Cygnus Module development

Project to send a balloon into lower space with payload
Daniel Wee
Site Admin
Posts: 2070
Joined: Wed 25 Feb 25 2009 8:00 pm

Cygnus Module development

Post by Daniel Wee » Mon 06 Dec 06 2010 10:27 pm

The Cygnus module will be one of the two function-duplicate systems (the other being Mark's Scorpion module) making up the payload. These systems will have mostly independent functions, sensor and communication systems. This will list some of the key features planned for Cygnus:-

1. Bi-directional data communications on 434MHz
2. Uses the LS20031 MTK GPS module
3. I2C based sensor bus
4. Logging of data on-board (EEPROM or SD card or EasyUSB)
5. Switched primary power supply
6. Temperature control possibility via PWM MOSFET
7. Dual peripheral bus

Daniel

Daniel Wee
Site Admin
Posts: 2070
Joined: Wed 25 Feb 25 2009 8:00 pm

Sensor list

Post by Daniel Wee » Mon 06 Dec 06 2010 11:12 pm

The list of sensors are as follows:-

1. External temperature sensor (IR sensor, MLX90614 I2C)
2. Internal temperature sensor (DS18B20, TMP100/101)
3. Board level temperature sensor (DS18B20 One-Wire)
4. GPS module (LS20031 MTK)
5. Primary battery voltage sensor
6. Secondary supply voltage sensor
7. Current draw sensor (INA169)
8. UV sensor
9. 3-axis accelerometric sensor (BMA180)
10. 3-axis gyro sensor
11. Barometric pressure sensor (BMP085 - 9km limit)
12. Low-range barometric pressure sensor (?)

Daniel
Attachments
Maxim-DS18B20.pdf
(218.99 KiB) Downloaded 380 times

Daniel Wee
Site Admin
Posts: 2070
Joined: Wed 25 Feb 25 2009 8:00 pm

Re: Cygnus Module development

Post by Daniel Wee » Tue 07 Dec 07 2010 12:58 am

Cygnus will have an X-Mega A4 managing communications, and a dsPIC33F doing data logging. The current plan is to use a 1W 434MHz module as the main communications channel. This is an SPI local interface so it will need an enable line.

Daniel

Daniel Wee
Site Admin
Posts: 2070
Joined: Wed 25 Feb 25 2009 8:00 pm

Re: Cygnus Module development

Post by Daniel Wee » Thu 09 Dec 09 2010 1:39 am

In the worst case scenario, we can expect the balloon to rise to about 40km altitude and 20km lateral translation, which yields the highest distance of about 45km LOS. We know from existing tests that 500mW on 433MHz can get us out to 30km with correctly polarized antennas. Perhaps a turnstile ground station could be sufficient or a cross-polarized Yagi. We can also increase our link budget by lowering the data rate further and increasing the output power (at the expense of battery life.) From 30km to 45km would require less than 6dB of additional gain, which could easily be provided by the high-gain antenna. This is assuring.

As for video, Thomas has done a test with high gain antennas (Yagi) on the receive end and a 1W VTX on 2.4GHz, across a span of about 50km. The image wasn't great but it was possible.

The crucial blocks would be the logging, and subsequent position reporting. Consideration needs to be given to the optimal antenna positioning as the balloon changes altitude from ground to apogee.

As far as telemetry data goes, I think only the GPS coordinates, altitude, temperature(s), battery status, maybe RSSI, need to be downlinked in real time. This information would be sent at about a 0.5Hz rate which is more than sufficient. This can also serve to conserve power.

Logging may have to take place at a higher rate but would be subject to the write-rate. Accelerometer data only needs to be stored as a compound vector. Gyro - actually this may not be needed at all.

Daniel

Daniel Wee
Site Admin
Posts: 2070
Joined: Wed 25 Feb 25 2009 8:00 pm

Re: Cygnus Module development

Post by Daniel Wee » Thu 09 Dec 09 2010 7:24 pm

The GPS itself can sit under the foam shields and should operate without issue. This will allow it some heat shielding but it would be crucial to test the operational temperatures for this module.

Power wise, I'm trying to see how little power I can get away with for the entire system. The real pinch is the 9V needed for 1W output power.

Daniel

Daniel Wee
Site Admin
Posts: 2070
Joined: Wed 25 Feb 25 2009 8:00 pm

SD card interface

Post by Daniel Wee » Fri 10 Dec 10 2010 5:57 pm

The SD card can be accessed via the SPI interface. The following link(s) are useful references:-

http://www.k9spud.com/sdcard/
http://www.k9spud.com/sdcard/schematic.php
http://www.brushelectronics.com/index.php?page=software
http://mondo-technology.com/sdcard.htm
http://www.captain.at/electronics/pic-mmc/
http://dbaspot.com/forums/arch/382345-p ... rface.html
http://elm-chan.org/fsw/ff/00index_e.html
http://www.microchip.com/forums/tm.aspx ... 0%BB%9D%83

The SD card will probably get its own dedicated SPI port, and the RF module its own as well. Hopefully this way we can avoid latencies and channel collisions.

Daniel
Attachments
Microchip-SD-MMC AN1045.pdf
(559.79 KiB) Downloaded 355 times

Daniel Wee
Site Admin
Posts: 2070
Joined: Wed 25 Feb 25 2009 8:00 pm

Re: Cygnus Module development

Post by Daniel Wee » Fri 10 Dec 10 2010 6:04 pm

At this point, we need to determine exactly how the sensors will be mounted, how many, and what type. This is for board and power supply considerations. One of the sticky points now is the 9V required by the RF module in order to output 1W. At 5V, it is going to be putting out much less than 500mW which could be a problem.

Assuming a 2500mAh 3S battery, which yields around 27.5Wh or about 19Wh at 70% efficiency. If we intend to run this over 5-hours, we are looking at no more than about 3.8W per hour consumption. At 5V, this is no more than 760mA total continuous current draw. Our primary current draw is going to come from the RF module, followed by the control board and GPS. I think we may be able to manage 760mA, especially if we are prudent with the RF duty-cycles.

Daniel

Daniel Wee
Site Admin
Posts: 2070
Joined: Wed 25 Feb 25 2009 8:00 pm

Re: Cygnus Module development

Post by Daniel Wee » Fri 10 Dec 10 2010 6:22 pm

The RF module draws a maximum of 380mA at 9V and typically 350mA. Assuming that we just went with a linear off 3S, we would end up with about 3.85W input for 1W output. Also, assuming a 64-byte packet at 1200baud, this will require half a second or so to send. If we send the data once every second, that's a 50% duty cycle, and the equivalent power drawn would be about 1.925W. This is probably a bit on the high side so we need to either increase the baud rate, or cut down the packet size.

In any case, we do need to determine now PA_VCC is going to be supplied and how many volts. We could conceivably use a dedicated switching PSU for this but this would almost certainly require an input supply of 3S or more. At this point we are looking at a 9VDC rail, a 3.3VDC rail and maybe a 5VDC rail. The 3.3V rail looks like it will require a switching supply but I will use larger inductors. This should be sufficient to handle at least 600mA. It might be possible to use a SEPIC buck/boost for the 9V rail, capable of about 600mA or more.

Daniel

Rris
Posts: 66
Joined: Mon 12 Oct 12 2009 9:17 pm

Re: Cygnus Module development

Post by Rris » Fri 10 Dec 10 2010 8:11 pm

I think 3hrs would be the max flight time. Balloon calcs reckon 2:45 or so. I'm working on a 3.5hr budget.

Daniel Wee
Site Admin
Posts: 2070
Joined: Wed 25 Feb 25 2009 8:00 pm

Re: Cygnus Module development

Post by Daniel Wee » Fri 10 Dec 10 2010 9:45 pm

Yes but we may not be able to recover it within one hour. So I'm just planning for a big margin.

Daniel

Post Reply