SDR cubesat question

Go To Last Post
17 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hello all,

I've been tasked with developing (or at least investigating) the possibility of a software defined radio for my university's cubesat program. I'm using a PlutoSDR to experiment with different data modes in GNUradio. I've been able to do quite a bit, and love this device, but after looking at the datasheet, I'm not sure if it would be a good device for the project due to power consumption. All of the AD936x transceivers have rather high power consumption. (1/4+ watt receive). This may not sound like a lot, but it is in space. Power is as valuable as gold and all that heat needs to go somewhere. 
It needs to be able to transmit 9600b ax.25 packets and possibly basic image transmission (faster than SSTV but slower than ATV, we haven't selected a mode yet). Also, it needs to be able to receive over-the-air updates in some format for other systems (likely higher baud ax.25).

An SDR is not required for the next cubesat launch, but I'm just curious, knowing the advantages, is there a good solution for a low cost low(ish) power SDR? Should I stick to an embedded microcontroller and off the shelf fsk transceiver? 

The wisdom the fellow freaks have on here never ceases to amaze me and has helped me with many projects. 

Thank you and 73, N2EEE.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What a cool product, and an equally interesting project.

 

I looked at a number of links and still didn't see the RF power output vs frequency data?

 

Are you sticking to Ham frequencies?

 

Question:  I see the software supports OSX, Win, Linux.

Is that just to configure it, then it is stand-alone?

Are you flying an RPi with it to program it in real time, while in flight?

 

It seems to me that one of the benefits of an SDR is its flexibility to tune both the Tx and the Rx.

 

It is not clear to me why that would be beneficial in your spacecraft project?

I would have thought that you would pick a frequency based upon hardware, licensing, and antenna configurations and then used that for your primary in-flight comm's, and add a second, low power, omni-directional VHF Tx with GPS coordinates for any satellite you wish to actually recover.

 

I thought, (with zero personal low earth orbit project experience...), that the heat generated is not a problem.

The heat generated is hopefully contained within the insulated spacecraft and benefits to keep your batteries warm, and the electronics and cameras, and their lenses, within a reasonable operating temperature.

The downside of the heat generation, obviously, is that that is energy "lost" and not available for the micros, cameras, GPS's, RF links, data storage modules, etc.

 

I trust you have access to the last year or two of Nuts and Volts magazine and read their LEO project columnist in addition to the other literature out on the web.

 

JC

 

Edit: KD8HKD

Last Edited: Sun. Oct 21, 2018 - 03:34 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

A quick survey of launch costs for a cubesat gives rough estimates of USD 50.000 to USD 100.000.

 

So if  you're nit picking about the cost of a SDR module you are really working from the wrong end of the specturm if you ask me.

Haven't looked much into cubesats, but it seems these often get heavily sponsored, but in my book that is not a good excuse to send the cheapest cobbled together junk into space you can get away with.

 

Just curious:

Are there build plans circulating in the cubesat world for foldable solar panels?

If it's not (and if it's allowed) then such a design could benefit a lot of other cubesat projects.

A nice folding mechanism would be with a spring and a wire around a (metal film) resistor, which melts the wire at the right time (G sensor / vacuum / light intensity + timer?)

Paul van der Hoeven.
Bunch of old projects with AVR's:
http://www.hoevendesign.com

Last Edited: Sun. Oct 21, 2018 - 09:12 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

DocJC wrote:

I looked at a number of links and still didn't see the RF power output vs frequency data?

There wasn't a graph, but there was data on the current consumption @ 1.3v Receiving one channel at 800 MHz. We'll probably end up using 70cm or 23cm, but the ISM band is not out of the question. 

 

DocJC wrote:
Question:  I see the software supports OSX, Win, Linux. Is that just to configure it, then it is stand-alone? Are you flying an RPi with it to program it in real time, while in flight?

We will use linux for the ground station (likely mint, nothing too crazy), and an SOC with embedded linux for the cubesat if it needs it.

 

Paulvdh wrote:
A quick survey of launch costs for a cubesat gives rough estimates of USD 50.000 to USD 100.000.

That's why one of our next satellite projects is a low-cost satellite. I'm not sure how low cost but it certainly helps to have price in mind.

 

Paulvdh wrote:
Are there build plans circulating in the cubesat world for foldable solar panels?

I believe it won't have solar panels that fold out. The previous launch didn't. I hope it does, though. As you can tell, it's pretty early in the project so nothing is set in stone. 

 

I have a design review of the PlutoSDR completed and ready for presentation, but just wanted to get any additional information or ideas for it from more experienced people out there. 
Thanks and 73, 
N2EEE

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

N2EEE wrote:
All of the AD936x transceivers have rather high power consumption. (1/4+ watt receive).
Similar for SDRPlay RSP2 at 850mW (5V * 170mA, USB VBUS), 1.5MHz to 2GHz, or, 1KHz to 30MHz

RSP2's ADC is up to nearly 11MSPS with 10.4 ENOB (12b ADC) at 8MSPS; that's PIC32MX PIC32MZ capability with several sequenced 12b ADC with DMA but its current is significant (100mA roughly)

Can AVR do SDR?  Yes

XMEGA128A1U, 12b ADC, 2MSPS, two ADC with DMA, EBI for RAM if required

XMEGA256A3U - likewise though without EBI, 16KB internal RAM

Popular cubesat MCU are PIC24, dsPIC, and MSP430; didn't search for PIC24 or dsPIC (iow didn't make a run through MAPS)

 

https://sdr-kits.net/SDRplay-RSP2-RSP2pro-10kHz-to-2000Mhz-Wideband-SDR-Receiver

http://sdr-kits.net/documents/RSP2_Specification.pdf (page 2 of 2, last half)

https://www.microchip.com/maps/

 

Edit: strikethru

 

"Dare to be naïve." - Buckminster Fuller

Last Edited: Sun. Oct 21, 2018 - 03:44 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

DocJC wrote:
It seems to me that one of the benefits of an SDR is its flexibility to tune both the Tx and the Rx.

 

It is not clear to me why that would be beneficial in your spacecraft project?

I would have thought that you would pick a frequency based upon hardware, licensing, and antenna configurations and then used that for your primary in-flight comm's, and add a second, low power, omni-directional VHF Tx with GPS coordinates for any satellite you wish to actually recover.

Two or more radios are common for spacecraft and large passenger aircraft (low gain antenna for telemetry, high gain antenna for data and position, low gain antenna is the alternate antenna)

One SDR could do both (low ISM for TM, high ISM for data) reducing mass and power.

 

https://en.wikipedia.org/wiki/ISM_band#Frequency_allocation

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

N2EEE wrote:
... and an SOC with embedded linux for the cubesat if it needs it.
SAMA5D2 SoM is 3.3V up to 450mA so maybe 600mW effective.

There are arm Cortex-M SoM running uClinux that might meet the power requirement some of which have a fast enough ADC.

Will need a watchdog for the system controller to handle the radiation (Galactic Cosmic Rays (GCR)) and the single-event upsets (SEU) due to GCR and high energy electrons (South Atlantic Anomaly)

Microchip's Microsemi has radiation-tolerant FPGA that can run RISC-V or has an arm Cortex-M1.

N2EEE wrote:
I'm not sure how low cost but it certainly helps to have price in mind.
Price of launch is (was?) huge.

Bring Stratolaunch and etc.

N2EEE wrote:
I believe it won't have solar panels that fold out.
Most Cubesat tumble though, IIRC, there are some that have attitude control.

 

P.S.

Microchip has radiation-tolerant megaAVR one of which is megaS128 (has an EBI)

VORAGO has radiation-hardened arm Cortex-M0.

 


https://emcraft.com/home

https://www.microsemi.com/product-directory/fpga-soc/1640-rad-tolerant-fpgas

https://www.avrfreaks.net/forum/stratolaunch

https://www.microchip.com/wwwproducts/en/ATmegaS128

https://www.voragotech.com/vorago-products

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

...reducing mass and power...

 

Obviously some interesting design tradeoffs.

 

The transceiver mentioned has a very wide operating bandwidth.

Presumably one might be using two different bands, (VHF and UHF, for example), for different power levels, data rates, etc.

But as soon as one does that with a single transceiver one now needs an RF relay, more connectors, (RF signal loss), and power to drive the relay, or a very very carefully laid out RF PCB with some discrete transistors to route the signals, (non-trivial).

The SDR mentioned also appears to need a Linux PC attached to it to change the configuration, so more weight and power and 100% bug free software just to change channels.

 

Two "credit card" ham HTs would seem to be much lighter, likely more power efficient (no PC or RF relay), have switchable RF power output levels, and be easily connected to two antennas.

 

Lots of ways to approach this task!

 

JC

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

N2EEE wrote:
I have a design review of the PlutoSDR completed and ready for presentation, ...

  • The Linux kernel is about one megabyte (plus or minus)

1MB into any orbit is possible though not necessary for SDR.

Consider an RTOS or event framework; the reduction in memory usage is an order or two of magnitude.

 

  • Bulk CMOS survival duration at LEO and VLEO versus solar maximum and solar minimum versus Cubesat flight duration (before natural de-orbit) at solar max and min

32b MCU can be made radiation tolerant; likewise with FPGA (RISC-V)

The transistors in 8b and 16b MCU are larger in area with thicker gate oxide; so, will survive some GCR and high energy electrons.

A 16b MCU might be a best fit for SDR with some SDR functions possibly in a FPGA.

Some PIC24 have a 1MB EBI (8b widith, it's 16b for PIC32)

Maybe an off-the-shelf radiation-hardened board with 32b MCU and "lots" of RAM :

https://www.voragotech.com/products/seb2-%E2%80%93-pa32kas-development-board

CubeSat board though the bulk RAM is SPI-attached :

https://www.voragotech.com/products/rhobc1

https://www.mouser.com/new/vorago-technologies/vorago-rh-obc-1-board/

 

Edit: Mouser

 

"Dare to be naïve." - Buckminster Fuller

Last Edited: Mon. Oct 22, 2018 - 08:07 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thank you for all the replies. If I end up developing the transceiver, I'll most likely use a ATMEGAS128 or similar rad-hardened MCU because of familiarity. If not, then I'll have an MSP430 in the design, since that's what's taught in the curriculum. I'll probably pair this with an SI4463-B1B-FM or similar transceiver chip. The last launch used 70cm with a kill switch on 2m. We'll most likely use 70cm/2m again, but we may add 23cm. 

gchapman wrote:
The transistors in 8b and 16b MCU are larger in area with thicker gate oxide; so, will survive some GCR and high energy electrons.

Huh, awesome. Never knew that. Thanks!

DocJC wrote:
The SDR mentioned also appears to need a Linux PC attached to it to change the configuration

It is possible to use it standalone if the programs are stored on the flash. 
 

DocJC wrote:
Two "credit card" ham HTs would seem to be much lighter, likely more power efficient (no PC or RF relay), have switchable RF power output levels, and be easily connected to two antennas.

I can almost guarantee it'll require pcb design, but we were analyzing if taking the pluto, adding a PC-104 cubesat bus, and building the transceiver only in the software was a good idea. The more I look into it, the less favorable using an SDR seems to be. I looked into this "credit card" ham HT Alinco DJ-C(1-7) and it seems pretty cool. Might be able to mod for digital transmission

 

gchapman wrote:
Price of launch is (was?) huge.

Yep, the cost of development was huge. That's why we're working on a low cost satellite for future usage. I don't know if the launch costs filter into this cost, though. 

 

gchapman wrote:
CubeSat board though the bulk RAM is SPI-attached : https://www.voragotech.com/produ...

That is quite an angry price, but thank you for the link. Saved for later.

Thank you for your ideas! Design review is in 2 days and I'll keep you updated on whether we'll pursue this SDR transceiver. 

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

N2EEE wrote:
If not, then I'll have an MSP430 in the design, ...
A concern is a 16-bit MCU might not have enough compute power for SDR.

Two of the several ways to overcome that are

  • Dual processor (dual MSP430)
  • Co-processor (FPGA)

If a HDL is too much to bite off, a soft processor can be in the FPGA (RISC-V, MicroBlaze, Nios II, AVR mega103, etc) to run an application by your preferred computer language.

N2EEE wrote:
That is quite an angry price, but thank you for the link.
You're welcome.

The following is one way to reduce the price :

Embedded

The challenges and evolution of CubeSat electronics

by Ross Bannatyne, VORAGO Technologies (marketing director)

January 23, 2017

https://www.embedded.com/electronics-blogs/say-what-/4443299/The-challenges-and-evolution-of-CubeSat-electronics

...

Figure 3. Using a rad-hard MCU with a COTS FPGA to optimize cost and mitigate radiation effects (Source: VORAGO Technologies)

...

Ross's article describes how CMOS can be destroyed by latch-up (for too long) due to GCR and high energy electrons and how CMOS is upset (SEU) by likewise.

Mouser has in-stock the VORAGO radiation-hardened arm Cortex-M0.

A radiation-hardened MCU can implement the system controller (watchdog, power control to quickly overcome latch-up, current sensing and control, battery charging, antenna switching, etc)

A radiation-tolerant MCU can do likewise except for the watchdog (megaS128, megaS64M1)

Price can be further reduced by use of typical CubeSat MCU as these seem more than adequate for reliability.

There are radiation-tolerant FPGA (Microchip's Microsemi) but an industrial FPGA apparently will be reliable enough for LEO; might inquire at your preferred FPGA manufacturer about what would be recommended for VLEO or LEO (an older model, current model with redundancies and EDC and ECC, ...)

 

P.S.

N2EEE wrote:
[MSP430] since that's what's taught in the curriculum.
A good choice that parallels FSF's choice of a 16-bit MCU for GCC.

MSP430 lacks an EBI; PIC24 and dsPIC have either EPMP or PMP for either EBI or a synchronous interface (display, camera, etc) (8 bits wide by up to 16 address bits)

MSP430, PIC24, and dsPIC are in CubeSats.

A dsPIC with a "lot" of internal RAM and a PMP :

https://www.microchip.com/wwwproducts/en/dsPIC33EP512GP806

via https://www.microchip.com/maps/Microcontroller.aspx (Family : 16-bit DSC, SRAM 53248 bytes, Digital : Parallel Port PMP)

 

P.P.S.

Ross's article shows F-RAM.

MRAM is an alternative for some satellites; MRAM has internal ECC.

Parallel-bus MRAM has significant idle current.

https://www.everspin.com/aerospace

 


https://www.embedded.com/user/Bannatyne

https://www.mouser.com/vorago-technologies/Semiconductors/Embedded-Processors-Controllers/_/N-6hpef?P=1ye0wuz

https://www.microchip.com/design-centers/aerospace-and-defense/radiation-tolerant

https://www.microsemi.com/product-directory/fpga-soc/1640-rad-tolerant-fpgas

 

Edit: 1st URL

 

"Dare to be naïve." - Buckminster Fuller

Last Edited: Wed. Oct 24, 2018 - 01:23 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

A dsPIC with a "lot" of internal RAM and a PMP :

[dsPIC33EP512GP806]

The improved dsPIC was announced this summer.

The following is a one core that's faster, improved ADC (more speed and one more), and PMP though less internal RAM (can add more on PMP) :

https://www.microchip.com/wwwproducts/en/dsPIC33CK256MP208

dsPIC33C has a two core version though it apparently doesn't have PMP.

 

Announcements :

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

N2EEE wrote:
The more I look into it, the less favorable using an SDR seems to be.
Can a dsPIC be effective at 40MHz ISM (Earth satellite)?  Yes

Can a dsPIC be effective at 434MHz EMEA ISM?  Am uncertain; might be able to answer given data from the datasheets or consider a bench test (proof-of-concept)

Can a FPGA be effective at 434MHz EMEA ISM?  Yes

 

SDR Cube Transceiver

at Midnight Design Solutions (Maryland, designers in Maryland and Finland)

https://www.sdr-cube.com/products.html

[dsPIC33F, hardware audio codec for ADC and DAC over I2S, up to 17m band (18MHz)]

Microchip logo

Microchip

dsPIC33CH Curiosity Development Board

https://www.microchip.com/Developmenttools/ProductDetails/DM330028

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Can AVR do SDR?  Yes

A previous version of SDR Cube was on an 8051 at 67MHz.

SDR Cube

A Portable Software Defined Radio Utilizing An Embedded DSP Engine for Quadrature Sampling Transceivers By George L. Heron N2APB and Juha Niinikoski, OH2NLT

https://www.tapr.org/pdf/DCC2010-SDRcube-N2APB-OH2NLT.pdf

(page 24, third paragraph)

The rest is all digital processing. The [Cypress] PSoC 3 family of parts have an interesting internal hardware feature they call a digital filter block, or DFB, and it consists of a 24-bit fixed point, programmable limited scope DSP engine. This is a dedicated hardware accelerator block that operates independently of the main 8051 processor [at 67MHz]. It consists of a dedicated multiplier and accumulator that calculates a 24-bit by 24-bit multiply 48-bit accumulates in one system clock cycle. It is optimized to implement a direct form Finite Impulse Response (FIR) filter that approaches a computation rate of one FIR tap for each clock cycle. This block is used as two independent, 64 tap, digital filters.

(page 25, bottom)

Performance:

The minimum discernible signal (MDS) is approximately -117 dBm. A typical receiver would normally be more towards the -120 dBm number. However, on the lower HF frequencies, where atmospheric and man-made noise dominates, the performance is sufficient. In fact, in a head to head comparison with my Flex Radio Flex-3000, I find it difficult to hear any difference. Not bad for a $50 receiver.

http://www.cypress.com/products/psoc-3

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ADALM-PLUTO is a value when compared with SDR-Kits SDRPlay.

https://www.ebay.com/sch/m.html?_nkw=&_armrs=1&_ipg=&_from=&_ssn=sdrkits&_sop=15 (sdrkits store on eBay)

 

ADALM-PLUTO vs SDRPlay :

  • double the bandwidth
  • transceiver
  • slightly less noise

 

SDRPlay's frequency is two orders of magnitude more accurate than PLUTO.

SDRPlay can reach to and beyond the long wave time signals (WWVB, DCF77, etc)

 

http://sdr-kits.net/documents/RSP2_Specification.pdf

via SDRplay RSP2 and RSP2pro 10 kHz - 2000 MHz Wideband SDR Receiver

 

"Dare to be naïve." - Buckminster Fuller