Some of the required equipment is:
Let's start at the beginning. I have 2 old Compaq laptop computers, and access to another. I hoped these would be adequate for the job, as they were not very valuable and I had an extra as a backup.
The analog to digital convertor could be built, but DataQ sells a small, simple A/D convertor for $25, it includes the software and a cable to run the A/D board into the computers 9 pin serial port via a RS 232 cable. The newest version of the board is the DI-194RS, it has four analog input channels at a maximum throughput rate of 240 Hz. If you use all four channels the throughput of each channel is 60 Hz (60 samples per second). I don't think I could build one for $25, as the cable alone is over $10. So I ordered one, and it arrived in just a couple of days.
The DataQ DI-194RS
The A/D convertor has an input voltage range of +- 10 volts. So I decided I was going to make a simple load cell using a potentiometer and a compressed air piston to drive it. I sketched out a few ideas, and quickly decided to scrap the idea. I checked e-Bay for load cells, it seems there were plenty of surplus load cells and pressure transducers selling for $10 to $15. These items sell for $350 to $850 new, so I was pleasantly surprised.
Above is one of the two, 500 lb. load cells I bought. They are the bending beam style, the side with two holes is bolted stationary. The side with one hole is where the load is applied.
Most load cells require and input voltage of around 10 volts, and the output is about 2 mV per volt of input. The mV output is too low for the A/D board to use. So a signal amplifier would be needed to create a useable signal. Pressure transducers are a little easier to find with an output of 0 to 5 volts. So my first purchase was two 5,000 psi pressure transducers for a total of $25. Next I won an auction for two loads cells rated at 500 pounds. While the load cells only had milli volt output, the 500 pound range was just about perfect for my use. They were $44 for the pair. The last two sensors I picked up were pressure transducers rated at 1,000 and 1,500 psi, they were $33 for the pair.
Above is my quartet of pressure transducers. Maximum pressures run from 1,000 to 5,000 psi.
I hooked up one of the 5,000 psi transducers to a 9 volt battery, then into the DataQ A/D board. I used my nitrous tank as a pressure source. After calibrating the software, I added and released pressure to the transducer and watched the computer display the information, it worked perfectly. After a couple of minutes the calibration started to drift off, the low setting dropping to -300 psi. As I had expected, the voltage would need to be regulated. The transducer simply relates pressure as a voltage relative to the input voltage, as the input voltage from the batteries drops, the corresponding output from the transducer drops.
So I looked around for a voltage regulator, and found the LM 317. It is adjustable by means of a resistor, to voltages of 1.2 to 35 volts. It is a very simple circuit, requiring only two resistors to work. The voltage needed for my transducers varies, two of the better/new ones require 28 volts. So to give the LM 317 enough driving voltage, I needed to make a battery pack of over 30 volts. I simply wired four 9 volt batteries in series to get the required voltage.
Here is a schematic of the LM 317. For R1 I use a 240 Ohm resistor, for R2 I use a 4.7K Ohm resistor. The Regulated output is 28 volts and goes to the load cell positive excitation. The Load cell excite negative goes to the battery negative. For R2 you could use a variable resistor to adjust the regulated voltage for different applications. You could add a couple of capacitors to the circuit, but I haven't found them necessary.
This is a picture of a completed regulated power supply using the LM 317.
Above is a picture of the regulated power supply now mounted in a project box. I added a power switch and an output jack. The output jack is a solder type RCA connector, each pressure transducer that uses this power supply can simply be plugged in, making the power supply useable with several of my pressure transducers.
The load cells were going to need a signal amplifier to bring the millivolt signal up into the volt range that the DataQ could read. I found a very simple IC that would do the trick, the INA 125. The INA 125 would also generate the regulated excitation voltage of 10 volts the loads cells needed. The first one I built didn't work. So the second one I built I used jumper wires instead of copper strips on the board. It works, but for some reason it's intermittent. I don't really need the load cells, my test stands use a hydraulic cell to transfer pressure to a gauge, so I can just replace the analog gauge with a pressure transducer. But, I think the load cells would be more accurate. So I'm pursuing the load cells as well as the transducers.
Here is a photo of a completed signal amplifier using the INA 125 chip.
Here is the INA 125 now mounted in a project box. I did a fair amount of testing, and found I needed an extra ground connection to make the unit stable. After adding the ground connection, it now works perfectly and is very stable. It runs on 2) 9 volt batteries. I added a power switch and three RCA jacks. The label on top reads: INA 125; 10 volt Excite Output; for HBM Model BBS 500; Load Cell. The RCA jacks are labeled: Yellow Amplified Signal Output; Red Excite to Load Cell; Blue Load Cell Output. I'm starting to get so many different electronics I needed to label things so I don't screw up and plug the wrong thing into the wrong place.
Here is the schematic for the INA 125. Keep in mind this schematic is again my own way of looking at it from the bottom side, not the component side. The only component you need is a resistor between pins 8 and 9. A 1 K ohm variable resistor may be used, or you can use a fixed resistor once you determine what gain you need. You can just play with a few resistors under 1 K ohm, or start out with a variable resistor adjusting it to the desired gain, then read the variable resistor with a meter and install a fixed value resistor of the same value. The formula for gain is; Gain=4+(60,000/R1).
Pins 14, 15 and 16 are for different output voltages to drive your load cell. My load cell needed 10 volts so I tapped off pin 16. Pin 15 is 5 volts and pin 14 is 2.5 volts. You will likely only use pin 16 or pin 15. If you use pin 15, you could get by with a single 9 volt battery instead of two wired in series for 18 volts.
I first soldered a 16 pin socket onto a 1.5" perfboard. I then used point to point wiring to make all the connections on the bottom side of the board. After determining what size resistor I needed, I soldered a fixed value 1/8 watt resistor between pins 8 and 9 to set the gain. Two 9 volt batteries in series were used to get 18 volts, then a toggle switch was inserted between the board and the battery. To finish up, I wired all the inputs and outputs to RCA type jacks and mounted everything in a project box.
The idea will be to use channel one as a direct thrust measurement. Channel two will be to measure chamber pressure. For hybrids, channel three will measure oxidizer pressure. I have plans in the future to build a nitrous oxide/alcohol bi propellant engine, in that case I could use the fourth channel to measure fuel pressure.
I decided to rework my solid motor test stand. I was installing the load cell anyway, so I added a mount to install a launch rod on the stand. I also made provisions for a blast deflector on the stand.
Here is a picture of the test stand with the new load cell installed.
Here is a little wider shot of the stand.
In this picture you can see how the launch rod attaches to the stand, the deflector plate will need to be larger, but you get the idea. The launch rod just slides in place, and the deflector plate attaches with just one bolt. I tapped the upper support with a 3/8" nct tap. It's easy to convert from test stand to a launch rod stand in just seconds. I also retained the hydraulic load cell mount on the stand. In case I want to use it again in the future. All in all, the revamped stand should work well, and be versatile as well. I'll post more details on the electrical system once it's been put to use.
Update: Feb. 23, 2004
I used the pressure transducer for the first time a couple of days ago. You can see the results at the bottom of the 2" motor construction and theory page. The results of the first use were encouraging. The only down side I could see, was the plumbing going to the transducer was nearly plugged by residue. That may be a problem with sugar propellants that may not have an easy solution. I suppose relocating the port to the side of the motor may help, but that would mean a dedicated motor just for static testing.
I reworked the test stand again today. I installed a new motor support tube to accommodate the larger 2" motor. The first time I used digital data acquisition, I could not use the load cell because the motor was a little too large for the support tube.
With the test stand reworked, and another test of the 2" motor looming. I decided I needed to check out the load cell and try to calibrate it. The pressure transducers have a nice feature, they have 2 extra pin outs, which when shorted give you either a 75% or 80% of full scale calibration reading. That makes calibrating the transducers easy, the load cell on the other hand, would have to be calibrated with physical weights.
I have two scales, an analog bathroom scale, and a digital shipping scale. Needless to say, I trust the digital shipping scale more than a bathroom scale. But, I decided to use them both to see how close they were. Lucky for me, I have about 200 lbs of lead scuba weights. So I piled them onto each of my scales. They were both within one pound of each other.
I decided to make another 100' data wire before any testing. So I again used RG-6 cable TV wire, I soldered a connector on one end to plug into the amplifier box, and a couple of tinned leads on the other end to go into the screw terminals of the A/D board. I then taped the new wire (load cell) and the old wire (pressure transducer) together. I used electrical tape and made 2 or 3 wraps every 8" or so. There wasn't any sense in having the wires separate, and I was afraid they would become a tangled mess.
With the wires ready. I plugged everything in and started the data acquisition software running. I inserted the motor into the stand, as if it were to be tested. The software was then low calibrated to zero. For high calibration, a stack of weights (previously weighed on the digital scale) were added on top of the motor, and the high calibration set to that point. To test the calibration, I put alternating amounts of weight on top of the motor, and checked the software's reading to the known weight. I was very pleased, the load cell reported within one pound of the scales reported weight up to 266.8 pounds. It would be nice to check it all the way up to 500 lbs, but 266 was as much as I could safely get on the load cell.
The first test of the 2" motor reported a maximum chamber pressure of almost 1,200 psi, that would translate to a thrust of about 550 pounds. While that is 50 lbs over the load cells capacity, the cell is supposed to work up to 750 lbs and be safe up to 1,500 lbs. So I don't think there will be any problems there.
Tomorrow will be the second static test of the 2" motor, and the first time I will be able to record both thrust and chamber pressure. I'll post the results after the test.
Update: Feb. 25, 2004
I used both the pressure transducer and the load cell for the first time. The test can be see here. The motor was my largest to date, holding over 2 pounds of propellant and was a new design. As it turns out, the motor did not perform as expected, and produced a very short high duration thrust. The data acquisition worked well, but for some reason the load cell quit reading, or maxed out at about 350 pounds. The actual thrust at that point was more than likely well over the load cells rated capacity, and that may have been the reason it froze at that point. Once the thrust level dropped, it again functioned properly. Subsequent testing indicated the load cell was still calibrated properly and functioning normally. Update:April 21, 2004: I did a little thinking on the load cell maxing out at 350 pounds, and came up with the reason. The Data acquisition board reads 10 volts, only it's not 0 to 10 volts, it's -5 volts to +5 volts. And I had my load cell amplifier was set with too much gain and at 350 pounds of force it was going over +5 volts. So all I needed to do was adjust the amplifier to a lower gain setting. The load cell now reads full scale.
Update: Feb. 26, 2004
The test stand was used again to test the T-2 motor, the results can be seen here. Only the load cell was used in this test. With the test stand and it's instruments functioning well, I'll bring this web page to a close. It has been fun and interesting setting up the data acquisition equipment. The extra information this equipment is providing has already proven to be very useful.
Update: February 9, 2007
It's certainly been some time since I updated this page. I've been using pretty much the same system for several years now. While I'm still using the same, very first amp I built, I have been making my own PC boards and I have more information on those boards and some new load cells on this page.
Current set up as of February 2007
One thing that always frustrated me was the data cable going from the amp to the notebook PC. I started out using some RG6 cable, but it was to bulky, then I used cat 5 cable for several years, but the cat 5 cable is not very flexible and is solid copper wire, which I fear may break in time. So now I decided to go with 1/4" phone plugs and the extension cables available everywhere. In the picture above you can see two 25' cables, one regular wire and the other is the expandable/stretch style. These cables make hooking things up a breeze, and are shielded audio cables so they should be at least as good as the cat 5 cable in terms of blocking interference.
On the test stand now is one of my new, canister style load cells. It is set up for smaller motors, the motor tube can be unscrewed from the load cell and replaced quickly if need be. The same holds true for the load cell, one bolt and it can be replaced with another load cell. I have a 100 pound load cell on it now, which is rated accurate up to 150 pounds and safe to 200 pounds. I also have 500, 1,000, 2,000 and 10,000 pound load cells. I left the mount in the center intact for testing larger motors, so I can bolt in one of my larger load cells in that location.
The new (to me anyway) notebook computer works fine with the DataQ adc hardware. So that solved the problem of having to lug a full sized PC in the vehicle with me.