• 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.

Signal Conditioning Device for Material Testing system

Status
Not open for further replies.
Hello,

I am working for a company that produces video based measurement systems. One of the key measurements in materials testing is the measurement of strain. Strain is the amount of extension of a material compared to its original length, normally this extension is induced by a force applied by a test machine.

This is normally measured using an extensometer:

Extn.png

There is a port in the Universal Test Machine (UTM) that allows the extensometer to be plugged in. Through this port a reading of extension is given as a signal.

I however would like to use a video measurement system to generate the extension signals, however I need a device to convert from the signals given by out DAC to the signals required given out by an extensometer (so I can plug them into the UTM).

Vid.png

There will be a single extension output from our DAC:

· +/- 10 volts

· + and – signals (differential)

An Extensometer however is more complicated. It normally consists of a Wheatstone bridge, where one or more of the 4 resistors is/are strain gauge(s):


The extensometer port therefore will have four connections:
· +/- excitation

· +/- signal

One crucial element is that the signal is proportional to the excitation voltage, because the excitation voltage can vary from 3 to 12 volts.

Instead of strain being applied to the device, it will be a voltage signal being applied. Therefore it also has to be proportional to the strain signal. Therefore what I will need is a way to multiply two signals together with very high accuracy.

So far I have gathered a few ideas from people:

· Convert everything to a digital signal. I could then use a microcontroller to do the multiplication.

· Use digital resistors in the same bridge layout.

· Use FETs in the region where they act like voltage controlled resistors.

· Use an off-the-shelf multiplier IC.

I haven’t great experience with any of these methods, so I would like to know your thoughts on these ideas and whether I may have missed any other ideas.

Let me know if this doesn't make sense.

Thanks,


Neil
 

spec

Well-Known Member
Most Helpful Member
Hi NM,

Interesting proposal.

The first thing is that what you want to achieve is doable. I say that without analyzing your original post (OP) in detail because of the axiom that anything (within reason) that can be stated concisely in words can be represented in mathematics and and thus can be done in hardware.

Of the approaches that you mention, a processor based systen would be the simplest, cheapest, and most flexible. It would also give you the shortest development time and most chance of success. Of course this approach will involve some computer programming, but nothing terribly complicated.

My feeling is that a Raspberry Pi would be a good choice for the processor, although an Arduino may hack it.

In addition to the processor core, you will need some power supplies and specially designed electronics.

The key to progressing this project will be to produce a definitive design specification giving exact details of the transfer function needed and then produce a functional block diagram.

And that is my ten cents worth.:)

spec
 
Last edited:

OBW0549

Active Member
When you posed this same design problem elsewhere back in late August I proposed a solution (attached) which would have done exactly what you needed (i.e., making the simulated strain gauge output proportional to both the simulated strain and the gauge excitation voltage), but I didn't hear back from you. Did you try it?
 

Attachments

JonSea

Well-Known Member
Most Helpful Member
Do you have enough resolution using video? I just did a quick search and found that for a "505" steel sample, the deflection at failure was about a third of an inch, measured to +/- 5 mils. The strain in the linear region is much much less and this is probably the biggest area of interest. The strain in the Linear region is presented as micro-inches/inch, so the amount of movement is astonishingly small.
 

LulaNord

New Member
Hi...i am a new user here. As per my knowledge a processor based systen would be the simplest, cheapest, and most flexible. It would also give you the shortest development time and most chance of success. Of course this approach will involve some computer programming, but nothing terribly complicated.

pcb assembly
 
Last edited:
Status
Not open for further replies.

Latest threads

EE World Online Articles

Loading
Top