Fun with a ISU Inclinometer

For a while now, you've been able to buy digital levels. Amazingly, they will display angles within 0.1 degree.

As some of you know, I've been on a quest for a good way of mapping hills, and I found that you can order up a inclinometer from the same company that makes the digital levels. Which I did.

It's a pretty cool device - you give it 5 volts, and then it gives you the current angle back through a serial port. In the simple mode, it will send you this data every 5/8 of a second.

So, this weekend I hooked it up to a level converter (the inclinometer speaks TTL rs-232, so you need something like a maxim 232 to convert that to real RS-232), connected it to my laptop, and exercised the new Whidbey SerialPort object. I wrote a little logger, and headed out to gather some data on the hills around my house. Thirty minutes later, I had my data, and I imported it into excel.

And it looked weird. Some of the data looked really nice, and some of it had big spikes in it - both in the positive and negative directions.. Took me about twenty minutes to figure out the problem.

Care to guess what I forgot, and how I'm going to fix it?

Comments (10)

  1. Niall says:

    Speed humps.

    Or small children.

  2. Matt says:

    You attached it to the spokes

    You like pulling wheelies and stoppies

    You used a short for your data storage and over flowed 🙂

  3. Does acceleration and deceleration affect the readings?

    If so I suspect you’d attempt to correct for it by using your gps readings to figure out your acceleration and determine a correction factor. But that seems complicated and depends on having very frequent samples of location info…

  4. Was it the fact you were using "Whidbey" and not "Visual Studio .NET 2005"? =)

  5. Radeldudel says:

    Maybe you forgot to align it properly to the ground?

  6. ericgu says:

    Stuart is correct.

    When you accelerate or decelerate quickly, the gravity vector is skewed towards the front or back, and the inclinometer responds to it.

    My plan already includes a rotation sensor on my back wheel, so I’m going to use that to figure out the speed, take the derivative to figure out the acceleration, and then use that to adjust the level data.

    Which means that in the end, I’ll have built both a hill mapper and a g-meter.

    Radeldudel, I don’t bother aligning it. I just start with the vehicle in a known spot, record that angle, and then finish at that same spot but in the opposite orienatation. The average value of those is the level angle.

  7. Karch says:

    How much did the ISU unit cost – what are you using for a rotational sensor?

  8. ericgu says:

    The ISU unit was around $100. Yes, pretty pricey.

    For a rotational center I’m just going to use the magnet and sensor from a bicycle computer.

Skip to main content