I haven’t really had a chance to dig into comparing the SDR and FlightAware requirements with the Raspberry Pi Pico capabilities, but I bet someone already has!!
$4 for the potential that little Pico has makes me a little tingly. Even if it required the stacking of 3 or 4 Picos to handle the load, one for SDR, one for running FlightAware, one for handling the data packet negotiations with the server, etc. etc.
Anyone? Consider or Try?
I believe it is designed to be a micro-controller and is similar to an Arduino.
The ARM Cortex M0+ cpu is about 9 years old according to this ARM Cortex-M - Wikipedia
So I’m not knowledgeable enough to understand the implications of that vs. how a Raspberry Pi 4 is designed. The maker community are going nuts with the new Pico and using our it to run some pretty complex stuff, so is it possible to hack it and make it work for whatever you’d want?
Does that mean it has no onboard RAM, or something? I’ve been running my PiAware and Nooelec Nano 2 with a Zero W for the better part of this year, so I figured the PiAware image is light enough for that, maybe this.
I appreciate the reply and schooling in advance!
The RaspberryPi 4, or the Zero for that matter, are small computers. You install an operating system, attach a mouse, keyboard, monitor, or ssh into it. You can ran software like Piaware, or use it like a desktop (if you are inclined to be of a patient nature).
The Pico is a controller. You’ll have to code something in eg python, uploaded via the usb port, and then it runs this code and interacts with devices that are attached to its gpios.
It would be very cool if it is possible, but unfortunately I don’t think the Pico will ever be able to manage it.
It has very little RAM - 264KB to be precise. Plus it does not have any WiFi or ethernet capabilities. It is very similar to an Arduino; it is a microcontroller instead of any other Pi which is a computer.
It doesn’t have any USB ports, so even if it met FA requirements, a special GPIO dongle would have to be created.
It would be terrific if it was possible, but it’s just like replacing a Pi with an Arduino.
In the future, Raspberry Pi might consider making a more advanced version of the Pico, it would still be a microcontroller but could feature USB, WiFi, etc.
The ESP8266 is a microcontroller has wifi, and I think there are Arduino ethernet boards. So it is completely possible to add these to a Pico. But only time will tell, but at present, it wouldn’t be possible at all.
Edit: I forgot to mention that it’s programmed with MicroPython. It doesn’t run an OS so everything would have to be run through several Python scripts.
It might be that somebody makes a small dongle, similar to the existing ones, that does more work itself then passes the data onto the Pico, which would send it to FA. So basically it would just take the data from the dongle and pass it to relevant service (FA, FR24 etc…)
Ooof. Could you go ahead and do that for me. My expertise on Python begins and ends with the term ‘sudo’, which I have no idea the meaning of.
I’ve always had a tenuous relationship, at best, based on my frustrations with computers that stopped doing what they were supposed to be doing because nothing. This would usually come to a head whenever I was consulting on a specific software development or configuration where we were trying to build a very specific tool to do a very specific thing. Whenever they would tell me that one of the stated requirements for the software was not possible or only possible at a level which would address just a fraction of what we wanted, I’d drop my philosophy of all things computers:
“You guys work in a universe made up of just two things, 1’s and 0’s. Why can’t you just find out which of those 1’s and 0’s are needed and what order they need to be in to make the stupid thing do what we need it to?”
Please, don’t feel the need to answer that question, I am aware of the ludacris nature of it, and have been schooled on at least 2 dozen of the however many legitimate reasons exist that answer the question.
Full disclosure, it was really just a manifestation of my frustrations that we wouldn’t be getting tools as powerful and capable as we wanted. Who amongst us can say that they don’t get a little passive aggressive (PAGRO) from time to time?
My reasonably high expectations actually been be explained beyond my just being ignorant (which is a contributor!). I’ve been a civil aviation industry careerist for more than 30 years, and includes time as a commerical pilot, aircraft mechanic, and spent a LOT of time both engaging with or an employee of the FAA. With respect to the vast majority of work I was involved with, if any one of those components integrated into the aircraft or FAA air traffic management systems randomly failed or blue-screen-of -death at a rate we see with computers and digitally controlled equipment seem to fail at, the FAA would have kittens and they would be immediately pulled from service. So I’m essentially brainwashed into believing that any failures are unexceptionable. So hopefully explains WHY I am so unreasonable.
Oh well, thanks for the engagement and knowledge! I’ll talk to you again once the next good idea pops into my head!! 
Now THAT I didn’t even think about, talking about the last thought you had. That skinny, no processing and just push the raw data, actually meets the requirements of our project.
The primary requirement is to reduce the nominal operation power draw to the greatest degree possible, so where I have a habit of building or purchasing tools which must would consider inefficient overkill, the ideal controller solution will have no unnecessary parasite voltage draw from unused features or background OS scripts running for some unneeded ‘pondering the meaning of life’ type processes.
I think the Zero W is going to end up the best solution, but I saw that Pico, and some of the killer kit being built with it by a bunch of 8 to 11 year olds who are way smarter than I ever will be.
The biggest obstacle to my maker skills is that whatever part of the brain which handles math, languages, abstract processing, and as a result, coding, either was deprived of O² for too many minutes at some point, or is so caked with malted hops buildup and Popeyes Chicken fat, it’s effectively dead and doesn’t work.
Every once in awhile I get a wild optimistic hair up my bum, and try to learn some foriegn language (Farsi), coding (Java Groovy cause it has a neato name and I’m a home automation dork), or math (algorithmic concepts). I can do the vocab and memorization parts without much issue, but the moment some ‘not exactly like English’ lexicon is introduced, the dumpster gets filled with old tires, soaked in gasoline, and promptly ignited, and a guarenteed spectacular failure soon after.
The only 2 reasons that any of my Pi projects work or my home automation integrations don’t cause the house to burn down is 1. Github and 2. Copy and Paste. I guess giving credit to the 8 year olds I liberate all that code from deserve a some credit too, so 3 reasons.
The Zero is an excellent solution, I have two running PiAware. Takes around 450mA running itself and an ADS-B tuner and easy able to cope with the amount of air traffic I see. Live around 14 miles from Heathrow so pre COVID used to see between 1500-2000 aircraft daily.
The 1s and 0s are really on and off switches, 1=voltage, 0=No Voltage. Amazing that millions of on off switches create our images, the wonder of technology.
Geoff
I actually am fascinatrd by the things we can do with those crazy microscopic diodes, you’re spot on with that.
Good to know on the amperage draw figure. What SDR does that represent, and is that figure derived or measured?
Apologies I had to scrap my previous post there were too many errors. Didn’t get much sleep last night
Let me try again:
Do you pick up aircraft taxiing/on ground at Heathrow?
I am about 30 mi away and see about 800 per day with Pro Plus stick and FA antenna.
Compared to yours, my lower amount would make sense as I am only seeing most aircraft >1000ft.
How many do see now during lockdown?
The figure was measured using a USB volt/amp meter which shows the draw of connected devices.
The actual device was bought from a UK Raspberry Pi seller but they seem to have discontinued it and now sell the FlightAware device.
I’ve been very pleased with my ADS-B tuner, I have two and I am sure the same or similar are available via online outlets.
Details follow as to my device;
Device: ID 0bda:2838 Realtek Semiconductor Corp. RTL2838 DVB-T
Device Descriptor:
idVendor 0x0bda Realtek Semiconductor Corp.
idProduct 0x2838 RTL2838 DVB-T
(Bus Powered)
MaxPower 500mA
The above is an edited section of the output details from the computer of the device.
Geoff
No, I dont get them taxiing at Heathrow, lose them really below about 5000 feet. Get decent range with my standard non optimised antenna when aircraft above 8000 feet but as they drop below that then likely to lose contact.
My location is on a Heathrow flightpath, approaching and departing and on either they are around 8000 feet. Bit noisier when climbing than on decent but at 8000 feet certainly not an issue.
I think, due to sound issues, Heathrow alter the flight path subtly every so often but guess if you live in Hounslow you’ll get it all the time.
Geoff
I’ve just ordered a few of Nooelec’s new Nano 3 which will be here soon. They are tiny, designed so that you can fit one in every one of the Raspberry Pi USB ports at the same time. I was concerned about heat management because I knew that they run red hot, but they sent me the benchmark stress test data as well as some IR images which seemed to indicate the design isn’t retaining much of the heat when installed in the open.
Once I get them, I need to see how they do when enclosed inside a case. I’m optimistic it won’t be an issue with our requirements dictating the need for just a single SDR. The one feature that they lack is an integrated bias tee toggle option if we end up using an active antenna, filter, or preamp. I’ll let you guys what I find when they get here!
That is interesting, strangely when FlightAware brought out their tuner a few years back I immediately bought one despite them being about double the price of the one I had.
Ended up sending it back as I was concerned about the heat developed whereas my current tuners are merely mildly warm. I know many devices are designed to run hot but must admit I don’t like too hot. I still have 4 original Pi devices (One Chinese and three British made) running various projects, absolutely no problem with processor heat.
Geoff
So there is a heat management option which is called ‘Thermally Conductive Potting Compound’. From what I’ve read and been told, it apparently works very well to effectively draw out any parasite heat away from the components and transfer it to the case material (metal being the best option as far as this goes) to be shed into ambient air.
I haven’t used it yet, but will be able to determine if we need to use it for all or some applications once we see how the test articles subjected to the stress testing regimen tolerate the simulated austere conditions. There are some very valuable tangential protective benefits for all the components which go beyond just heat management.
The compound is a 2-part, 1 : 1 ratio mix, silicone based material which seems to have the consistency of fresh construction adhesive. Once properly mixed, the thick liquid compound is poured into the enclosure fully submerging all of the components which have been already properly configured, connected, filling as much of the empty volume as possible.
I read somewhere that it takes a LONG time to fully cure, and should be left unmolested for at least 96 hours before being ops tested and put into service. Once finished curing, as expected with a silicone based material, the compound forms into a rigid rubbery cacoon and provides the desired heat management benefits, but also protects all encased components and their physical connections from water or moisture and sand infiltration, as well as from sudden kinetic shock to a certain extent.
With all that goodness, are there any drawbacks? Well, while the components are Fort Knox level pprotected from the elements they would eventually succumb to on a long enough timeline, any hardware troubleshooting, repairs, or component replacement that may be needed will have a few more steps added to that process, steps which create various risks if damaging more components in the process. I don’t know if you’ve ever had to carefully carve something out of a hard rubbery material, multiple inches thick, and containing various circuit boards and integration cables, because I haven’t, and it sounds awful.
But before we even get to the opportunity to break them so they need fixing, there is one concern I have which we’ll just need to wait until we have a fully assembled and compound cured test article in hand. The potting compound, being silicone based, would be a poor conductor of heat, cause the components to retain their parasite heat energy, and likely result in component overheating and premature failures of the sensor system. The color of the potting compound on both it’s cured and uncured state, is a dark muted gray, which I’m concerned is the result of some sort of metallic additive which is used to act as the thermally conductive material when combined with the silicone.
The enclosure case being used is manufactured to act as a large heat sink surrounding these components. The requirements include the ability to connect OTA using Wi-Fi protocols, a function normally handled without issue by the appropriately equipped Raspberry Pi board in most scenerios. That said, early testing of this set-up was conducted to determine if the metal case would cause a significant degradation of RF signals being transmitted and recieved by the onboard Wi-Fi radio, and the results were encouraging. There was a curtailment of the Raspberry Pi’s Wi-Fi signal strength as expected, but remained at levels within what was needed to move ahead with no changes. Now we’re introducing a thick layer of material which is possibly saturated with metallic particles, and it’s unclear whether silicone’s lack of electrical conduction includes radio frequency energy or not. There are certainly mitigation options if the 2.4GHz signals being recieved and transmitted through the Raspberry Pi’s native onboard Wi-Fi antenna are unable to penetrate the compound and case materials, but the preference is always to use the least number of unique components and physical connections while still meeting the stated requirements of the build.
I’m really looking forward to getting the opportunity to tinker with and develop an understanding of the capabilities of a new material which can provide a working solution to multiple challenges which were addressed individually in the past.
I’m a Certified FLIR sUAS Thermographer, so I’ll be capturing a number of Thermal-JPEG images as one element of our analysis of the stress test data. I’ll let you know how it all pans out when we know more.
FTR this is correct. The pico does not have the CPU speed (133MHz M0+, likely to be <<10% of the speed of a Zero) or the I/O bandwidth (USB 1.1, 12Mbit/s - existing demodulators need more like 40Mbit/s) to act as the software part of a SDR for ADS-B.
If you have the demodulator somewhere else (e.g. a Mode S Beast) then it can probably keep up with the decoded data stream, since that is a few orders of magnitude less data. The next challenge is then what do you do with that data - the pico itself doesn’t have any network connectivity, you’d need a board with separate wifi or ethernet.
A zero with wifi is so cheap that running from a pico would not be viable other than as an interest project.
Great detailed post by @pete3500
Geoff
I haven’t been able to get my hands on a Pico yet, but my initial read is that there is not enough compute power on the device, though I would still like to try 
Griffin, welcome to the group.
I’m not sure how a pico could be used for ADS-B - pretty sure if there was some interesting use then the Arduino (similar type of device) may have been utilised but don’t think it has.
Geoff
Found this today, so maybe a ½" step towards my dream coming true!!!
Hackaday: The Raspberry Pi Pico As An SDR Receiver.