Basejumper.com - archive

Posted this to the wingsuit forum on dropzone.com as well. Afterward, I thought, "Who's really most interested in glide angle information?" My apologies if this is the second time you've seen this post.

A couple of years ago, I started work on an audible glide angle indicator. I put together a prototype based on a Palm Pilot and Bluetooth GPS, and did a bit of testing. There was a small amount of lag, but the information was usable.

I recently started playing with microcontrollers, and thought I'd work on a hardware incarnation of the device. You can see what I've got so far in the attached photo.

The device connects to a Garmin Bluetooth GPS. and uses velocity data to calculate the glide angle. This is converted into a periodic tone whose pitch varies depending on your glide angle. By plugging a pair of in-ear earphones into the jack, you can listen to this in freefall, and get nearly instantaneous feedback on your true glide angle. Other highlights are:

- Device measures 50 x 50 x 20 mm.
- Rechargeable battery (charging circuit built-in).
- Logs data to microSD card.

As you can see, I'm just finishing off this prototype. I've been thinking about what I'd like to do next, and a couple of things come to mind:

- Recharge and transfer data through USB.
- Built-in 20-channel SIRF III GPS.

With a built-in GPS receiver, the new device would be mounted on the back of your leg, with headphone wires running up the inside of your suit.

The upshot of a built-in GPS receiver is, it costs about the same for me as the Bluetooth module, and makes for fewer components in the system. The downside is less flexibility. With the Bluetooth setup, if a higher rate GPS becomes available, you wouldn't need to replace the whole thing.

Anyway, I figured I'd write up a quick post to put the idea out there, and to see if anyone has ideas, suggestions, or is interested at all in where this is going. I realize this kind of thing is not everyone's cup of tea, but if it is yours, I'd love to hear what you think.

Michael

P.S. As with my previous work, I'd be happy to make this stuff "open source", so interested jumpers could build their own. If demand exists for a finished product as well, I'll look into that. Cost will be under $200, I think.

sign me up

why use GPS when you can use a gyroscope on a chip? from analog devices.

are you using a 1 of 5 hz GPS?

A gyroscope will tell you what the orientation of the device is, although it tends to wander a bit over time. This can be corrected using a Kalman filter and an accelerometer, but you're still just getting the orientation of the device, and it's hard to say how this is related to the actual flight path of the jumper.

So, in the end, to get complete information what you really need is a GPS, gyro, and accelerometer. I looked into this for a while, but the complexity is, in my opinion, just not worth it. What matters most for long flights is, ultimately, true glide angle, and this can be measured using a GPS alone.

I'm totally someone who obsesses over data. In the end, though, I felt like this was taking me further and further from what started it all: flying. My goal with this device has been to create something simple, which would kind of give you a "sixth sense" about glide angle, like when you jump off a cliff and can suddenly gauge very accurately how far away the ground is.


The current prototype works only with the Garmin GPS 10, which outputs 1 Hz PVT data.

The SiRF III receiver I'm looking at also looks like a 1 Hz receiver, but I think it may be possible to increase the rate. For me, this would be a great reason to go with built-in GPS.

I should be able to adapt the Bluetooth version to work with any SiRF III receiver, which opens up the possibilities a bit compared to "Garmin GPS 10 only".

Michael

how much does it wander over time? 1 to 2 minutes should not a problem. of course, an inertial navigation system using chip rotation sensors and chip accelerometers seems like the way to go along with gps, but how accurate is the gps for heights or glide angles? do you have a truth test for the glide angle data derived from gps?

The wandering of the gyro isn't too much of a problem. That part can be corrected with a Kalman filter and an accelerometer. The hard part is integrating the acceleration data to get velocity.

To give you an idea what the data looks like, I've attached a plot from a jump I did with a unit built around the Freescale MMA7260Q accelerometer. This is a relatively inexpensive ($10) unit. The vertical lines are every second, with one second of data shown before freefall, and one second after canopy opening. This is 250 Hz data, oversampled from 15 kHz measurements.

The high frequency noise seems to be wind noise. This is with the device sitting in a pouch under my armour's cumberbund, a small pack, and a tracking suit.


The GPS is actually quite accurate for velocity. Using finite differences on position an altitude is a notoriously bad idea, but velocity data derived from Doppler measurements on the GPS signal gives an accuracy of about 0.1 knots. With velocities in the range of 80 knots, that's about as good as you could ask for.

The main problem with GPS is lag. With data coming once a second, you need to hold a particular position for a couple of seconds to determine the effect. A 5 or 10 Hz GPS would be just about perfect for the job.


I have not actually verified the glide angle figures derived from GPS data, but the accuracy of the velocity figures gives me quite a lot of confidence.

Michael

I never completed my inertial navigation system using chips. damn ex-wife took the computer. anyway, my plan was to put it on car dashboard, film the speedometer, and path and drive around for a while to see how well the thing would do.

One day "for the hell of it" we tried to integrate gyro data and accel data for a relatively simple control task. Despite being underwater (and quite well damped) and using a relatively high-end ($1800) sensor, the data was only good for maybe 5 seconds before going entirely out of whack.

What about throwing a triaxial magnetometer into the mix? I know you can get your orientation from 3 accel and 3 mag with some weird matrix voodoo, but don't know how flight would affect the readings.

Steve

sounds like a more difficult task than i had originally anticipated. i may try a "for the hell of it" experiment too. my stuff costs less than $200. geez, how do those space system inertial navigations systems work so well?

I just talked to one of my colleagues who said he got great results from his accelerometer and gyro system built on the cheap. I wonder where the discrepancy in performance lies?

Good question. Beats me... I'm just the embedded guy. The controls gurus tried this setup because the mag sensor was being thrown off by nearby motors doing weird things, but we never could get it to be accurate... oh well, we just got the mag working instead.

As for spacecraft, I think they use these weird fancy laser ring gyros or something. But we can't afford those... strapping one to a WS may be fun tho :)

Seems like you could just use a paragliding Vario many of them are audible offer GPS tracking, glide angle, air and ground speed, and many other forms of information. I was thinking of taking mine on a jump a while back but never got around to it you are kind of inspiring me to give it a shot.

Interesting. The basic idea was inspired by a vario. Do you have a link for a vario that does audible glide angle?

Michael

I think its similar, a vario indicates climbing(more climbing), decending(more decending), so as you fly you know, weather you gained/saved altitude, to fly or not.

You need the drag polar of your equipment (WS/PG) to get a correct glide angle, which is influenced by airspeed and windspeed. Therefore some Varios i.e. Compeo Plus http://www.brauniger.com/.../iq_compeo_plus.html have the McCready function.

Ah, I think the quantity I'm measuring is not properly called glide angle. All I'm calculating is the horizontal speed divided by vertical speed. This is useful for maximizing flight distance, but is sensitive to winds.

With past prototypes, I've spent considerable effort trying to estimate the winds based on a model entered by the user, and then subtract the wind out of the measurements. However, this becomes fairly complex in a hurry, and you never really know if you've got the model right, anyway.

Finally, I decided to keep things simple. If you're flying downwind, the "glide angle" measured by this device will be artificially increased. This isn't really a big deal, since the device is largely comparative. That is, you're mostly interested in the change due to a change in body position. Unless you happen to have perfect pitch, you'll likely have a very hard time estimating glide angle from the emitted tones.

Michael