/diy/ - Do It Yourself

>>623344
If you're going to work in any branch of engineering, you'd damned well better start learning to kickstart your own research rather than asking anonymous people on an imageboard. All the answers you need are on the internet, you just need to figure out the right search terms.

>>623347
You'd better learn how to design a PCB board, too, since there's no way in the nine hells that an Arduino connected to a breadboard would be able to survive the trip.

>>623347

Yeah, I know. I have my own idea of how to do it, but I am open to new ones, or pottentially complements to mine.

I have support at the university, but I'd like to see suggestions of /diy/ and compare with what we have in mind

>>623348

Surely. That's one of our biggest concerns. The "winning" option was the Arduino, because it's friendly with newbies, and strong. But we dont know if it's toughness is enough.

>>623350
Is it something else than measuring acceleration and integrating twice? No?

>>623352

No. The rocket can be launched with an certain angle, not perpendicularly, so the height measure would be wrong.

We'd like to recieve real values, not theorical/expected ones

>>623344
many modern camera phones have GPS & timestamp. if the device can't "timelapse" or take pics at intervals then you may have to hack something together.
Parafin wax is a great potting medium. once your homebrew system is checked and working you can dip the whole thing in warm wax. The wax coating will help insulate, and provide much shock and vibration dampening.

arduino isnt fast enough for a rocket. maybe a parallax would work. each sensor and thrust controller controlled by a separate processor for realtime processing.

>>623391

Is it way more complex, or basically the same of arduino?

>>623344
>electronical details
>height and speed of the rocket

Nigger you do that with tried and true mathematics. No need to put finicky electronics on a model rocket. Don't waste your time. It will be easier, cheaper, and probably what your professor had intended.

You need to know the factors of your rocket, but you're not making your own fuel so it's even more simple using store bought rocket engines.

Links
http://www.rocketmime.com/rockets/rckt_eqn.html#Method

http://www2.estesrockets.com/pdf/2844_Estes_Math_of_Model_Rocketry_TN-5.pdf

>>623344
what's your major?
This is baby stuff.

>>623348
what is perf board?

>>623370
this is good advice, though GPS is pretty shit for elevation data

You want an altimeter. I'm sure someone has already gone throught the hassle of creating the library for a relatively cheap one if you want to connect it to arduino.

found this after 3 minute with google:
http://www.apogeerockets.com/Electronics_Payloads/Electronics/Jolly_Logic_AltimeterTwo

Since you were too lazy to do the search yourself, i'm sure you'd rather take the lazy option like this.

>>623391
you're a fuggin idiot

>>623350
Here's a tip for you for posting on /diy/:
If you already have ideas of your own, post them along with your request for other ideas/comments. It'll help get you taken more seriously.

If you decide to use an Arduino, use the commercially available version for your prototype, then make a custom PCB for your 'production' version.

>>623392
parallel processing means its 8 or more arduinos in one chip. also, more complex.

although, really, you're most likely going to be burning solid fuel which you can't really control the burn on. if you were using liquid fuel then maybe but you're going to need a complex(read:heavy) way to control the rocket with various thrust nozzles and high speed sensing circuitry and some kind of air control and fuel injection system that is going to be far more complex than a simple intake manifold.

I'd recommend a mobile phone with a finder app. gps can measure height. and at its apogee, it'll stay at a particular height for long enough to get a reading because apogee dwell

>>623411
GPS has shit accuracy for measuring altitude

>>623344
Just use a pressure sensor to measure atmospheric pressure

These here should be plenty accurate to measure your altitude:
http://www.st.com/web/catalog/sense_power/FM89/SC1316

Just be careful testing them, they don't work indoors very well.

>>623360
>>We'd like to recieve real values
They you are SOL, there are no sensors out there that measure height.

They may give you a very close measure of your height, but not an exact measure of your height. All sensors are wrong, and you will have to accept this fact.

However, using an extended kalman filter with a NICE IMU(not the ones from sparkfun) should give you a pretty damn good estimate of your rocket's path and speed.(pic related is probably too nice)

You probably don't need the position any better 1 meter right?

Uh, 1 km is pretty lousy for a rocket. At least aim for a couple miles.

>>623447
This.
Im right now building a little Altimeter for model rocketry using an BPM085 (there is a better one now) and an OpenLog to save hight data to an SD-Card.
Everything is available on breakout boards and arduino compatible so this isn't realy hard.

>>623523
i ment BMP

Relying on satellites in space or getting accurate measurements of the ambient air pressure around a rocket moving at Mach 8 seems like it would be tricky at best. The smallest tweak to the line/tube equalising the pressure between outside and the barometer could cause air to be either pushed in (increasing the pressure and thus give too low altitude reading) or maybe even Venturi air out of the rocket (decreasing pressure, too high reading). Would it be possible to measure altitude directly by having the rocket communicate with a unit on the launch platform, like some Doppler stuff or signal return delay or even laser ranging? I'm not an electronic engineer.

>>623460
>Deviate from your PDS by a factor of three, that'll be sure to impress you professors.

>>623530
>>623523 speaking. Atleast for my purpose pessure of muving air is irrelevant. Im mostly interested in the top altitude so the rocket won't be moving verry fast there.

>>623535
Fair enough, but OP wanted speed readings as well.

>>623538
I guess accelerometers are the way to go for speed. I think i actually saw an altimeter that uses the speed measured by accelerometers to account for changing air pressure with speed.

Another way would be a pilot tube.

>>623344
Hey, I was involved in something similar with my university. We had an 8ft fuselage with a custom cast solid composite motor. It was capable of 2 miles AGL with the nozzle we built at about .8 Mach. I built the IMU, launch and recovery electronics.

So... absolute pressure transducers like the MPX line from Fresscale provide more than enough resolution and accelerometers from AD is what I used. The key is that you have to understand the VERY nonlinear relationship between pressure, temperature, and altitude. You can get amazingly accurate results from just pressure if you know your altitude above sea level and current atmospheric pressure conditions. The same basic information can be acquired with 3 axis accels but the math becomes more involved. But with accels you can determine flight profile anomalies like CP/CG precession and motor specific impulse.

I'd be more than glad to give more info if you like.

>>623541

Yes please.

Can you describe with more details EVERYTHING you can remember from your project?

>>623541

and please remember to detail how did you transfer data from the rocket to your pc(bluetooth, anthem, etc)

>>623539
Pitot not pilot tube.

http://en.wikipedia.org/wiki/Pitot_tube

>>623911
Sorry not a native speaker.

Part 1/2

>>623344
Hey sorry, slow reply is slow...

Ok, so first... I built my IMU (inertial measurement unit) before the Arduino/RasPI craze started so I had to build stuff from scratch. And our first rocket was 3.5" in diameter and about 6 or 8 feet tall so a huge dev board would have not fit.

Here's what I designed:

PIC18F2620 running at 40MHz which is 10MIPS
Freescale MPXM2102 absolute pressure sensor
ADXL78 accelerometers for 3 axis sensing
FT232 serial to USB converter for serial output and function programming
Compact flash 1GB for data storage
Audible beeper for finding the rocket in a field, you walk right up on it and not see it sometimes
No telemetry/data radio in this version
4 channels of high current mosfets to fire the pyro charges and separate the recovery section

At the time, I liked PIC's since they are stupid simple to learn and program, documentation is exemplary and programming interface is simple. Also, you can get chips for free if you have a business mailing address. Everyone seems to love the pre-built dev boards but the PIC needs nothing more than power (3.3V or 5V) and a couple capacitors and you're up and running. Get a C compiler for free at MikroEletronika and you are dropping code onto the chip in minutes. If you want low level chip debug and program functionality the programmer is $30 from Digikey or Mouser and it programs everything in the Microchip lineup. With that said I have moved onto ARM core processors at this point mostly due to my job pushing me in that direction but PIC is a great place to start. It is absolutely not a second run junk processor, it's just not open source so libraries are usually proprietary but they are easy to come by in the hobby community.

Part 2/2
>>624194
I would use a fully signal conditioned integrated pressure sensor now instead of the MPXM2102 sensor since the 2102 requires a differential amplifier to get a nice signal and Freescale has sensors with very good linearity that have integrated diff amps built in.

There were no MEMS accels in 3 axis packages at that time so I had to build daughter boards that soldered into the board at right angles for a 3 axis orthogonal sensor. I would use one of the integrated packages now with up to 8g or 10g range and SPI interface for the type of rockets I flew on. Sensing rates can be in the 100Hz range since you won't see much above this range unless you want to do acoustic analysis of the engine.

I would also migrate from compact flash (only thing fast enough when I built mine) to SD since the SD spec is fully fleshed out and is plenty fast now.

I found the PIC ADC to be fast enough and accurate enough if using oversampling and averaging to get 12 bits of resolution. This limited my overall sample rate to 10k samples per channel or something like that but that's actually fast enough for low end acoustic analysis. If I remember, I was dumping data to the CF card at around 160kilobytes/sec running full tilt. That's almost 1.25 megabits/sec which is no simple feat for a relatively low end system like I had.

Part 3/2
>>624195
Later versions of this project incorporated a 434MHz beacon that pulsed my schools letters in a simple morse code pattern every 30 seconds. We used a HAM radio fox hunting technique with a highly attenuated directional Yagi antenna to track down the beacon. This worked so incredibly well we never flew GPS on this rocket. And yes, you can fly the right manufacturer's GPS on a rocket, I flew a 10Hz output GPS on an incredibly fast motor (Mach 7) out of the California desert with very little trouble. The only problem is the lock time, you have to let the GPS find a constellation and then don't break the speed and altitude combination limit. Not all manufacturers implement the speed/altitude limit properly, it's supposed to shut down if you are too high AND moving too fast. So basically most GPS's will auto deactivate during the motor burn but will reacquire after you fall below the cutoff speed. This is a US government imposed limit and the manufacturer's further voluntarily limit it for liability reasons so you can't build a guided missile.

Well, I don't know if I helped or overwhelmed you so what else do you need to know?
Are you able to make your own schematics and layout PCB's?
The best advice I can give is write a simple spec and then find parts that fit the spec and then read data sheets completely and thoroughly (can't stress this last part enough).
For what you are looking to do there isn't much off the shelf available, that's why I built my own.

>>623344
Hey op, I went to a university that has taken first place in a 1 mile national competition several times.

The way to win is first you need good avionics. Secondly you need a rocket with enough umph to get past the mark and use pop out air brakes to slow you down the prescribed amount as computed in flight.

dont die now thread

https://www.youtube.com/watch?v=P5JxXka50Gw

>>626147
These guys are fools, whatever they do don't.

You can easily buy rocket engines that will exceed 1km.

>>623395
could the op use a gyroscope to measure z axis. and its distance from 0

asd