Maker Pro
Maker Pro

Strain gauge conditioner/load cell amplifier

kevnas

Oct 10, 2017
4
Joined
Oct 10, 2017
Messages
4
Hello. I am trying to come up with a miniature strain gauge conditioner/load cell amplifier for use with a full bridge sensor/load cell that has around 2mV/V sensitivity and 1100ohms and having a flexible 5V/10V excitation and a flexible supply anywhere in the range of 3-50Vdc.
There are a few similar solutions to what I want but no schematics/circuit diagrams available anywhere.

http://www.mantracourt.com/products...conditioners/miniature-strain-gauge-amplifier
I just want the 0-10Vdc output..so it does not necessarily need 4-20mA, -+10V or 0-5V output if that makes things a little simpler.
I would appreciate if anyone is able to reverse engineer some kind of a circuit diagram that I can work with just so I have somewhere to start? Recommendations of the types of components would be very much appreciated too.

Please I hope you do not see this as me trying to get someone to do my work, I have really tried to do this myself but I am a beginner & I have struggled for a long while.
some of the solutions I have found are like the ones below which use a chip & therefore not fully analogue but they seem to make things more complicated than the solution above, I am looking for a solution that does not require programming or use of an arduino or any other mcu?:
http://www.mechtechplace.net/mech-t...g-a-low-cost-strain-gage-load-cell-amplifier/
http://www.elane.net/Datasheets/Datasheet_Load_Cell_Amplifier.pdf
 

shrtrnd

Jan 15, 2010
3,876
Joined
Jan 15, 2010
Messages
3,876
What you're looking at is an industrial standard interface module. In the States the most-used standard is 4-20ma, 0-5V. You get one with the other, not either/or. The old original standard was 5-50ma, 0-10V. Looks like you're looking at one that does 4-20ma, 0-10V.
The reason I'm going over this, is that you're probably going to wind-up using that spec for anything else in your circuit, like a display, which will also work on the 4-20ma/05/10V circuit.
The reason for the 4ma 'zero' value is so that when you're troubleshooting a current loop that you're not sure is working; the 4ma would tell you that your circuit is working and you have a 'zero' value. If the DMM read 0ma, you would know that the circuit is actually dead, because you don't see the 4ma 'zero' value.
I understand your quandary in being new at this and I just wanted you to have the basic reason for why you see those specs for the device you're looking at.
That said, there are tons of information from manufacturers called 'application notes' to help people see how to use their products in applications that people want to use their wares for.
I would Google strain gauge/load cell mfgrs, and review their websites for 'application notes' to try to identify what you want. Almost all of those sites also have 'contact us' features, where you can ask questions about applications concerning the products they sell, and which of their products will work in your application.
What you're asking concerns very specific tolerances, which means you need components that work within your specified operating range. Manufacturers are the best source of that kind of specific information. And most are more than happy to direct you to 'application notes' that can probably help you.
If my advice is of no help to you, maybe someone will offer you better information in a following post.
I'm thinking you don't need to reinvent the wheel here, somebody probably already has what you need available, and you just need to find it.
 

kevnas

Oct 10, 2017
4
Joined
Oct 10, 2017
Messages
4
Thanks for the reply & explanation. I did contact some of them but they only directed me to their data sheets which do not show circuit diagrams which brings me back to square 1. I know there are many different solutions available, but for me to know which one to go with that gives the solution which I attached is the problem. Most of the solutions have differential or instrument amps or have a digital solution which the first two are probably the direction I might end up going in if I cannot find any help to reverse engineer that circuit. It is just that I cannot find any solution with two potentiometers, some resistors, caps and transistors as you can see on that circuit diagram. I would have liked it to be that solution because I could understand everything in the data sheet, whereas other solutions that I have been looking at just confuse me even more.
 

hevans1944

Hop - AC8NS
Jun 21, 2012
4,878
Joined
Jun 21, 2012
Messages
4,878
... some of the solutions I have found are like the ones below which use a chip & therefore not fully analogue but they seem to make things more complicated than the solution above, I am looking for a solution that does not require programming or use of an arduino or any other mcu?:
http://www.mechtechplace.net/mech-t...g-a-low-cost-strain-gage-load-cell-amplifier/
http://www.elane.net/Datasheets/Datasheet_Load_Cell_Amplifier.pdf
The "solutions" in the above two links neither use nor require an arduino or any other mcu. Both are analog circuits.

To be clear, are you trying to reverse engineer the mantracourt product and build one for yourself with a 0 to 10 V DC output? Is the form factor (size) an important parameter? How many amplifiers do you need?

Is your load cell always in tension or compression, eliminating the need for a bi-polar output? Or, if not, can you work with an offset output where zero tension or compression is represented by a fixed positive voltage?

You have stated, "I am a beginner & I have struggled for a long while." So what are your qualifications now (after a long while struggling) to design a miniature strain gauge amplifier and construct it? If you are having trouble comprehending the contents at the two links above (mechtechplace.net and elane.net), I doubt you are qualified to design, build, test, and calibrate a miniature strain gauge amplifier, even with a schematic and parts list in front of you.

Have you tried to contact Mantracourt to see if they will build or modify a custom strain gauge amplifier to your specifications? Or is their product line too expensive for your application?
 

kevnas

Oct 10, 2017
4
Joined
Oct 10, 2017
Messages
4
Thanks for the reply.
I am a recently graduated electronics engineer so contacting Mantracourt defeats the whole purpose of me building one by myself & learning from the experience.
"To be clear, are you trying to reverse engineer the mantracourt product and build one for yourself with a 0 to 10 V DC output?" - Yes
Is the form factor (size) an important parameter? - yes-ish but not necessarily for now, maybe after I have one working then I can reduce the size. similar size to the one they have would be good or slightly bigger is okay.
How many amplifiers do you need? - I just need to build one or two to have a spare.
"s your load cell always in tension or compression, eliminating the need for a bi-polar output? " - Wheatstone full bridge type & therefore not always. I would like it to work with one of the inputs positive (e.g 12V and the other set to ground).
"Or, if not, can you work with an offset output where zero tension or compression is represented by a fixed positive voltage?" - Yes
I do understand the solutions on the links, but in terms of calibration they used a NI Daq
This solution seems good, could you please have a look & if it is any better than the INA125P, it seems to have the ability to get a 0-10V output which I am not sure I can get from the two solutions you proposed?
Another thing is the Temperature which is rated at -40degC to 85degC compared to the 0-50degC..The AMP04 Itself is rated to -40-85degC so which components might be reducing this down to 0-50degC and what can I do to keep the rating to -40-85C?
https://www.industrologic.com/sgauman.htm
 

hevans1944

Hop - AC8NS
Jun 21, 2012
4,878
Joined
Jun 21, 2012
Messages
4,878
This solution seems good, could you please have a look & if it is any better than the INA125P, it seems to have the ability to get a 0-10V output which I am not sure I can get from the two solutions you proposed?
I agree. The Analog Devices AMP04 appears to be a better instrumentation amplifier than the venerable Texas Instruments (Burr-Brown) INA125 in many respects. I would try both of them and operate from ±15 VDC power rails.

I haven't proposed any solutions yet, and probably won't, lest the state of Florida accuse me of practicing engineering without a PE license. The two links were provided by you.

Neither approach will produce a 0 to 10 V DC output swing because both use a single-ended 5 V DC supply. However, if you decide to try using the INA125, it's capable of operation with a single-ended supply up to 36 V DC, so obtaining a 0 to 10 V DC output is not a problem. This is may not true for the AMP04, which will operate from either a single-ended 5 V DC supply or ±18 V DC rails, with no mention of what happens if you try to operate the AMP04 with a single-ended 36 V DC (or lower) supply. Probably be okay because what does it know about where "ground" is?

The Elane Electronics product is limited to a single-ended 5 V DC power supply because it uses an AD8541 to provide offset voltage to its AMP04, deriving its power supply voltage from a three-terminal regulator driven by a 9 V DC power source, typically a "transistor radio" battery. No way to bump the supply voltage up because of AD8541 power supply voltage limitations.

Building strain gauge instrumentation amplifiers that actually work reliably in the field is also a character building experience. In the late 1960s, fresh out of a four-year term of enlistment in the U.S. Air Force, I talked my way into a prestigious electronics technician job at a local research institute. My boss was an "old school" ex-General Electric supervisor who had retired to Dayton, but wasn't ready to give up the thrill of new technology yet. He encouraged me to experiment. One day we were looking at spec sheets for a new Burr-Brown instrumentation amplifier, a classic three op-amp implementation assembled and potted into a "brick" with wire leads protruding from the bottom side designed to be mounted on a printed circuit board, in a socket if necessary. IIRC the cost of the little beastie was upwards of a hundred bux. You had to add your own bridge completion, offset nulling, and gain tweaking adjustments to hook this thing up to a load cell. And we had a contract with a nearby Air Force base to service and maintain hundreds of load cells used on materials testing machines. So Clyde (my boss) asked me to try to design a similar instrumentation amplifier, only with better specs and cheaper cost.

I completed that task about a month or so later, learning a lot about op-amps, common mode noise, offset drift as a function of temperature, gain stability, and load cell calibration and hysteresis. I "improved" on the classic design by incorporating a pair of adjustable trimmer capacitors to null-out 60 Hz common mode noise appearing on the shielded twisted-pair load cell wires, and used a pair of matched NPN transistors on a temperature-controlled substrate for the "front-end" input stage to minimize input offset voltage drift.

The completed design worked like gang busters, and only cost ten times what we paid for the Burr-Brown device. Well, at least ten times, maybe a little more... But Clyde never complained about the time or money I spent on this "in house" research project, and he would proudly show it off to any visiting firemen who came around. I think it made him feel young to support "cutting edge" electronics. Anyhoo, a few months later, integrated circuit instrumentation amplifiers became widely available and my "advanced design" was instantly obsolete. Couldn't compete with laser-trimmed parts, adjusted automagically during manufacture to produce the best amplifier specs possible. Didn't bother me. I had a sign on my office door: For More Than One, Please Consult the Factory. It was like that for most of the years I worked there, until finally graduation and a bachelor of engineering degree.

I recommend that you breadboard whatever design you choose using DIP ICs for ease of assembly and probing with test instruments. When everything is working to your satisfaction, change over to SMD and have a board house manufacture and assemble your "prototype" amplifiers. Trying to do it all yourself as a newly minted engineer is a waste of your education and extremely expensive "fun".
 

kevnas

Oct 10, 2017
4
Joined
Oct 10, 2017
Messages
4
Thanks for the reply. "Neither approach will produce a 0 to 10 V DC output swing because both use a single-ended 5 V DC supply. However, if you decide to try using the INA125, it's capable ofoperation with a single-ended supply up to 36 V DC, so obtaining a 0 to 10 V DC output is not a problem. "
I do not think I understand this statement!? So how would I obtain the 0-10V with this?
Another question. If it produces a 0-5V output can I not have a none inverting amplifier added to the output with a gain of 2 to make it into 0-10?
 

hevans1944

Hop - AC8NS
Jun 21, 2012
4,878
Joined
Jun 21, 2012
Messages
4,878
Thanks for the reply. "Neither approach will produce a 0 to 10 V DC output swing because both use a single-ended 5 V DC supply. However, if you decide to try using the INA125, it's capable of operation with a single-ended supply up to 36 V DC, so obtaining a 0 to 10 V DC output is not a problem." I do not think I understand this statement!?
Holy op-amps, Batman! Where did you get your engineering degree? The output of any amplifier is determined by three things: (1) power supply voltage available to drive the output; (2) input signal amplitude; and (3) the gain of the amplifier. Divide into 10 V the minimum signal, in volts, that you expect should result in 10 V DC output. The quotient is the gain to which you should set the amplifier. Now pick a power supply voltage that will provide 10 V DC output with a little head room, say 12 V DC.

... So how would I obtain the 0-10V with this?
See discussion above. You may have to farkle around with offset adjustments so that "zero" bridge output (neither in tension nor in compression) results in zero output voltage.

... Another question. If it produces a 0-5V output can I not have a none inverting amplifier added to the output with a gain of 2 to make it into 0-10?
Sure, you can, but why not set the gain to produce the desired 0 to 10 V output to begin with? And your "gain of 2" approach doesn't have to be a non-inverting amplifier with a gain of two. An inverting op-amp configuration would work just fine with proper power supply polarities and load cell connections.

If you are afraid of "letting the magic smoke out" of your components while bread boarding, try one of the many free SPICE programs available online, and also free to download for non-commercial use. This is the 21st Century: let the computers do all the work while you stay focused on the big picture.
 
Top