GPS audible glide angle

[crwper 7]

Here it is! I had a chance to test the new changes this weekend, and everything looks fine. It was a bit windy, but I was able to make two jumps. The data from the second jump is attached.

On the first jump, I tracked straight into the wind. The software seemed to give almost immediate response to changes in body position. For this jump I maintained a glide ratio of about 0.3, so I set the program with a range of 0.1-0.4 for the second jump.

On the second jump, I decided to try something a bit different. Immediately on exit, I tracked crosswind. The audible beeped like crazy. I changed to a more neutral body position. It was still maxed out. So I turned into the wind and tracked that way. This brought the glide ratio down. I turned crosswind and tracked a bit more, increasing the glide ratio, until shortly before pull time when I returned to a neutral position. This is all shown in the attached screenshot from the software. Note that the sharp increase in glide ratio around 40 seconds is me pulling. Sorry, no simulated audio to upload yet.

The software is now capable of dumping out GPS data to a comma-separated file. I've attached this file as well.

I’ve done some analysis on the data. It looks like the “black box” Garmin uses to calculate vertical velocity is not well-suited to this application. In particular, I’ve attached a plot of the vertical velocity over time. It seems pretty clear that they are fitting chunks of data to a quadratic function, which does not match velocities estimated from altitude data very well. I'll be thinking about ways to improve this.

Finally, I've attached archives with installation files and a bit of documentation ("bin.zip") and source code ("src.zip"). If you use the software, your feedback would be much appreciated.

As it stands, the program will work with any Garmin GPS which can be connected either by Bluetooth or through the cradle connector. You will need a GPS, Palm OS handheld, and a pair of earphones you can hear in freefall.

Coming soon:

• Multiple graphs of data for post-jump analysis (like Tracking Derby)


• Store and view more than the most recent jump


• Audible feedback for values other than glide ratio

That's where I'm at. Let me know what you think.

Michael

[Costyn 0]

Very cool, thanks for all the goodies.

Now where to find a Palm PC. Are there minimum requirements for the model? (USB/bluetooth and earphones plug are obvious, of course).

Cheers,

Costyn.

Costyn van Dongen - http://www.flylikebrick.com/ - World Wide Wingsuit News

[SkymonkeyONE 3]

You techies amaze me. I wish I had enough brain power left in my beat-up 43 year-old body to begin to understand that shit. Seriously. That's cool shit.

Chuck

[cpoxon 0]

Very cool Michael! Not only for developing it, but for releasing it as well.

Unfortunately I don't have a funtional Palm at the moment. I wonder if it will work under a Windows Mobile Emulator? Emulators tend to struggle when you get down down to the port layer though.

Skydiving Fatalities - Cease not to learn 'til thou cease to live

[crwper 7]

Quote

Costyn: USB/bluetooth and earphones plug are obvious, of course

Well, the program uses an ARM callback to generate the sound stream just-in-time, so I suppose the PALM will have to use an ARM processor. Anything newer than my old Tungsten T should do it. I think any Palm running OS 5 has to be ARM.

Other than that, the connection is either through the cradle connector (which goes way back) or through Bluetooth. If you go through the cradle, there could be a bit of tinkering to come up with a cable. I made my own using an old Garmin serial cable, Palm serial HotSync cable, and a null modem adapter and gender bender between.

If you find a Palm with a connection other than these two which you would like to see supported, please let me know. It would take about 2 minutes to add another port to the list.

Quote

cpoxon: I wonder if it will work under a Windows Mobile Emulator?

I doubt it, but would love to know the results if you give it a shot. More than the port stuff, I think the ARM callback might throw it for a loop.

Michael

[Waldschrat 0]

Is it also possible and usefully to set a range between 1.0 to 2.2?

Can i export the data from the Palm to Paralog for my Jumpbook?

Thx

[crwper 7]

Quote

Is it also possible and usefully to set a range between 1.0 to 2.2?

The minimum and maximum glide ratio can be set to anything between 0.0 and 25.5. The tighter the range, the better you will be able to distinguish small differences in glide angle. On the other hand, if the range is too tight then a bit of wind will push you right out of it, so it's a bit tricky.

Quote

Can i export the data from the Palm to Paralog for my Jumpbook?

There is no direct export to Paralog at this point. However, it is possible to export data to a comma-separated file in Memo Pad. This is usually synchronized to the PC on each HotSync, and can be accessed in Palm Desktop. A sample of the data output is found in "GPS Data.txt" in my previous post. If Paralog can read that, you'd be set.

Michael

[ove 0]

Hello!

Sounds good!
Can I connect my Garmin Etrex Vista C to my palm? Where to buy cabel? Bluetooth adapter?

Special earplugs? Both ears?

Ove

Do, or do not. There is no try.

[crwper 7]

Quote

Can I connect my Garmin Etrex Vista C to my palm? Where to buy cabel?

You can. The cable shouldn't be too hard to get. Google gives a few hits for eTrex to Palm Universal cables. The problem with this kind of set-up is that the Palm Universal connector isn't really meant to be used in mobile applications where stress is applied to the connector, so it has a tendency to come undone.

Quote

Bluetooth adapter?

For the Palm, or the eTrex? Again, Google turns up a few hits on Bluetooth adaptors for the eTrex, but they all seem to be something that plugs into a cigarette lighter. The eTrex doesn't supply power through its connector, so you'd need to find something that has its own power. There are many Bluetooth adaptors available for the Palm itself.

If you can do it, wireless is the way to go. I built my first prototype using a hacked-together Palm-Garmin cable and a Geko 301. I actually never got it to work, because every time I headed for the door, the Universal connector would come undone. I currently use the Garmin GPS 10 and a Palm Tungsten T (has Bluetooth). The difference is incredible. Instead of fussing with a cable, I only have to manage the Palm and earphones on the way up. On jump run, I tuck the GPS into a pocket on my ankle and I'm off.

Quote

Special earplugs? Both ears?

I use Etymotic Research ER-6i earphones. These provide excellent noise reduction. They are always available on eBay. I have been wearing just one when I jump, so that I can hear other jumpers just before exit. There is no problem at all hearing the audio in freefall.

Michael

[vidiot 0]

Quote

Quote

Special earplugs? Both ears?

I use Etymotic Research ER-6i earphones.

Sony Earplugs work well too, or better get one of the noise cancelling types if you regularly jump a Skyvan.

My Logbook

[crwper 7]

Quote

or better get one of the noise cancelling types if you regularly jump a Skyvan.

I've heard that consumer "noise-cancelling" earphones are not as good as "isolating" earphones like the ER's or the Sony MDR-EX line. Have you tried them?

Michael

[TomAiello 25]

Quote

I've heard that consumer "noise-cancelling" earphones are not as good as "isolating" earphones like the ER's or the Sony MDR-EX line. Have you tried them?

I've used both. I think the "isolating" ones work better than the noise cancelling variety.

Also, I found that some of the cheaper noise cancelling types could give me a headache over time.

-- Tom Aiello

[email protected]
SnakeRiverBASE.com

[Costyn 0]

Yep, Team Fly Like Brick can vouch for the in-ear earphone types like the Sony MDR-EX71. We have all jumped occasionally with these in freefall, for listening to music during the flock. Good fun, but a bit distracting.
Yes, they are expensive. Yes, they are worth every penny.

Costyn van Dongen - http://www.flylikebrick.com/ - World Wide Wingsuit News

[crwper 7]

Quote

I’ve done some analysis on the data. It looks like the “black box” Garmin uses to calculate vertical velocity is not well-suited to this application. In particular, I’ve attached a plot of the vertical velocity over time. It seems pretty clear that they are fitting chunks of data to a quadratic function, which does not match velocities estimated from altitude data very well. I'll be thinking about ways to improve this.

I've looked a little more into this. As it turns out, my GPS was set to "automatic" velocity averaging. I've set it to "none", but it will be a couple of weeks before I can try it out. Hopefully this resolves the problems with vertical velocity.

Michael

[ove 0]

Hello

Might buy an gps 10, arn't that expensiv.

Is there a palm version, and a pocket pc version, or are they all compatible?

Ove

Do, or do not. There is no try.

[Cloggy 1]

Quote

The minimum and maximum glide ratio can be set to anything between 0.0 and 25.5. The tighter the range, the better you will be able to distinguish small differences in glide angle. On the other hand, if the range is too tight then a bit of wind will push you right out of it, so it's a bit tricky.

Just an idea: how about a self-adjusting range? Don't know if that would be feasible on the palm or work in practice, but it could give you both a wide range and good sensitivity combined.
Like a small range that centers itself again on the average glide ratio. Mixed with beeps that indicate the average glide ratio to give a reference to the audio feedback.
Or several distinct overlapping windows (like 0-6, 4-10, 8-16). With the audio feedback mixed with e.g. 1, 2 or 3 background beeps every few seconds, so you know in which range you are.

(It's fun playing with the Palm: tried years ago to convert my IIIc into a variometer for paragliding. Never got that to work...).

[crwper 7]

Quote

Just an idea: how about a self-adjusting range? Don't know if that would be feasible on the palm or work in practice, but it could give you both a wide range and good sensitivity combined. Like a small range that centers itself again on the average glide ratio.

This is something I've considered. Basically the effect is that the beeps would indicate change in glide ratio, rather than the glide ratio itself. I think this might actually be more confusing in the air. Since it is not possible, really, that you would constantly be improving your glide ratio, you would always wind up chasing small improvements, without any real indication of how they stack up over time.

Quote

Mixed with beeps that indicate the average glide ratio to give a reference to the audio feedback.

If I understand correctly, I think this could also be confusing. Having tried this thing out a couple of times, it's certainly fairly obvious when you're doing well and when you're not, but if you try to include more information in the beeps (for example, indicating the window position), I think it would become considerably less intuitive.

Quote

Or several distinct overlapping windows (like 0-6, 4-10, 8-16). With the audio feedback mixed with e.g. 1, 2 or 3 background beeps every few seconds, so you know in which range you are.

Again, I think this may add more complexity than is necessary. My feeling is that the device should be as simple as possible. If it turns out you have very complex wind conditions, for instance, it might just not be the right day to play with this particular toy.

What I have in mind at the moment is just to subtract an assumed wind vector from the horizontal velocity before doing calculations. Data would be stored without the correction made, so that when you get down to the ground you could make adjustments to your assumption and re-interpret the results. Basically, you would use the results from one jump to adjust the wind assumptions for the next.

The wind vector could be entered manually, or possibly calculated from an existing jump if certain things are known about the jump itself (for example, if you can select a period of time where you performed a constant-rate turn).

Michael

[vidiot 0]

Quote

I've heard that consumer "noise-cancelling" earphones are not as good as "isolating" earphones like the ER's or the Sony MDR-EX line. Have you tried them?

I used both extensively while on commercial flights. The noise canceling type has a slightly better noise surpression but they are not worth the added bulk and $$$, IMO.

@ Costyn: The Sony MDR-EX71SL earbuds are only about $30.

I am not too happy with the cables during skydiving, though, so I'm going to try a BlueTooth headset with PocketParalog over the weekend, assuming the wheather holds...

My Logbook

[Costyn 0]

Quote

The Sony MDR-EX71SL earbuds are only about $30.

When I bought them 3 years ago they were about 50 euros. Funnily enough I've repaired for about the 4th time just tonight. The tiny wires inside get worn out after a while and require patching up with a soldering iron and some gaffer tape. But then again I have used them nearly every day since I bought them.

Costyn van Dongen - http://www.flylikebrick.com/ - World Wide Wingsuit News

[crwper 7]

Here is the latest version of the vario program. As before, I've attached archives with installation files and a bit of documentation ("bin1.zip and bin2.zip") and source code ("src.zip"). I have not yet had a chance to update the documentation, so I'll describe the new features briefly here.

First, I've opened up the display a bit. The "status" area at the top is reduced to a single line. On the right is GPS status and on the left (if you're connected to the GPS) is some piece of information from the GPS. If you tap on the status area, this will switch between time, altitude, climb rate, and horizontal/vertical position error.

Second, you can view more than just glide angle (although the beeps are still glide angle only). To change the plot, use the drop-down list in the top-right corner. You can switch between altitude, horizontal distance, glide ratio, vertical speed, horizontal speed, and bearing. The glide ratio plot goes between the minimum and maximum defined in the preferences. All the other plots go between min and max for the current data set.

Third, these plots aren't quite as simple as they seem. Under Options, I've added a "Wind" menu. Here, you can define the winds at altitude. The wind speed and direction (actual direction of the wind, i.e. where it's going to) are subtracted from the data for (a) calculating glide angle, and (b) displaying the various plots. That is, the data is relative to the moving air mass. If you don't like this, just set the wind to zero speed. Otherwise, it can be used to correct for wind so you still get meaningful data.

That's it for the big changes. For those looking at the source code, I apologize that I have not yet had a chance to clean it up very much.

I also have not had a chance to test the latest changes (except testing on the ground). If you download this version, please save the old one just in case. If you try out the new version, let me know how it works!

On the way... I will be changing things so the jump data is stored in the Palm database, and so you can flip between many past jumps to compare. I'll also be adding options for the beeps, so you can have the program beep depending on your horizontal speed, for example.

Have fun!

Michael