Analog to Digital Board for Aquisition of Rocket Thrust Data

An analog to digital converter supplies a convenient way to transfer thrust information from a rocket engine test stand to a computer. This information must first be produced at the rocket engine test stand by a sensor that produces a signal that can be translated by the analog to digital converter (from here on referred to as an A/D board) into a digital signal. The digital signal can then be grabbed by a computer and conveniently analyzed, graphed, plotted etc. The rest of this page describes in detail how to reproduce the prototype described herein. The purpose of this page to provide information & software for average hobbyist so they may make their own rocket engine test stand.

Rocket engine test stand exerts force on a load cell. The load cell design used in this project came from Nakka's Experimental Rocketry Web Site. There are other load cell designs discussed there, but for the first implementation I recommend sticking with design that has been tested. Thy hydraulic load cell produces a pressure as force is exerted on the load cell. A pressure transducer produces an increasing voltage as the pressure increases. This voltage is converted to a digital signal by the A/D board and transferred to a computer via a RS-232 serial connection. A custom software application then reads this data and exports to a comma separated text file that can be imported into Excel as a spreadsheet. The data can then be graphed, plotted, plugged into an equation, etc.

The design for the test stand used in this project is loosely based on the static test stand built by Dan Thames. Much of the steel was salvaged from a scrap metal heap at my father's ranch in NV. I did end up paying $13 for the C-channel that runs the length of the stand (upper and lower beams). My father welded it together for me (My welding is atrocious) so the total cost was very low, probably less than $20 including the paint and mounting hardware. A drill press was used to drill a hole for the hinge point of the upper rail and all of the mounting holes for the engine holder and the hydraulic load cell. If you end up paying a professional welder to construct an engine test stand of this style it will end up costing you over $150. The stand was designed to take a large difference in engine sizes. This would be accomplished by moving the engine holder (round pipe piece) on the top up and down the length of the top rail. The load cell can also be moved up and down the length of the bottom rail. The bolt holes were carefully measured off to accommodate 5 setting positions on both the top and bottom rails for the engine holder and load cell. This adaptability will allow for a large range of engines to be fired in the stand. Smaller engines will have the engine holder at the opposite end of the rail from the fulcrum (hinge point), with the load cell near the fulcrum. A configuration for firing larger engines would require the engine holder to be positioned near the fulcrum and the load cell to be positioned as far from the fulcrum as possible. I estimate this engine test stand to be adequate for engine sizes from about G through L. I think I may make heftier side spars when I fire my first L engine though, just to be safe.

The Load cell configuration very closely matches Nakka's 'Hydraulic Load Cell for Thrust Measurement'. The one important thing to know about a hydrolic load cell before it is installed is the surface area. For the load cell described on Nakka's site this would be the surface area of a single side. Instead of attaching a needle gage however, a pressure transducer is attached instead. The pressure transducer used for this project was a Measurement Specialties Inc. MSP410-2P2-ND which Digikey retails for $100. For those who find this price inhibitive, an alternate route would be to use the strain gauge based load cell described on Nakka's web site. It is important to make the mounting hardware for the load cell before drilling the holes in the bottom rail, otherwise the mounting holes in the bottom rail may not match up with the boltholes on the top rail.

The A/D board is a simple design. It is laid out on a small prototyping board, and comprised of a voltage regulator, RS-232 driver, A/D converter, and a few capacitors. My soldering is even worse than my welding; so I decided to lay the parts out on the board, ensure I had a working configuration, and then epoxy everything in place. If you use the epoxy method, use epoxy that has a long working time. The A/D board was inexpensive relative to the price of the pressure transducer. All together I think it only cost me about $35. The board is run by a 9-volt battery, which will last it many, many hours. The two connectors on the board are a RS-232 serial connector that goes to the computer, and a basic 4-pin connector. The later connector can be replaced by anything that is convenient, I picked the voltage regulator I use up at Radio Shack, which is where I got a couple of the capacitors and the voltage regulator as well. The wire that connects the pressure transducer to the A/D board needs to be shielded twisted pair, if it is not you'll end up with to much noise to have a useful signal. I used a connection length of about only 15ft. If you desire a longer connection length 50ft should be no trouble, although you may notice a slight voltage drop in the signal strength.

Click to see a full size schematic

UPDATED!

Click here for a parts list including prices

The RS-232 driver chip used is a MAX232EPE. The A/D converter is a Texas Instruments TLC1549. It is a 10 bit A/D converter. There are 4 1.0 uF capacitors that drive the MAX232. There are also 3 capacitors that help cut noise in the circuitry of the board. (470uF, 220uF, and 100 uF). All of the capacitors used are electrolytic, so pay special attention to polarity when connecting them to the board. You can view the full size schematic by clicking on the thumbnail to the left.

The software used is a custom implementation designed specifically to work with this board, and was developed as a Win32 application. It supports a sample rate up to 1000Hz, and a variable length record time for data acquisition. After recording the thrust data simply export the data (as comma separated values) to a text file and import it into Excel. (Sorry, I don't think Open Office supports importing comma-separated values into its spreadsheet yet.) Right now the software will need to be run from a laptop, I intend to do a port to Windows CE as soon as I get a pocket PC late this year. Below are a couple of graphs from Excel using data exported from the software.

Graphing the data in Excel should yield results typical of the test run below: (simulated since I didn't have any engines to fire off)

The frequency response of the system is pretty good, this sample was taken by dropping the top rail of the engine test stand down onto the load cell. This test well demonstrates the scan resolution of the system. Now when your engines CATO you'll be able to see the resulting force spike in the data curve ;) WARNING: don't drop contact point of the thrust stand onto the load cell if you don't leave about a 1cm^3 air bubble system, otherwise you risk over pressuring and bursting the pressure transducer from the extreme shock that will result without an air cushion - I found out the expensive way!.

If you have any questions, comments, suggestions, requests, etc, please feel free to contact me: ben@mccart.us