Question on Measuring Lift Force for MAV Ornithopter for Student Project using Load Cell

[Ecibob]

Hey everyone!

I am new to these forums so bare with me, please. I am part of a senior design student group working on making a micro aerial vehicle ornithopter (generates lift by flapping its wings) similar to this Design.

We are using this load cell connected to an AD620 Op-Amp circuit that basically has two input wires from the load cell coming into the op-amp with a gain of 100 (there is an Rg resistor with 500 ohms in the circuit). The load cell and op-amp circuit are both powered correctly with 10 V and +-15 V, respectively. There are two output wires coming out of the circuit. One comes out of the 6th terminal of the op-amp, and one is connected to ground. If we turn on everything and test the output of the circuit with a multimeter there is little if no fluctuation in the output voltage, which is about .25 V with zero load. If we add precise weights and form a linear graph of the output voltage vs. weight, we can find that there is about .0045 volts/gram relationship.

Here comes the part we we are having trouble. It seems as though the circuit it working correctly, but when we try and hook it up to our USB-6221 DAQ Board though the ai0 (or any analog input) and confirm a LabVIEW program to read the output, we get a ton of noise along with a weird looking graph.

Here is a picture of the front panel that shows the output we are getting http://img130.imageshack.us/i/frontpanelg.jpg/

Basically, I would have though we could get a straight line that continuely increases or decreases depending on the load on the load cell. Why would it graph a ton of vertical lines? I would think it would be just one data point per time value?

Here is the simpler block diagram just to test the output: http://img26.imageshack.us/i/blockdiagramsmall.jpg/

Here is the main block diagram that we were going to use to convert to force, etc. but its not working (no need for while loop?): http://img15.imageshack.us/i/blockdiagrambig.jpg/

The properties of the DAQ Assistant VI is as follows:

input range: -10 to 10 v

RSE terminal configuration

acquisition mode: n samples and not sure how many samples we want to read or how fast. we had 1k hz and 100 samples and the image above is basically what we got it the program runs continuously. 

I don't know what else I can tell you, except that if anyone helps me you would be like a godsend to our group. It would be very much appreciated 🙂

Thanks!!!

[LV_Pro]

As I am unable to look at youor images (the corporate filter here is blocking the url) I can't see what you are describing. I will say that a multimeter will not show fast transitions that will show up in a LabVIEW/DAQ application, so using a meter usually doesn't tell you that much about signal "cleanness". How long are the wires from the op amp/load cell to the DAQ? How long are the wires, and what type (shielded, twisted pairs, untwisted pairs, etc.) from the load cell to the op amp. The output levels of the load cell are usually quite low, so the possibility of also picking up "environmental" electrical noise is very high, and this will be amplified as well. Since the output of the op amp is higher, the noise introduced between it and the input to the DAQ is less of an issue, but there as well. So making sure that signal wires are properly shielded and grounded is important, and there are long papers describing this for low level analog signals. Since the load cells desired output will probably vary relatively slowly (a few Hz) it may also be helpful to put a simple low pass filter before the DAQ.

You should insert the images into your post, as jpg, or png (you have jpg images) Not bmp! (this is for others that might read your post, yours are jpg which should be ok). Generally most of us won't or can't go to links, corporate network filters and corporate policies frown on it, too much maleware lurking at the end of unknown URL's