Group SD0628
From ECE Department Wiki
Truck Brake Augmentation and Electric Jackhammer Page
EE Members
Dan Reineke
Brian Scheer
Anthony Brown
Advisor: Joel Aslakson
EE Member Areas of Responsibility
Dan Reineke: Truck brake project manager
Brian Scheer: Jackhammer 2 project manager
Anthony Brown: Floater
ME Members (Jackhammer)
Tyler Budke
Jesse Ringgenberg
Advisor: Dr. Wayne Reitz
ME Members (Truck Brakes)
Dallas Jorgenson
Matthew Russel
advisor: Dr. Robert Pieri
Truck Braking System
The truck break augmentation project is in it's third year of design here at NDSU. It was formerly at the University of Minnesota, however due to a lack of progress, it was brought to our school.
The basic idea of our Break Augmentation project started around 20 years ago when our sponsor Jim Skarie (a Minnesota native) wondered why someone hadn't invented something for our cold icy winters that would help a vehicle stop on ice.... maybe by dumping sand in front of the tires. After a few years with this in the back of his mind
BASIC IDEA
Jim finally decided to implement his new project. He started working with the University of Minnesota but after a coupleo of years and unsatisfactory results, he moved on to NDSU. For the first two years, the project was strictly a Mechanical Engineering project. In the fall of 2006 Electrical Engineers were added to the team.
Jim Skarie: Inventor and sponsor of our project.
The Mechanical problems have greatly been solved: airflow to bring sand down to the tires, sand hopper design. But how to interface this sand dumper with the vehicle on an operational level and how to detect situations when sand would be needed were problems presented to the Electrical Engineers.
Starting from scratch, we determined that the ideal time to release sand and increase traction was when the Anti Lock Break (ABS) system was not stoping you. We simply called this situation ABS failure. Detecting ABS failure was our first job. This is not as easy as it may sound.
Anti Lock Breaks are really quite simple. When you try to stop on a slippery surface your wheels often lock up. When your wheels lock up you lose almost all of your traction. Anti-Lock breaks release your breaks so that your wheels are no longer locked up and then reapply the breaks until they lock up again. They repeat this process until the wheel rotations when the brakes are released are at a satisfactorily low rate.
Unfortunately Anti-lock breaks don't always work perfectly. If you are on too slippery of a surface, the ABS themselves will lock up and you will have ABS failure.
To detect ABS failure we had to decided right away that we needed to bypass the truck's own braking system which was failing, and go right to the source of the information. We decided to constantly look at the speed that the wheels were turning. From that we decided to analyze when the ABS did and didn't work.
This is a speed sensor from a truck along with it's ideal output
Speed sensors on cars and trucks are also quite simple. They work with a hall effect sensor
(a sensor with a magnet built in which detects magnetic materials) and a big gear with
between fifty and one hundred teeth. The gear is attached to the wheel who's speed it is
sensing and the hall effect sensor is placed so that it can detect the number of teeth
passing passing by. It sends out an electric pulse every time one of the teeth passes the sensor.
Hall effect sensor Output from the attempted inductive sensing below
We wanted to tap into the speed sensor wires directly but didn't want to ruin the truck's own speed sensing setup. We talked to the gentlemen in the shop at Fargo's largest Ford dealership and they told us that it would be impossible to tap into the truck's speed sensors without seriously damaging the truck's speed sensing capabilities. So we decided to see if we could get our information without breaking any wires.
Our first attempt to detect ABS was by inductive sensing. We tried to pick up the electro- magnetic field from the wires leading from the speed sensors to the vehicle's computer. This attempt proved rather futile however because no accurate information about what was going on inside the wire was accurate under a frequency of between 1 and 10 khz. Amplitude of the signal was a problem too. We were able to sense approximately 25% of the signal's amplitude at 1 khz and around 50% anywhere above 10 khz. Since a truck speed sensor wouldn't be able to put out this kind of frequency until it reached around 100 mph. we figured this approach should be thrown out the window.
Since Inductive sensing went out the window, we just decided to brainstorm for a while and finish more of our design. We put together a basic design of parts we might need (depending on what signals we could get from our speed sensors.) We designed a couple of basic sets of pic processors that would be flexible enough to allow for different requirements. We decided to control the sand dumping system we would use the PIC16f876A because of it's simplicity and our previous experience with it.
Because we weren't sure what kind of output we would be getting from the speed sensors or from inductive sensors, we designed a couple of simple amplification and rectifying circuits.
According to our friends at Luther Family Ford we would only be getting milivolts from the system. Receiving this information we started looking for equipment to measure and record whatever signal we were getting: namely a DAQ (data aquisition unit.) We looked over some high quality lab equipment but didn't like the pricerange.
We contacted our sponsor and it turned out that Jim's son James is an EE Masters student at the University of Minnesota and he had a pretty decent piece of equipment for us. In January we came in posession of a 2002 Ford F-150 test vehicle and the DAQBook 120. Here are some of the DAQBook 120's capabilities:
• 100-kHz channel-to-channel scan and gain switching (10 μs);
200-kHz for DaqBoard/2000 Series and DaqBoard/2000c Series Boards.
• 512-location sequence memory that can be loaded with any combination of channels and gains.
• Ability to access up to 256 different channels of DBK signals while maintaining the channel-tochannel
scan rate. The DBK expansion options can accommodate mixed-signal inputs from
thermocouples and RTDs to isolated high-voltage inputs and strain gages.
• Ability to handle 8 differential or 16 single-ended signal inputs without DBK expansion units.
• Ability to handle fixed digital I/O up to 4 TTL lines in and 4 TTL lines out (accessible only if no
analog expansion cards are in use).
Quite a powerful piece of equipment, we were pretty confident we would be able to trace down any signal the speed sensors could put out... and maybe some of their reflections. With the truck and the DAQ we also received a laptop: Panasonic Toughbook with a Pentium 4 processor. The downside of the laptop was that it was running Windows 2000 and only had 256 Mb of RAM. With this equipment one of us stopped out at a salvage yard, cut some wires and harnesses off of a couple of Ford trucks and made some connectors to tap directly into the speed sensors.
Sadly to say, it was nearly impossible to work at any reasonable rate with the laptop. It came with all of the necessary software to run the DAQBook 120 but as to weather or not it could take in data, we never found out. For nearly a month and a half (between the end of January and the middle of March) the electrcial engineering side of the Truck Break group concentrated their efforts on getting readable signals from the DAQ.
It was around this time that we learned that is it worth the investment to get high quality equipment. Through
Dr. Bob Piere we were able able to acquire an older but absolutely sufficient DAQ being used under Dr.Michael
Stewart of the NDSU Mechanical Engineering Department. We reserved the DAQ for the weekend of Easter Break.
Unfortunately the DAQ was lost somewher between two research groups and we were never able to use it.
Since we were already there, Anthony Brown and myself hooked up our home-made sensor taps, 
hopped into the truck
with an Oscilloscope and checked to see if all the planning we had done and the assumptions we had made were worth
the stock we had put into them. 
We were more than extatic to see a well defined signal on the oscilloscope screen!
The guys at the Ford service shop were wrong and we were right. We were getting a perfect sin wave with an amplitude of around
5 volts and a frequency of approximately 10Hz per MPH. At that point we came across the idea of
recording the signal as a .WAV file in a laptop. The main advantage of .WAV files is that they can easily be read
into MATLAB and we could do any analyzation we needed.
Through the help of DR. Bob we got a Laptop computer from the IAAC and got working on our testing. The big problem that faced us now was that it was over a week into April and we wanted to take in data when the ABS were failing. To do this we needed ice. Our only chance was to take advantage of the barely freezing temps in the middle of the night and make our own ice. Between 10:00 pm on April 11th and 3:30 am on April 12th we dumped water on a 100 ft. stretch of the parking lot on the West side of the Fargodome.
Our first test run was at 0412 on Thursday morning. There was ice on many of the streets and although our home-made ice wasn't ready for driving ( or in our case slipping) we had plenty of area to test partial brake failure on.
The official testing began at approximately 0530 on Thursday morning. Here is a video of some of our test runs:
Our data was almost exactly what we had predicted and happy is an understatement as to how we felt. Here are some of the plots of the data we received from the truck on different test runs:
Notice the smooth tapering of the frequency of the signal. Each pulse in this graph is from a tooth on the speed sensing gear mentioned above, passing by the hall effect sensor. Lower frequency = slower truck.
Here is a zoomed in section of the same graph.
NOTE ABS PULSES..................pulse..........pulse.................pulse...................pulse
This is a quick stop at 10 Mph on ice. ABS failure is almost not an issue here.
Now at 20 Mph we finally see a total ABS failure
Again we zoom in on this graph
After finishing these test runs and acquiring our data, we were finally able to construct a reasonably good ABS failure
detection algorithm. It will take some tweaking but is nearly foolproof:
The program based on this flowchart takes data strait from the speed sensors and analyzes it to check for ABS failure. Control of the sand dumping contraption will be left up to another chip for now, for the sake of simplicity in future fine tuning and troubleshooting. There will be one chip for each speed sensor initially. However in an ideal production model another chip would be selected to do the whold job.
One thing that is sometimes hard to remember is that we are the consultants. This is only so much our project as our clients say it is to be. At this point there is little more we can do other than write this basic PIC program and wait until completion of the rest of the machine before fine tuning and testing.
Your ECE ABS Augmentation Team
ANTHONY BRIAN DAN
Jackhammer
The Jackhammer project is in its second year for a senior design project at NDSU, but this is the first year that there was an EE team involved. Basically, the EE team acted as consultants to the ME team, and performed all the electrical research and design portion of the project. The patent for the jackhammer describes a vibration-less, compact, portable and affordable electric jackhammer. The specifics of the patented design call for a mass hereafter referred to as the ram to be engaged by two counter-rotating flywheels, driven via belts from 2 motors, propelling the ram into a chisel. The current team decided to utilize solenoids to engage the ram. Another specific prescribed by the patent calls for the use of a position sensor in the design. This design utilizes an optical sensor to sense the position of the ram so that when the ram reaches the top of its travel, the solenoids will then engage the ram, throwing it into the chisel. The chisel is returned to the top of its travel by a spring, and, to reduce vibration, is damped by a weaker spring. The current design utilizes 120V-AC to power the machine. Some of the more challenging aspects of the project for the EE team were to research the best motors for the job and to down-size the previously utilized electrical controls to fit inside the jackhammer enclosure. The challenge of finding the best motors for the job came due to the fact that there was very little useful documentation from the previous design team. After exhaustive research on the subject, it was found that the best type of motor for the job is the Brushless DC design currently being used. Later on, at the beginning of the Spring, ’07 semester came the challenge of researching the best solenoids for the job. At the beginning of the Spring, ’07 semester, the team had decided upon the use of rotary solenoids in the design. Exhaustive research led to the possibility of several manufacturer’s products. Unfortunately, in all cases, the solenoid having the required size and torque specifications were a made to order product requiring a minimum order of 1000 pieces, making that option prohibitively expensive. The next choice was to use a low profile solenoid with a flat plunger face. This design has a very high force (about 40 lbs) when the travel distance is kept to 1/32” or less. Although these solenoids are a made to order product like the rotary solenoids, a source was found, and parts were ordered. Once the solenoids were specified, it was possible to begin building the power supply to fire two solenoids simultaneously as well as the proper control circuitry. The design utilizes a robust optical sensor to sense the ram’s position. When the ram is near the top of its travel the sensor enables a one-shot multivibrator circuit to send a delayed pulse, adjustable from 1-10 ms, to a second one-shot which outputs a pulse adjustable from 22 – 220 ms. These are preliminary settings, and can be easily changed to fit the design; with the target speed of 600 to 1000 beats per minute, these preliminary settings are in the ballpark. At the time of this writing(April 24), the mechanical design team is awaiting the specified machined parts to be completed. Once these parts are completed, the machine will be assembled and the circuitry will be mounted in the enclosure and the testing and fine tuning phase of the project will begin.
