Hi Everyone, just thought I 'd share an update on something I 've been working on

[G+_Chris O'Riley]

Hi Everyone, just thought I'd share an update on something I've been working on. I posted some early details last year, but it's come a long way since then.

 

I used to love making little parachutes when I was a kid. There was just something about watching them unfurl, open and then gently glide to the ground. When I'd make them for my son, I noticed that he couldn't reliably throw them to the same height. If they didn't go high enough, they wouldn't open until they were almost to the ground (if at all!). So I'd fold and wrap the parachute looser, and then he'd throw it high and it would open too early on the way up.

 

So I thought I'd try throwing some technology at the problem! Long story short, I'm reading a Bosch BMP 280 barometer with an ATtiny85 as it's increasing in altitude (thus the atmospheric pressure is decreasing) and controlling a servo to release a parachute as soon as it detects that the pressure is starting to increase, indicating that it's starting to come back down.

 

I started by breadboarding an Arduino Nano triggering a buzzer instead of a servo, wrapping it all in foam and throwing it up to make sure my code was working. It'd beep when it got to the top, so I'd know the basic concept was working. I moved on to a rat's nest of wires, soldered directly to the legs of an ATtiny85, just to make sure it would work on the actual chip I ultimately wanted to use.

 

With that working, I designed a small circuit board in KiCad and had one made up through OSH Park. Last week, I started on the housing and release mechanism, which I have working with the first version of my circuit board. I had hoped I'd be able to power it with a few large coin cells, but the current draw of the servo is just too high, so I picked up a small LiPo and it looks like it'll work great. I expanded the circuit to include a LiPo charger IC and a boost regulator to bring the single cell LiPo up to 5 volts. I sent that out to OSH Park last weekend and started finalizing the housing.

 

Here's a video showing the release mechanism, running from the first version of the circuit board. To activate it, I'm just pressing on the BMP 280, which reads that as an increase in pressure which triggers the servo. I left the bottom of the bottom shell open so I could inspect the mechanism as it was working in case anything was jamming or not moving far enough. So far, it seems to work pretty reliably.

 

By the time the updated circuit board arrives, I'll have a full bottom shell printed and should be able to test everything in a pretty darn close to complete state. Once any issues are worked out of the housing, I'll probably have Shapeways print one in nylon on one of their SLS printers, but I already checked and that'll cost about $50, so I definitely want to make sure the design is good.

 

Anyway, I figured I finally had this working enough to show as a generally working concept vs. an idea and a hobbled together bunch of bits and pieces!

 

Here's a video of the release mechanism working: https://www.youtube.com/watch?v=-NDtWjtEi0Q

 

I'd love to hear what people think!

https://www.youtube.com/watch?v=-NDtWjtEi0Q

[G+_Chris O'Riley]

Here's a photo of the circuit board. The pin header on top lets me reprogram the Tiny85. On the breadboard I was able to reprogram it without having a shared ground between my circuit board and the Nano I was using to program it, so I didn't include a ground on my board. As it turns out, I do need that, so I had to drill a hole for an extra pin, luckily there was just enough room for this and it was in a solid ground plane to solder to. I updated this on the new version of the circuit board, so it won't have this horrible hack!

[G+_Michael Larson]

Brilliant! I need a 3D printer...

[G+_Rickbearcat]

You should patent your device. Find a company over in China to make these for a buck then market them to toy stores and chains for $4.99. I'd bet you'd make a killing selling these to 5-10-year-old kids. If you could find a way to display the height of parachute release, this could become a game of "who can throw theirs the highest" contest.

[G+_Robert Gauld]

I doubt it'll be patentable since he's shared the idea publicly and I believe I'm right in saying the model rocketry folks have something similar already (meaning there's prior art).

That said neat idea. Since you know the difference between release pressure (highest reading on the way up, the highest overall will be the ground when it lands) and highest point reached (the lowest reading) then you can easily get the difference and calculate the height thrown. I don't recall the formula but it should be easy to google.

[G+_Paul Hutchinson]

Very clever.

 

FYI barometric pressure varies roughly 1 inHg per 1000 feet.

[G+_Chris O'Riley]

Thanks everyone.

 

Regarding the patent, yea, I'm pretty sure I'm not the first one to use barometric pressure to release a parachute, like Robert said, the same concept has been used in rocketry. And a patent doesn't really protect you unless you have the means to enforce it... which I don't. Too busy making toys for my son! ;)

 

I'm still refining the code, but so far, I've gotten the best results not worrying about conversions to actual altitudes or pressures, instead just reading the raw pressure value from the sensor and comparing it to prior readings. I could certainly store the high value in eeprom memory to be read back later. To save power, as soon as the parachute is released, I put the Tiny85 into sleep mode until a press of the button, which just resets the chip.

 

The BMP 280 is read over i2C, so I could add a small i2C screen to display information. I've been trying to keep compact so I can fit as large a parachute as possible inside.

[G+_Sushal Penugonda]

Great work, I feel this had good applications in drones. However the trigger mechanisms would be different: detect g - forces, and power to drone to detect free fall and then deploy safety chute to save the drone from crashing to the ground

[G+_Chris O'Riley]

Hi Sushal, that was actually where the idea originated! I figured it could make a fun toy for my son as well, so that's what I started on. The toy is a good test bed to refine the code, which is where I need to focus next.

 

Firmware/flight controllers like APM/Pixhawk are already capable of detecting a failure and deploying a parachute, you just have to connect a standard servo to one of the outputs. Not sure if Betaflight/Cleanflight/iNav have similar functionality. If not, the barometer could be replaced with an accelerometer, or just programmed to detect a decrease in altitude above a given threshold.

 

I have some ideas for a release mechanism for drones that I'll investigate after the toy is finished. Biggest issue is clearing the props if any are still spinning. Arducopter conveniently has the functionality to stop all the motors before deploying the parachute, but it'd be nice to be able to reliably get a parachute out even if some were still running.

[G+_Black Merc]

Chris O'Riley try for a patent anyway but, focus on your improvement over other patents... You can still get a patent if it improves on 'prior art'. Just credit the 'prior art'... Open Source has the same rules, your improvement sence the previous artist finished. Just give credit...

[G+_Chris O'Riley]

Here's an updated PCB. I'm now powering it from a single cell 110mah lipo, and didn't want to mess with a battery compartment to get the battery out for charging, so I integrated a lipo charger IC on the board (under the USB port), and replaced the step down regulator with a boost regulator to increase the lipo's ~3.5 volts to 5 to power everything.

 

So far, I've been testing it for about a week, and it hasn't reached the 3.3 volt cut-off I have programmed. It doesn't use much power - only about 20 milliamps while it's reading the barometer, a second or two of ~80ma while the servo is activated, and then I put the ATtiny into a sleep mode until the button is pressed again. In sleep mode, I think it's under 2ma. I'm curious how many times it can be thrown before the battery needs to be charged.

 

It's all working, I'll try to get some footage of it soon, but I already plan to revise the release mechanism and housing to work better/easier.