Hobby Rockets - Where to find accurate sensors and How to transfer data?

[gurenmarkv]

Hey there. I am currently researching hobby rocket stuff. My team plans to launch a rocket 30,000 ft in the sky. I am looking for help on three things.

1. Microcontroller - Looking for a fantastic microcontroller to implement the sensors and communicate data. So far just looking at options from Arduino due to various options and robust library. 

2. Sensors - Where to find high precision sensors that will work with said microcontroller. Need stuff like GPS, Gyroscope, Temperature/Humidity, Pressure maybe even camera. The type of accuracy I am looking for is stuff like real time GPS, sensor able to recognize up to 30G pressure, gyro that can give rotational data etc. 

3. Communication - Looking for a method to communicate the data from the sensors live from the rocket. Current idea is to use HAM Radio. Currently limited to the public wavelengths that do not require license. So transmitting the GPS, Gyro, Pressure in real time is something I am looking for. Best possible option would be a method that can also transmit video of any quality. 

Currently thinking of getting arduino setup to work with sensors, transfer that data to a raspberry pi. The pi would then be connected to the radio stuff to transmit the data to a base station. The base station would do calculation and plot the data. Unless if the arduino is capable of transferring the data through a module of some sort. 

Thank you for any help. 

[Gavinmc42]

Hack one of these? Add GPS module
http://erlerobotics.com/blog/pxfmini/

Not enough sensors?
http://www.pi-in-the-sky.com/

A combination of these two/
LoRa for telemetry.

30,000ft is not a hobby, that's rocket science

[gurenmarkv]

Hack one of these? Add GPS module
http://erlerobotics.com/blog/pxfmini/

Not enough sensors?
http://www.pi-in-the-sky.com/

A combination of these two/
LoRa for telemetry.

30,000ft is not a hobby, that's rocket science

Interesting. Thank you for this information.

[Elfen]

For something small and lightweight, I would recommend the Pimoroni's Enviro pHat sensor board for the R-Pi. It's the same size as the R-Pi Zero/Zero W and has more or less the same sensors as the R-Pi Sense Hat but without the LED Array and joystick. and its about 1/2 the price of the Sense Hat, depending on where you go.

You do need to import the Python Libraries, but if you are at this level of sophistication, this is no biggie for you.

The Enviro pHat and a R-Pi Zero, though you need to solder headers and things, will get you most of what you need. Things like GPS and other items you need to add on separately.

Enviro pHat Link: https://learn.pimoroni.com/tutorial/san ... nviro-phat

[knotty]

Hi,

My sons and I are venturing on a similar project however with slightly lower (altitude) aspirations!

We are using ESTES rockets with BT60 payload sections so ‘space’ is limited (excuse the pun).

We are more focused on recording the flight video and overlaying telemetry data which can be viewed after the launch.

Our setup is a Pi Zero W, ZeroCam FishEye, BerryGPS-IMU v2, PowerBoost 1000C and a LiPo battery.

The BerryGPS-IMU v2 is a great bit of kit and will provide all the telemetry info and more.

I’ll post again when I’ve got more info/photos/etc.

Regards,
Paul.

[Gavinmc42]

Cool, that BerryGPS_IMU stuff is nearly perfect, just needs a LoRa transceiver.

Got into electronics nearly half a century ago just for rockets and rc planes.
The first things I made were tiny FM transmitters so I can track and find those rockets/free flight planes again.
The Estes Rocketcam kit with the 110 camera was expensive and cost a bit of pocket money.

Making PCB loop square spiral antennas with a Dalo pen was the start of my PCB design career.
I still design PCBs to this day as my job, but it all started due to Estes.
Now my PCBs plug onto Pi's.

What I find interesting is this Zero/GPS hat solution is about the same size as the 110 camera.
I just had to wait 50 years for a movie version with full sensor payload
Hmm, can roll rate be determined or is another sensor needed,
I used to use a light sensor modulating the FM and recorded it on a FM receiver/tape setup.

Wonder what G's the rockets get up to these days? That IMU goes to +-16G's.
Is there a rocket Hat?
Rocket glider with GPS homing? Bring the egg down for safe landing?
Use optic flow code for landing, Pixart or Broadcom sensor.
First surface mirror for the camera for launch/ground view, extra Zero for another side view camera won't add much weight.

Hmm, that is nearly the same tech as cruise missiles.
Flashback to the 1984 Tom Selleck movie Runaway, guided rocket bullets, Gene Simmons from Kiss as the bad guy.
Heat seeker bullets- 80x60 flir sensor?
I slowmoed those spider robot scenes over and over to figure how they worked.
Short Circuit in 1986 pushed me over the edge and back to school for a Grad Dip in Robotics.

It all started with those Estes rockets, back away Paul, save your kid

[bensimmo]

Egg down safe is all part of the CanSat by the ESA (which is what we base our own school fun we I mean education challenge on).

http://www.esa.int/Education/CanSat

So far we've only use Arduino and loads of add-on boards and then simplified with Pi A+, ublox6 and SenseHAT and camera.

Things have moved forwards but no kids are doing the challenge at the moment, so I don't get to use Zero's with them yet.

[bensimmo]

The Envirophat would make it easy to start to play with but not for your final use as your specs are higher, but you can buy the sensor modules individually and hook up over I2C, though ADXL377 (+-200G MEMS) is 3x analogue, so 'harder' to hook up to a Pi. (May not be suitable for your use).

Don't know what with your specs though, look around see what is there for your forces, accuracy and frequency of update.
See what other high power rocketeers are using and what you country allows.

GPS via UART or usb. The advantage of not using a 'HAT' is you can place them where you like.
And a camera with a Pi Camera (if using the Pi).
Are you planning on any control of the rocket?

Communication is the harder part in my opinion, mainly as I've not looked into it in depth,. HAB have the benefit of being slow w.r.t. Rockets but they do the distance.


30k feet (~9km) is a lot higher than we go which is 1km (same rocket as the CanSat one) but we are capable of approx 4/5km with our second rocket.

But check out what they use https://ukhas.org.uk/general:beginners_ ... ballooning

[knotty]

Hmm, can roll rate be determined or is another sensor needed

I'm not sure about roll rate but I quite like the idea of inclinometer to show pitch/roll overlay in the corner of a video. It looks like a lot of the leg work has been done here:

http://ozzmaker.com/inclinometer-using- ... ry-pi-imu/

However I've got zero experience in coding and I'm struggling with the basics (if anyone can help!)...

viewtopic.php?f=63&t=108410&p=1250879#p1250879

The hardware aspect is nearly finished. As its all fitted inside a standard BT60 payload tube (for compatibility) I'm short on space so I need a low profile USB (I'm going to hack a wireless charging receiver) and then fix it to the tube as I have with the camera. I've attached a couple of pictures of the work in progress.

[Gavinmc42]

I started with Python on Pi's, a good place to start learning.

Is that a Zero W?
You want to overlay data onto live video?
For rockets you need to collect data fast so probably save into a memory buffer.
Don't try to save to SD card, until you are going down or landed.

I would collect video and data separately and do the overlay in post processing back on the ground.
You can overlay some stuff one line or so?
Pikrellcam uses this feature to timestamp videos.

Can it be done fast enough in Python?
A good place to start would be with Picamera software
http://picamera.readthedocs.io/en/relea ... index.html

A very simple video camera system can be done with the Ultibo examples
https://github.com/ultibohub/Examples/t ... V/RaspiCam
The advantage is you don't need to learn Raspbian Linux.

WiFi is not yet working in Ultibo, but I did make a basic webcam app very quickly by combining 2 examples.
https://ultibo.org/forum/viewtopic.php? ... bcam#p5877

Steaming video is not that easy to do.
MJPEG seems to be the fastest as the VC4 can make the JPEGs in hardware.

You might also try lowering the resolution of the camera to get higher frame rate.
Some amazing speeds are getting done now, above 600fps, but not beginners stuff.
viewtopic.php?f=43&t=109137&start=325

But perhaps stick to 640x480 90fps to start with, should be able to capture that to memory but I don't think you can steam that fast.
https://www.raspberrypi.org/documentati ... /README.md
Lots of info on google for 640x480 90fps slow mo videoing.

You might want to practice all this on a RC car before attempting a rocket, works out cheaper and easier to fix bugs quickly.
DIY drones should give you some ideas.

Ultibo 2.0 has all the Videocore stuff and OpenVG has a an extension that pops jpegs on the display.
https://ultibo.org/forum/viewtopic.php? ... enVG#p5412
Being playing around with it for glass cockpit displays but it could be turned into a ground station display?

I gave up trying to do this in Linux, I am not a good enough coder, it may just be possible for me to do it with Ultibo one day.
Collect data and video at the same time? That would a be a challenge
I do know i2c comms in Ultibo works better, those sensors are i2c or SPI, SPI will allow faster sensor updates.
Good drones get sensor data at 400hz updates, how fast is enough for a rocket?
But might be easier to use two Zero's.

This Sentral based IMU seems to be the best for lazy coders like me
ttps://www.tindie.com/products/onehorse/ultima ... -solution/
The Accelerometer is +-16G and those egg rockets get up to 20G's, so the accelerometer might get pegged on launch?

All possible, just not easy, but with Pi's you really don't need to be a rocket scientist any more.
Pi drones have been done now, Pi rockets the next frontier?
Scale ME163 with hybrid rocket motor? Anyone done that yet? ME263 would be better?

[bensimmo]

We have a larger payload to play with (are you parachuting it on it's own or is that to stay in the rocket ?)

WiFi and AP setup on the Pi will get you your initial setup comms (comes in handy to check it all works).
Also comes in useful to find once you get close to it and a) the WiFi hasn't detached itself, smashed or the battery hasn't had a jerk to disconnect it.

Loggin is pretty easy, but I guess that depends on how fast you wish to log at.
You log in a thread and then batch write out to SD at regular intervals as needed (in a thread)
you can run the camera in another thread, if video recording this is SD card intensive and locked up out A+ iirc, but still captures is possible. (you have a Zero so twice the memory we had to play with on the A+'s)
I also had another for GPS, but as they are 1/second I didn't log with the other data and ran a separate python logger for that alone. It is integrated into my logger, but it's never been used in the rocket (on my bike in a back pack it has)


Everything was easy on a Pi2.
SenseHAT datalogger by RPI Education section was the basis for most of my logging, stored in CSV files and jpg's
The tutorial page seems to have changed a bit but is probably the same underneath

(that Berry has the same IMU as the SenseHAT and the BMP180/280 pressure/temp is as seen on the EnviroPHAT, just an F.Y.I.)

I haven't done the streaming (other than video over an Android App at the start of launch)
I have played with visualisation but not direct from the rocket.
http://blog.bitify.co.uk/2013/11/3d-ope ... -from.html (they happen to be the individual modules I have, which was lucky for playing around with, so nicked them of the Arduino they where on )


Hope that's of some use for links to anyone reading
(I'm going to read that Berry link )

[bensimmo]

knotty, if you could do with the width, consider dumping the uSD power connector and connect to the GPIO 5V and GND, or solder your own snap fit power connector to the PP pins to make it more compact.

[knotty]

Is that a Zero W?
You want to overlay data onto live video?
For rockets you need to collect data fast so probably save into a memory buffer.
Don't try to save to SD card, until you are going down or landed.

I would collect video and data separately and do the overlay in post processing back on the ground.
You can overlay some stuff one line or so?
Pikrellcam uses this feature to timestamp videos.

Can it be done fast enough in Python?
A good place to start would be with Picamera software

http://picamera.readthedocs.io/en/relea ... index.html

Yes using a Zero W. I’m not planning on overlaying data onto LIVE video but onto recorded video onto the SD card. I’ve thought about post processing, and will go down that route if I have to, but currently I’ve got Picamera working in a basic form; recording 30 seconds of video with time and GPS overlay. I've attached a photo...

Here’s the code…

root@raspberrypi:~# cat overlay_gps_working.py
#!/usr/bin/python
import serial
ser = serial.Serial('/dev/serial0',9600,timeout = None)
import picamera
import datetime as dt

with picamera.PiCamera() as camera:
camera.resolution = (1280, 720)
camera.framerate = 24
camera.start_preview()
timelatlong = ""
camera.start_recording('timestamped.h264')
start = dt.datetime.now()
while (dt.datetime.now() - start).seconds < 30:
gps = ser.readline()
if gps[1 : 6] == "GPGGA":
timelatlong = gps[7:9] + ":" + gps[9:11] + ":" + gps[11:13]
timelatlong = timelatlong + " " + gps[18:20] + "." + gps[20:22] + "." + gps[23:27] + gps[28:29]
timelatlong = timelatlong + " " + gps[30:33] + "." + gps[33:35] + "." + gps[36:40] + gps[41:42]
camera.annotate_text = timelatlong
camera.wait_recording(0.2)
camera.stop_recording()
root@raspberrypi:~#

What I’d like to do is also add speed, course and altitude from the NMEA sentences but I’m not great at Python!

We have a larger payload to play with (are you parachuting it on it's own or is that to stay in the rocket ?)

WiFi and AP setup on the Pi will get you your initial setup comms (comes in handy to check it all works).

I have played with visualisation but not direct from the rocket.

http://blog.bitify.co.uk/2013/11/3d-ope ... -from.html

When I’m in the US in a week or so I plan on buying some more tubes/connectors. There are only a few suppliers in the UK and they are quite expensive. I’ve got the Pi setup as a WiFi AP and also a simple web server so I can pull recorded videos down to my MacBook. That visualisation looks cool and I’ll add it to my future list!

knotty, if you could do with the width, consider dumping the uSD power connector and connect to the GPIO 5V and GND, or solder your own snap fit power connector to the PP pins to make it more compact.

Yeah I’ve looked into this. Apparently if you use the GPIO there is no fuse protection so I’d rather avoid this. I did consider removing the Micro USB connector and soldering directly but apparently the pads aren’t strong enough. Any experience with this?

[bensimmo]

Switch to PYNMEA2 module for your GPS and your python can relax a bit.
https://github.com/Knio/pynmea2/blob/master/README.md

Just saved me from having to recreate code.


A bit of code I can find from a few years back (note Pi3 just stating era)

Hope it makes sense,

Code: Select all

##Based around https://www.raspberrypi.org/learning/sense-hat-data-logger/worksheet2/
##need to install pynmea2?

##### Libraries #####
import serial	#to read fro serial UART
import pynmea2	#NMEA GPS splitter

from datetime import datetime
from time import sleep
from threading import Thread



##### Logging Settings #####
FILENAME=""
WRITE_FREQUENCY = 10
TIMESTAMP = True
LATITUDE = True
LONGITUDE = True
ALTITUDE = True
NUMSATS = True
DELAY=0



##### Functions #####
def file_setup(filename):
    header =[]
    if TIMESTAMP:
        header.append("time")
    if LATITUDE:
        header.extend(["latitude","latdir"])
    if LONGITUDE:
        header.extend(["longitude","longdir"])
    if ALTITUDE:
        header.append("altitude")
    if NUMSATS:
        header.append("Number Sats in use")

	#creates (w write mode aka overwrites) file with headers, comma seperated
    with open(filename,"w") as f:
        f.write(",".join(str(value) for value in header)+ "\n")



def log_data():
    output_string = ",".join(str(value) for value in gps_data)
    batch_data.append(output_string)


def timed_log():
    while True:
        log_date()
        sleep(DELAY)

        

def get_gps_data(strip):
    gps_data=[]

	#looks at GPS lines for GGA tag, parses data
    if strip.find('GGA') > 0:
        msg = pynmea2.parse(strip)
#	print("Timestamp: %s -- Lat: %s %s -- Lon: %s %s -- Altitude: %s %s") % (msg.timestamp,msg.lat,msg.lat_dir,msg.lon,msg.lon_dir,msg.altitude,msg.altitude_units)

        if TIMESTAMP:
            gps_data.append(msg.timestamp)
        if LATITUDE:
            gps_data.extend([msg.lat,msg.lat_dir])
        if LONGITUDE:
            gps_data.extend([msg.lon,msg.lon_dir])
        if ALTITUDE:
            gps_data.append(msg.altitude)
        if NUMSATS:
            gps_data.append(msg.num_sats)

        gps_data.append(datetime.now())

        return gps_data 

        #append file comma seperated with selected data, shows on screen it's written
        #  with open(filename,"a") as f:
        #      f.write(",".join(str(value) for value in gps_data)+ "\n")
        #    print("Writing to file..")




##### Main Program #####
serialPort = serial.Serial("/dev/serial0", 9600, timeout=0.5) #opens serial port, may need to change ttyAMA0 to whatever is now used e.g. serial0
batch_data= []

#set filename based on generic or given filename and current date-time
if FILENAME == "":
    filename = "GPSLog-"+str(datetime.now())+".csv"
else:
    filename = FILENAME+"-"+str(datetime.now())+".csv"

#setup file and add headers
file_setup(filename)

#see if we are using a delay, thread out a log if so
if DELAY > 0:
    strip = serialPort.readline()
    gps_data = get_gps_data(strip)
    Thread(target= timed_log).start()

while True:
    strip = serialPort.readline()
    gps_data = get_gps_data(strip)
    
    if DELAY == 0:   #only use if no delay
        log_data()

    if len(batch_data) >= WRITE_FREQUENCY:
        print("Writing to file..")
        with open(filename,"a") as f:
            for line in batch_data:
                f.write(line + "\n")
            batch_data = []

I think you can do the others you want, maybe not course but the nmea document is also there so you parse that line and cut it out if you keep your method.

Also remember/figure out which GPS long/lat reporting style you wish to display. (Could be worth adding the units).

[bensimmo]

On the Zero there is no protection anyway even through uUSB. (for you the cells and voltage module will be you protection).
So any connection will be the same, be it the PP pads that connect to the usb input or the gpio pins.

[Gavinmc42]

I have powered lot of Pi's via the 5V on the header, it is fine if you know what you are doing.
I have only blown up one, 9V on a GPIO

Zero's should work with 3.6-4.2V on the 5V as the BCM2835 circuit schematic shows, the 5V goes to the Vbat on the chip.
ie the BCM2835 was designed to run from a battery, 5V is needed for USB and HDMI.

Nice to see the GPS on overlay works, might have to borrow that bit of code
I think I read somewhere it can only be one line?

I think the VC4 can use mmal layers, video on a different layer.
It is how the videocore example video on the teapot demo works.
But this would mean doing camera with mmal and not the usual raspivid software?

Raspistill and Raspivid were originally just meant to be demo programs.
One day I may get around to reading, the currently 178 pages of this
viewtopic.php?f=43&t=63276

I guess it could be used to capture flight videos, motion detect launch, save on motion detect stopped after landing.
Uses MJPEG which should be better than h.264 if you lose power during writing to SD?
I have not used this RPi-Cam-Web software as most of my camera stuff till now has been for security or multi/hyperspectral use.
Uses Raspimjpeg? That could be useful for drone FPV, RC car, robot etc use.

[IotBitOfficial]

Have You had a look at our products? We provide HAT's for Raspberry Pi's which give them mobile communication features such as 4G 3G and GSM. We also provide a Gas sensing Hat which has the capability to measure temperature and humidity. check it out @ altitude.tech

[westcoastRC]

IMG_1702.jpg (163.63 KiB) Viewed 12445 times

I too am working on a project much like yours, but without the camera system on board. The purpose of my project is to track altitude, speed and GNSS/GPS data via LoRa. The only issue is the Raspberry Py 3+ base station which I am trying to find a code/operating software that will display this information real time.

[bensimmo]

You might want to have a read of this.
pytrack & skygate HAB software.
https://www.raspberrypi.org/blog/pytrac ... -software/

[westcoastRC]

I have seen Skygate - but the software is written specific to particular RF modules and frequencies. MikroE LoRa2 click uses very specific codes to get it operating and is not compatible with the current software. Tried it already - great interface!