Chris Biedrzycki

Chris Biedrzycki

Tony Swanson

ME 224 – Final Project

6-06-03

Linear Positioning Device

Linear positioning devices are becoming widely used in today’s society. Some of these devices incorporate the use of non-contact sensors. This type of sensing is seen in industry today with varying degrees of complexity. Applications for a system like this include filling/monitoring liquid containers, taking position readings, and autonomous applications where positioning is a functional requirement. These sensors are very useful in manufacturing and production applications. Ultrasonic sensing formed the basis for our project. This type of sensing can provide accurate results, so we built a simplified linear positioning system that utilized an ultrasonic sensor and tested its functionality. We were curious as to the accuracy of results we could obtain with a crude lab setup.

Several features of ultrasonic sensors give them great flexibility. One important aspect with these sensors is that the sensor does not need to come in contact with the measured surface. Using traditional methods, it would be difficult to measure things such as liquids, very hot items (hot rolled steel), or items which vibrate heavily. Also, since it has a large range and can take readings very quickly over the whole sensing range, these types of sensors are good for acquiring targets far out of range of standard capacitive or inductive sensors. In addition, ultrasonic sensors can detect any item that will reflect a sound wave, so color and material composition will not affect readings.

The basic setup for this project consists of a motor that will drive a lead screw, which in turn controls the position of a lead nut. Attached to the lead nut is a sheet metal target which reflect the ultrasonic pings. The sensor is fixed such that it can sense the relative position of the reflective surface through feedback programs generated in LabVIEW. As the motor drives the lead screw back and forth, the ultrasonic sensor emits chirps that are reflected by the target, and returned to the sensor. Based on information received by the sensor, the motor adjusts the position of the lead nut until it satisfies a user-defined value.

Creating the circuitry and building the basic setup for this project required a thorough understanding of motor control, circuit building, and selection of non-contact position sensors. Mechanically, the basic setup will consist of a mounting plate that will serve as a base for all mechanical elements. On this we will mount a 24V DC gear motor, which will be controlled using the H-bridge circuitry explained in the following paragraphs. The lead screw is approximately 24 inches long, with the far end mounted in a sintered bronze bearing, and the drive end mounted to the output shaft of the motor via a shaft coupler and rubber spider. A target plate made of aluminum sheet metal is mounted onto the custom lead nut. Stabilizers on both sides of the nut prevent rotation. The sensor mounts on a plate above the motor. Since the sensor has a conical sensing range, alignment is not critical. The flexible coupling allows for some shaft misalignment.

Electronically, a small circuit board and the Data Acquisition Card (DAC) serve as the fundamentals for all circuitry. The DAC links LabVIEW to the electronics, which in turn is connected to the mechanical components. The DC voltage supply is used to supply 20V, which can power both the motor (via the motor driver circuitry) and the sensor. Once calibrated, the user will enter a value into LabVIEW and then the motor and sensor will activate and place the target at the desired location. The sensor will be a Model Ultra-30 ultrasonic sensor from Senix Corp. of Bristol, Vermont. It has a sensing range of 4-30 inches.

Bi-directional control of the motor is required, thus an H-bridge was constructed using TIP31 and TIP32 power transistors and 2N222A switching transistors. The schematic of the circuit we made is below.

Fly-back diodes are also built into the circuit to prevent any damage to the control circuitry because of inductive spike. The motor drive circuitry is optically isolated from the inputs as an added safety measure. The sensor does not require any special circuitry, due to a handy self-calibration feature that is an integral part of the sensor. The target is positioned at the closest point in its possible range, the zero point can be set, and once the target has reached its furthest distance, the 10V point can be set. The sensor then reads any distance in-between, with a linear output of 0-10V with an accuracy of +/-2 mm.

The completed project consisted of the following components:

(1) 24” lead screw

(2) Sintered bronze bearing

(3) Lead nut threaded to fit the lead screw

(4) Shaft coupler

(5) Molon 12V DC gear motor

(6) Reflective surface

(7) H-Bridge motor circuit

(8) Senix Ultrasonic sensor

(9) Stabilizing supports

(10)2 x 11/16 Ball bearings

(11)¼” connecting shaft

(12)2 x 1/4” snap rings

(13)Electronics board

To test the accuracy of the device we positioned a long ruler with 1/16” graduations along one edge of the base of the setup. Then the zero point was chosen at 6.75”. We developed a test program in LabVIEW that would allow us to turn the motor on and off bi-directionally and output the number of inches traveled from the zero point. The farthest point that the slide would reach, 23.25” was designated as the 10V point, so we had a total of 15” of travel once the depth of the lead nut assembly is taken into account.

Using the test program the lead nut was positioned at various 1” intervals along the length of the rod, and the corresponding output from the sensor was recoded. The results are as follows:

Distance / Distance
15.5 / 15.48 / 0 / 0
14.5 / 14.48 / 0.25 / 0.26
13.5 / 13.5 / 1.25 / 1.24
12.5 / 12.52 / 2.25 / 2.26
11.5 / 11.5 / 3.25 / 3.26
10.5 / 10.54 / 4.25 / 4.23
9.5 / 9.52 / 5.25 / 5.29
8.5 / 8.54 / 6.25 / 6.3
7.5 / 7.54 / 7.25 / 7.27
6.5 / 6.57 / 8.25 / 8.27
5.5 / 5.56 / 9.25 / 9.29
4.5 / 4.57 / 10.25 / 10.27
3.5 / 3.49 / 11.25 / 11.27
2.5 / 2.52 / 12.25 / 12.25
1.5 / 1.52 / 13.25 / 13.26
0.5 / 0.51 / 14.25 / 14.26
0 / 0 / 15.25 / 15.25
15.5 / 15.48

After the calibration tests, we created the positioning code. This proved quite difficult due to the lag time in the sensor’s readings. Also, the sensor could only perform accurate readings when the target plate is moving very slowly or stopped. This created a number of challenges in timing all the events correctly. A plot of the actual value and the desired value is created in the LabVIEW program while running.

The Senix Corp. specifies that their sensor is repeatable within 2mm, or .076in. We found from the trials using the positioning program that we repeated values at various points on the screw within 1/16”, which is under the manufacturer’s guidelines. In the calibration test, we found that the device was accurate to nearly .02”, roughly 4 times better than claimed by Senix.

As with any precision measuring device, there were numerous sources or error. Firstly, the base plate was thin relative to its length, so some deflection occurs when the lead screw turns. The lead screw itself is not perfectly straight, so it also causes the target to wobble. The target is not quite perpendicular to the lead nut, so if the lead nut rotates slightly (which does occur) the distance may change. Also, things such as air temperature can play a factor due to their affect of the speed of sound. Also, we had to switch which side the target plate was mounted on, because the small area on the front of the lead nut was reflecting pulses and causing interfering signals.

In all, the Senix sensor performed as we hoped it would, and though it did lag a little when positioning the target, it did so with accuracy and repeatability that surpassed the manufacturer’s specifications, even with our slightly flimsy setup. The sensor was as user friendly as it was touted. Aside from some LabVIEW programming problems, and reorienting the target plate, we encountered very few problems with construction or implementation. The circuitry was very reliable and trouble free. We feel we designed a reliable setup that produced good results, and we feel our project was a success.