Need help! Looking for a GPS solution

[cwestfall]

So, I'm at my witts end with trying to change the PmodGPS to 10 hz. Dennis Knutson has been a great help in trying to get the system configured, but we are at the point where I need to contact the mfg. I have tried three times prior to Dennis asking me to. They are of no help currently. I cannot get a call back or e-mial. 

Anyone have a GPS unit they like and will support a frequency of 10Hz or higher? I know there are a number of GPS units out there that will do this. I'm looking for one that is user friendly.

I'm trying to finish my assignment for school and I cannot get the PmodGPS to work at the higher frequency. I just need something I can buy, hook up, and find the code for. My assignment has nothing to do with the labView portion. My assignment relates to braking. I just need a way of measuring the braking distance and speed. 

I bought a book to start learning LabVIEW, but right now, I just need to get this testing done. Any advice or help is GREATLY appreciated.

[kevins33]

Hi,

Can you please provide some more specifications on the device that you need? What form factor do you need? What sort of measurements do you need to take? 

Also, if it is not a National Instruments device, I will most likely not be able to provide you a recommendation, however, somebody else might be able to!

[cwestfall]

Hey Kevin,

Let me start off by saying that this is my first real experience with Dataloggers (I have had a few classes where everything is setup and we just start/stop) and I bought the myRIO without having ever used it. I'm trying to find a solution to get this round of testing done. Once I finish, I will be reading the LabVIEW book I just bought and learning LabVIEW.

I'm trying to conduct some vehicle brake tests for a design we are working on (only one semester left!!!). The vehicle will be traveling 60 mph before we apply the brakes and slow the vehicle down to a complete stop.

The main measurement that I'm interested in measuring is distance. Since I cannot figure out an easy way of measuring distance directly, I'm trying to use GPS corrdinates. I'm currently recording time, speed, Long, and latt (in degrees). I have conversions already set up in Excell to take the informaiton and convert everything over to distance from reading to reading.

I'm currently using a Digilent PmodGPS https://www.digilentinc.com/Products/Detail.cfm?NavPath=2,401,1038&Prod=PMOD-GPS. This GPS receiver should be able to sample at 10Hz but I'm not having any sucess (consistantly). I have another thread on here and I have a few people that have been trying to help me get this working but it seems to no avail. It will work on minute then not the next five times I attempt to use it. (Keep getting 1073807253.)

So, I'm looking for something with a 10Hz or higher sample rate. If I can get the Digilent to work, then it will work for us. If not, I looking for a cost effective solution (I'm a student). I found a few National Instruments GPS setups but they were all in the thousands of dollars range. I'm hoping to find one in the tens of dollars (Ideally less than 100).

As an added bonus, it would be great if I could find the code to get us up and running now. I have spend three weeks on the Digilent and I still cannot get it to work (consistantly). I cannot afford to loose that much more time.

If you have any suggestions, I'm all ears.

Thanks,

Chad

P.S. I have thought about making a test rig. Using a wheel to measure the distance and then recording how many times it goes around. (If I know the distance travel in one revolution, and how many revolutions it goes, we could calculate the distance traveled.) I think I can get this done with a decent degree of accuracy (probably closer than the GPS).  

[kevins33]

Hi Chad,

Thanks for providing that information. Unfortunately, it doesn't look like we have a GPS receiver at the price point you are looking at. 

[natasftw]

---

@cwestfall wrote:

P.S. I have thought about making a test rig. Using a wheel to measure the distance and then recording how many times it goes around. (If I know the distance travel in one revolution, and how many revolutions it goes, we could calculate the distance traveled.) I think I can get this done with a decent degree of accuracy (probably closer than the GPS).  

---

If you use this as a solution, make sure you find some way to account for braking too hard.  In that event, your tire won't spin the same way and your measurements will be off.

I'm curious.  Are you braking to a complete stop?  If so, the end point will remain constant for much longer than 100ms.  Have you looked into triggering the GPS readings based on the brake being applied?  You wouldn't need to increase the frequency in that event.  The trigger would give you the starting point.  The ending point would stay constant long enough the slower sampling wouldn't be a problem.  Is there something more to the sampling rate that I'm missing?

[cwestfall]

Hey Natasftw,

The test is being conducted at an initial speed of 60mph or 88 feet per second. With a sample rate of 1 sample per second, our readings could be off by 88 feet, due to when we start braking vs when the sample was taken. If we hit the brake immediately after the sample, we could have traveled up to 88 feet before the next sample. I do understand that while we would be braking and the actual distance would decrease, there would still be error. On top of that, you take into account the error of the GPS itself and our reads could be off by almost 100 feet.

We initially conducted the test at 1 hz. The first set of tests the truck was unloaded and our reads varied between 179.xx to 218.xx (40 feet). Then we added 500lbs to the truck and conducted the tests again. This time our readers were anywhere from 141.xx to 271.xx (130 feet variation). The third sets of tests were conducted with a trailer hooked to the truck and a vehicle on the trailer. Our braking distances were between 261.xx to 349.xx (88 feet).

I understand that our braking distances won’t be exactly the same on each braking event, but this is way too much variance. Each test was conducted 6 times and the readings don’t show heat soak or anything else that would be causing increased braking distances. The readers are just all over the place mostly due to when the sample was taken vs when the braking event occurred.

If we end up fabricating a test rig, the wheel will have a disk connected to it with slots. We will use a Hall Effect sensor to measure RPM of the wheel via the slots. If we put a slot ever 45 degrees, then we will be able to calculate the distance traveled with an error of roughly .4*diameter of the wheel ((pi*d)/(# of slots)). A wheel 18 inches in diameter would give us an error of 7 inches roughly. That is way better than the 100 feet we are currently getting.   

Do you see any errors in that logic? We are open to suggestions. We would really rather not have to fabricate a test rig is possible.

[natasftw]

If distance is the measuring point, set the measurement to start on a trigger.  When the brake is pressed, the trigger can be set.  This requires enough of a rig to work with this switch.  But, it means you won't see the 88ft drift because your measurements won't start until you trigger them.

The second experiment idea is solid if you're sure you won't ever lock the wheels during braking.  If you do, there will be added error to the measurement.