there are some cheap circuits we could use to just use a PC for this for now? It does seem like we should have better resolution to ensure accuracy but really, for our structure test, if we round to the nearest 100lb, we are fine. We are breaking a wooden beam.
 
Do you have a link to the load cell please? There are some cheap DAQ's that should do this and interface with MATLAB.
Any reason you couldnt use a opamp to up the output?
 
I will dig up the numbers that I have on the load cell but it is an Instron and badge indicates several load ranges and the plug has much more than 4 pins so I am assuming they have multiple strain gauges setup for each range?

No reason what soever to not use an amp. Actually, I think that is the way it really needs done. I don't have much DAQ experience but something I need to gain experience in. All I really need to do right now is get reasonably accurate values from the load cell without breaking the bank. If that means reading from a DMM, so be it, but I would not mind spending a few bucks for a more proper setup.
 
Much of what you are asking depends on the load cell. If this will be a permanent affair then there are meters like the Omega DP-41 that will excite and be scaled to read a load cell directly. There are dozens of similar meters out there. If you want to just measure the mV out and convert it you can do that too. Look at the specifications sheet for the load cell. You normally will see a "Sensitivity" specification expressed as mV/V and you will see the full scale rating as well as the typical excitation voltage which is frequently a Maximum. For example a 500 Lb load cell has a full scale rating of 500 Lbs. The sensitivity might typically be 2.0 mV/V or 3.0 mV/V. If the sensitivity is 2.0 mV/V and the excitation is for example 10 volts the output at full scale will be 10 * 2.0 mV or 20 mV. If this is a 100 Lb load cell when 100 Lbs is applied with 10 Volt excitation the output will be 20 mV. So in this example 0 to 100 Lbs = 0 to 20 mV or 0.2 mV per Lb of force applied. Normally a load cell is calibrated as in my example the 2.0 mV / Volt is approximate. Meters like I linked to are designed to be scaled and provide excitation.

This all depends on the load cell, some designs have internal amplifiers. Also, in cases where an amplifier is needed to drive an ADC input or something there are amps available. Really depends on exactly what you want?

Ron
 
Again, I am working on getting the specs off the load cell but assuming a few standard specs, what is a simple option here to just get accurate values from the cell? I think if I amp the signal, my DMM will be more than accurate enough to gather the data. It seems no matter what is configured, the 'system' would need calibrated regardless so we will have to figure that one out.

I was going to look at amping the signal but that again sort of runs us down a path of unknown and expense.

Now, I realize the right way here is probably to use a scaled system but as long as I know the exc voltage being supplied and hold it constant, I would think I could get reasonable results?

Any thoughts on a basic cheap amp? Will I need to worry about filtering for this? I would estimate no more than 3ft distances and we could use shielded wire if that is needed.
 
Again, I am working on getting the specs off the load cell but assuming a few standard specs, what is a simple option here to just get accurate values from the cell? I think if I amp the signal, my DMM will be more than accurate enough to gather the data. It seems no matter what is configured, the 'system' would need calibrated regardless so we will have to figure that one out.

I was going to look at amping the signal but that again sort of runs us down a path of unknown and expense.

Now, I realize the right way here is probably to use a scaled system but as long as I know the exc voltage being supplied and hold it constant, I would think I could get reasonable results?

Any thoughts on a basic cheap amp? Will I need to worry about filtering for this? I would estimate no more than 3ft distances and we could use shielded wire if that is needed.

I omitted something above. About the excitation voltage, it needs to be stable, very, very stable. That becomes apparent when we look at the load cell sensitivity. Once you have some specifications and the cell range much more can be covered. With a good metering system there is really no need to amplify the signal off the load cell, doing so just adds another variable. Also, at a glance, Instron makes a wide range of cells for a wide range of applications. Your exact cell and data sheet are pretty much a must.

Ron
 
The output of a load cell is not the voltage on a single wire with respect to ground, but a differential voltage between two wires, neither of which is at ground. You can measure that just fine with a DMM of sufficient sensitivity, but you need to be careful when selecting a DAQ. Many can do differential measurements, but some do so by reading two lines and then calculating the difference. That's OK for some applications, but not all.

Best is to use an amplifier. But not just a simple op-amp, but an instrumentation amplifier. They can give you gain, as well as properly handle the differential signal from the load cell.
 
The output of a load cell is not the voltage on a single wire with respect to ground, but a differential voltage between two wires, neither of which is at ground. You can measure that just fine with a DMM of sufficient sensitivity, but you need to be careful when selecting a DAC. Many can do differential measurements, but some do so by reading two lines and then calculating the difference. That's OK for some applications, but not all.

Best is to use an amplifier. But not just a simple op-amp, but an instrumentation amplifier. They can give you gain, as well as properly handle the differential signal from the load cell.

What he said. :)

If you amplify use an IA (Instrumentation Amplifier) and if you run into a ADC you want one with a differential input. Most will spec "Single Ended" or "Differential". You want the latter of the two.

Ron
 
A couple questions. What would be a reasonable resolution for a 10k cell?

Also, the cell we want to use and still trying to get specs but is stamped with several ranges.. 500,1000,2500,5000,10,000. Would u think this uses a different strain gauge for each internally? Plug has a lot more than 4 pins.
 
A couple questions. What would be a reasonable resolution for a 10k cell?

Also, the cell we want to use and still trying to get specs but is stamped with several ranges.. 500,1000,2500,5000,10,000. Would u think this uses a different strain gauge for each internally? Plug has a lot more than 4 pins.

From my brief read on the brand you are looking at ran with a sensitivity of 2.0 mV/V so assume 10 V excitation the FS (Full Scale) output with 10 Volt excitation will be 20 mV. So you get .020 V / 10,000 = 2uV / Lb force. That assume pounds of force. You also have not mentioned if you are measuring compression or tension? I have seen a few dual output load cells but they were strictly precision cells used for special applications. Normally load cells have a single range. I have seen cells listing several ranges on a label but one of them is marked, hilighted or circled to identify what that specific cell is. All of this should be in the data sheet for the model you have. You need a data sheet to know what you have. Resolution means little without knowing the uncertainty / accuracy of the cell. A cell with an uncertainty of +/- 1% will be +/- 100 Lbs on a 10,000 Lb load cell and that is without a repeatability specification.

Ron
 
oops, I think I messed up the math on my units, comparing V and mV. for a 10K cell, 10V exc, and 2mV/V sens, that is .000002V/lb or .0002V/100lb units. ?

That is well beyond my meter's ability so I will need to amp that. What are my options for that? I think regardless of what cell we use at this point, they will all respond similar and we already know we will need to push about 10K of force.

I can read mV with good accuracy but uV, not so much.
 
This is where, given a choice, I like using a meter designed for the task. A meter like I mentioned earlier which will also afford rock stable excitation voltages. They can also be scaled to read out in engineering units and spanned for the range of the load cell. Also they allow for easy calibration of the finished system. That 2mV/V is not always the case, the sensitivity could be 1.8 mV/V or 2.2 mV/V it all depends on the strain gauge bridge.
If you go the instrumentation amplifier route you want a really good amplifier like the AD524 Precision Instrumentation Amplifier. While not cheap it offers easy to configure fixed gain and a pot is easily used to calibrate the gain to the load cell. Whatever you choose note the noise ratings.

Ron
 
Ron, to be honest, I don't think our budget will allow that Omega meter right now but I may be able to find something used or discounted. However, I am having trouble determining which if any units actually have internal amps, if they are needed, and how to know if a particular meter is right our our application?

I realize, that regardless of how we set it up, we will then need to tune and calibrate the system as the sensitivity is not guaranteed.
 
Ron, to be honest, I don't think our budget will allow that Omega meter right now but I may be able to find something used or discounted. However, I am having trouble determining which if any units actually have internal amps, if they are needed, and how to know if a particular meter is right our our application?

I realize, that regardless of how we set it up, we will then need to tune and calibrate the system as the sensitivity is not guaranteed.

No problem on the budget, that is why I suggested the IA (Instrumentation Amplifier) I suggested. The unit you have, does it have a nameplate? There should be a nameplate with the basic data that is needed. Until you have a pinout of the connector there is not much else that can be said.

Ron
 
LOL, that has become a point of frustration because the cell is an older Instron unit that lists out several ranges yet Instron indicates it only has 1 strain gage! It has 10 pins so obviously more is going on than just a 4 wire bridge. I was hoping if it had multiple internal gages, we could just connect the range we need.

If it helps, this is a tension/compression unit.

Instron is still looking for specs on the unit.
 
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.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top