[simpits-chat] Altimeter Update

[Brian Sikkema]

I really should stop calling it an altimeter... it's really just a timer at this point. By moving down to the Basic Stamp I rather than the II, I've sacrificed the memory necessary for datalogging, and hence, altitude assessment.

Anyway, I've managed to come up with a modified circuit to take care of what I need:
http://www.me.mtu.edu/~bjsikkem/mycircuit.jpg

Rather than using a Basic Stamp II board - a board containing everything, interpreter, memory, etc - I've moved down to the Basic Stamp I chipset, and simply integrated the PARTS into my design. This saved me some decent dough - the Basic Stamp I costs $34, whereas with my setup the total cost at this point is $33.something. It also saved me some parts. I realized looking at this guys design that he has a regulated 5v line for the accelerometer, which just happens to be exactly what the BS1 needs. The BS1 board comes with a voltage regulator on it, but that becomes unneccesary if you have a regulated 5v line already. I also removed the reset circuit, saving myself some more parts and weight.

The only thing not shown on my schematic at this point is the relay that will trip the ejection charge. That will connect to pin12 on the PBASIC1/P.

As I've said, I can't do any datalogging with this setup. The BS2 that this guy had contained 2048bytes of memory, some of which is used for program storage, and the rest held only 12 seconds of data. The BS1 contains only 248bytes of EPROM, again, some of which is used for the program itself. So no room for any useful storage (and no way to upgrade the memory, as the BS1 interpreter assumes 248bytes of memory!)

What I can do, though, is write up a little program to take a look at the acceleration data coming from the accelerometer, and time an ejection charge (or even a staging charge) with this. Here's how it would work:
http://www.me.mtu.edu/~bjsikkem/data.jpg

That graph contains some real flight data from the orginal altimeter, as well as what I suspect the data would have looked like beyond the point where his parachute ejection took place, had it not. As you can see, the phases of flight are easily discernable, and it should be fairly straightforward to discern when they change, during flight. At the beginning you have powered flight, where the rocket is accelerating very quickly in the positive direction (acceleration is in the direction of the rocket nose). Then the engine cuts off, and you have a fast negative acceleration (in the oposite direction of the nose) as gravity and drag pull back on the rocket. As drag slowly decreases, the acceleration approaches -9.8m/s. That's all shown in the actual data. Now for the theory (solid theory, though!). As the rocket passes apogee, the nose begins to point down. The acceleration is still obviously 9.8m/s due to gravity, but now the rocket is accelerating in the same direction as the nose is pointing again. Which means it would be positive acceleration again, at 9.8m/s. 

So, all I would need to do is look for the second time the acceleration passes through 0, and time an ejection charge accordingly. 

Any comments? It all seems like it'll work to me, but I could be missing something here...

Incidently, Analog Devices is about the coolest company ever. I ordered some samples from them yesterday - a couple accelerometers, and an AD converter. I got them TODAY!!! Yeah, that's right, they sent my measily little 3 ICs UPS Next Day Air. For free. I have no idea why. I ordered a sample from National Semiconductor, and they told me 3-15 days, which seems much more reasonable to me. But hey, whatever floats their boat! :)

Oh, and I'm also pissed because the local Radio Shack closed doors this summer. Yeah yeah, I know, crappy store, maybe, but now I have to order even the simplest electronic bits - which is not cool when I'm working on a project! All I needed was a proto board, now I've gotta wait a week for it to get here... bleh.

Brian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.simpits.org/pipermail/simpits-chat/attachments/20031105/41f3fadc/attachment.html