I also picked up a lot of 7 used load cells, ranging from 100 to 10,000 pounds capacity.
Above is the layout of the PCB. This really is a very simple circuit and easy to build. Click here for the circuit image at real size for downloading to print and make your own PCB.
(Instructions for soldering up this little amp and using it will be coming soon!)
Here is an assembled amp on the right. To the left was the first amp board I made, the problem was I forgot to mirror the image in my photo editing software, so the first board was backwards (or whatever you call it) and unuseable.
My load cells and pressure transducers arrived. The task at hand now was trying to identify the output pins. I went through the literature I had, to no avail. The only useful piece of information I had was that the both the pair of input pins and the pair of output pins should read 350 ohms. I did have a color code wiring chart that indicated which pins were which. So I opened up one of the load cells, I was pleasantly surprised to find the wires were colored the same colors as my chart indicated they should be. The problem was, the resistance readings indicated pins A and B were a pair, as were C and D. But the color coded wires indicated A and D were a pair, and B and C the other pair. This wasn't going to be easy.
For my first attempt I wired the load cell to my amplifier 10 volt regulated output, and used a voltmeter to read the output from the load cell. As soon as battery power was applied I got a 3.5 volt reading, yikes! That wasn't the correct wiring pattern. So I quickly turned off the power. Obviously these load cells were wired in a proprietary manner. So I'd just have to wing it using A and B for the excite side and C and D as the output side. Pin A was generally used as the positive excite pin, so I'd try it that way, just going positive/negative, positive/negative. Low and behold it worked! So I quickly hooked up the signal to the amplifier and the amplified signal into my DI-194 A/D convertor.
Within minutes I had the load cell calibrated and working to perfection.
Here's the setup with the first load cell working. The short lead wires soldered to the load cell will be potted in silicone, taped together and soldered to a plug for a wire extension leading to the amplifier.
| Pin | |
| A | Excite + Black |
| B | Excite - Red |
| C | Signal + Green |
| D | Signal - White |
Wiring pattern for my 300 and 2,000 pound BLH load cells.
This is the pin letter system of the 9 pin connector.
At the moment I have the adjustable resistor set to 329 ohms, it seems to be giving good resolution and reading up to 170 pounds (The most I could apply by leaning on the load cell on my desk).
Since all these load cells and pressure transducers are from the same aerospace company, and they all have the same connector pins, I hope they are all wired the same way. I checked into buying the plugs for the pin connectors, $30 each. Go figure! I'll solder the wires directly to the pins and buy my own connectors for that kind of money.
The next load cell I wanted to set up was the 2,000 pound cell. My second Defiance static test is coming up and I could use the new load cell on this test. So I wired up this load cell identical to the first one and wired it into the amp and my A/D convertor. Now I had to fine tune the amplifier and calibrate the load cell.
I'm no expert at this, but here's what I do to set up a load cell and the amp. The DATAQ A\D convertor reads an analog voltage from 0 to five volts, the load cell is rated at 3 mili volts per volt of excitation voltage. I'm using 10 volts excitation voltage, so my load cell should put out 0 volts at no load, and 30 mili volts at 2,000 pounds of load. That's way too low a voltage for the A/D convertor to read, so that's why we use the amplifier.
Now we need to figure out how much to amplify the mili volt signal. We want to use as much of the 0 to 5 volt range on the A/D as possible, so I divide the full scale 5 volts by the 30 mili volt maximum signal from the load cell. That give us 166.6, or about 167 times amplification to make 30 mili volts into 5 volts. Makes sense so far, right?
Now that we know how much to amplify the signal, we need to know how to adjust the amplifier. All that is required is changing the value of one resistor to change the amount of signal amplification on the INA125. There is a formula to calculate the gain. "R" is the value of the resistor in ohms.
Gain=4+(60000 / R)
Still pretty easy isn't it? The 60,000 is a fixed value and is based on a resistance value inside the INA125, so don't worry about that. In my case, I wanted to replace my variable resistor with a fixed value resistor to avoid any possible changes in the amplification level. I had two resistors in my parts drawers that came close, a 323 ohm and a 390 ohm resistor. The 323 ohm resistor would amplify my signal 189.75 times, which is a little high. What would happen is I would clip the top end of my signal at the A/D convertor because the voltage would go over 5 volts before I reached 2000 pounds on the load cell. The 390 ohm resistor is slightly under at 157.8, but it will allow full scale reading up to the 2,000 pound max of the load cell.
These load cells will generally function well over their rated limit, in most cases at least 150% of rated capacity with no damage. So I could use slightly less gain, to get even higher possible load levels, but the resolution starts getting cut down so it's best to try to stay as close as possible to the load requirements of the load cell.
Calibrating is an entirely new challenge. In the past, I'd hook up the load cell to the DATAQ and simply zero the load cell with no weight on it, then I'd place a pre-weighed stack of weights on the load cell, enter the weight into the high calibration window and be done. The software extrapolates a linear rate to the signal, so even if you only loaded 25% of the maximum load on the load cell, the calibration would hold all the way to the load cells maximum limit. I'm sure it's best to use as much weight as possible for the best accuracy. With a 2,000 pound load cell that's going to be a trick getting at least several hundred pounds loaded on it. You should be able to calculate the voltage if you know the load cells rating and the amount of amplification. Which I do, but I think it is still more accurate to actually calibrate with a know weight.
