Hi everyone,

First and foremost, thank you for any help/advice you can provide. I am a mechanical engineering student and my knowledge of circuitry is very basic and it has been a good 4-5 years since I last had a course in electrical engineering, so please bear with me.

As the title suggests, I am trying to signal condition a load cell that I have (+/- 3,000 lbs http://www.futek.com/product.aspx?stock=FSH00672&acc2=acc) but after working on it for the past week, I have gotten nowhere. I am afraid I might accidentally damage the load cell (if I haven't already done so). I found a few other posts on this forum about this particular subject but I was not able to solve my problem using that advice. Here is the website I used as a place to start http://www.nakka-rocketry.net/strainlc.html. I've also attached my modification of the circuit presented on that website here: http://imageshack.us/photo/my-images/85/circuitd.jpg/. Basically I modified it to +/- 10V instead of the 5V the site uses. Also, all GNDs are connected.

The way I chose 423 Ohms to be my amplifier's gain is by using the load cells theoretical output at max load (3k lbs) which by the calibration certificate's mV/V output chart comes out to 20.918 mV at 10V excitation. I wanted this to be at 10V, so 10/20.918e-3 = 478. Rg = 200,000/(478-5) = 423 Ohms. Now, I'm pretty sure this is wrong since I didn't read the tare voltage first but still, this isn't where I'm having my issues as far as I can tell. Next, I connected all of the grounds (which is where I think maybe I messed up because the voltage regulator chips are getting burning hot) and at first the voltage read about 10.3V DC but after screwing around a bit I got it to level out at about 0.3###V. Now theoretically (not sure if this is the right calculation) I should get 3,000/2^12 ~ 0.75 lbs resolution since I have a 12 bit ADC I'm reading the voltage with. However, even if i put a load of 40+ lbs on it, the reading doesn't change whatsoever besides the oscillations it goes through naturally with or without load (same #s regardless of applied load).

So based on the above, can anyone tell where I'm going wrong? Also, is my resolution calculation correct? I think my DAQ can read 0-10V but I noticed it does read negative voltages too down to -10V. So would my resolution span the 20V gap or just the 10V? i.e. 6,000/2^12?

Thanks for your time and help guys!

 

Anyone have any thoughts? I really need help on this ASAP I would really appreciate someone's input.

 

A schematic would get you faster comments. It's just too hard to follow a thousand words instead of one picture. Many here won't bother to follow offsite links to photos and such.

 

Thanks for your reply wayneh. Here's the schematic. Sorry not sure why it keeps flipping it : / hopefully you guys can still understand it.

 

How sure are you of your +10 and -10 supply voltages? You're supplying them with ±12V. Can your regulators operate with that low a dropout?

 

I'm sorry, I didn't read your whole post; I just focused in n this one statement:

I am afraid I might accidentally damage the load cell (if I haven't already done so).

My sentiments exactly; this is why I have no input about your circuit; I've never attempted this because I don't want to ruin a very expensive load cell. I always use ready-made strain gauge amplifiers. Here's one that's (relatively) cheap: http://www.omega.com/pptst/DRF-LC.html

 

Yes, dropout voltage is 1.5 V I believe. Though I noticed just now after rechecking that alone, working off the power supply the -10V regulator works fine, outputting the -10V. With the setup I have above, it only outputs +0.6V. Not sure why this is.

 

Here's one that's (relatively) cheap: http://www.omega.com/pptst/DRF-LC.html

I wish that was "cheap" to me, I can't afford spending $150. Also, time is a big issue for me and I really can't wait for things to ship in the mail. On top of that, I'm trying to make this portable so having the circuitry be as small as possible is ideal. Thanks for the thought though.

 

I don't quite understand the purpose of the negative supply. The IC doesn't require it except maybe for very near zero input from the strain gauge.

If your -10v regulator is toast, I'd just ground the IC instead of giving it a negative supply.

 

Not sure why this is.

Double and triple check the pinouts. Then check it again.

 

I don't quite understand the purpose of the negative supply. The IC doesn't require it except maybe for very near zero input from the strain gauge.

If your -10v regulator is toast, I'd just ground the IC instead of giving it a negative supply.

To be honest, I have no clue what I'm doing.. I was wondering why it had a negative supply but went with it because the site I mentioned above said it required a negative voltage to power the op-amp. I'm going to try your suggestion, hoping it works! Thank you.

 

I think that'll be fine as long as you're measuring load in one direction, so that you don't need a negative output. See figure 5 in the Burr Brown datasheet.

 

Tell us the part numbers of your voltage regulators instead of speculating on dropout voltages.

 

+10V regulator P/N: LM2940T-10.0 (dropout 0.5V @ 1A, I'm using a 1.3A supply at 12.1V)
-10V regulator P/N: LM7910CT-ND (took this out based on wayneh's recommendation)

So as of now, the voltage fluctuates a ridiculous amount around +/- .15V which should be about 450 lbs? Once I remove the INA122 and just power the load cell with +10V, the readings fluctuate a lot less staying between two constant numbers whereas with the INA122 the readings jump around a bunch of numbers that constantly keep changing and it never settles around a specific point. Should I be using capacitors to keep the voltage steady? If so where?

Here's my current circuit:

Thanks for all the help.

 

Should I be using capacitors to keep the voltage steady? If so where?

Yes! Follow the datasheet recommendations for the regulator. Also add a 0.1µF capacitor across the power pins of the IC as physically near to it as possible. Other experts here may suggest additional locations.

Also, I think you may need to adjust the reference voltage up off ground. Your unloaded bridge supplies 5v to each input, right?

 

Yes! Follow the datasheet recommendations for the regulator.

I'm laughing at myself right now for not reading through the datasheet. Apparently I need at least a 22 uF capacitor to keep the voltage steady. Better go buy that tomorrow.

Also, I think you may need to adjust the reference voltage up off ground. Your unloaded bridge supplies 5v to each input, right?

^Not sure what you mean by this. Supplies 5V to each input?

Anyway, thank you for all of your help today. I will post my new outcome once I have the required capacitors hooked up.

 

Yes! Follow the datasheet recommendations for the regulator. Also add a 0.1µF capacitor across the power pins of the IC as physically near to it as possible. Other experts here may suggest additional locations.

Also, I think you may need to adjust the reference voltage up off ground. Your unloaded bridge supplies 5v to each input, right?

No, the reference pin needs to be connected to ground, unless you want to set your output to some non-zero voltage when the bridge is balanced.
The DC voltage on the bridge outputs is a common-mode signal to the amplifier.

 

I defer. What I was thinking about was noise around the balanced position. A slight blip to the negative hits the negative rail, whereas a blip to the north causes a positive voltage. Would a reference floor eliminate that, or just shift it?

 

Ok so this is the mockup of the circuit I'm going to build tomorrow. If you guys have a sec I would appreciate any comments.

I'm not sure if I need that extra 1μF capacitor at pin7 (+V) since there's already a 22μF one right there. Also, would pin5 (Ref) need one?

Thanks!

 

Pin 4 needs to be connected to ground. No capacitor is necessary on that pin.