The Angriest Radio Telescope - Wifi Camera Part 2
Вставка
- Опубліковано 11 чер 2018
- Last week we started on the construction of a mini radio telescope with 2 main goals. The first is to design a scalable system so that later projects can be built with larger and larger antennas for imaging the sky and pulling data from far away satellites. The second is to use this prototype to take images of terrestrial frequencies, starting with WIFI.
If you missed last weeks video heres a link so you can get caught up: • DIY Radio Telescope Ve...
This will be a 3 part series as this project has been one of the more complex and is taking time to sort out all the bugs and kinks. This week we finished assembly and got everything moving and starting working on the radio capturing script. Next week we'll have some actual images and will go through the coding and programs used to run this robot.
Power meter GNU radio: www.tablix.org/~avian/blog/ar...
DC offset fix: www.rtl-sdr.com/removing-that...
GNUradio: gnuradio.org
____________________________________________________________________
Support the show and future projects:
Patreon: / thethoughtemporium
Nebula: go.nebula.tv/thethoughtemporium
Ko-Fi: ko-fi.com/thoughtemporium
Become a member: / @thethoughtemporium
Store: thethoughtemporium.ca/
______________________________________________________
Our Social Media Pages:
Tiktok: / thethoughtemporium
Instagram: / thethoughtemporium
Facebook: / thethoughtemporium
Twitter: / emporiumthought
Website: thethoughtemporium.com/
_____________________________________________________ - Наука та технологія
Not a criticism, just for those who might want to know: that shim isn't a shim, it's called a key. If it worked by spreading the axle apart to tighten the press fit to the base, it would be a shim. In this case, the key bridges the axle and base and takes forces 90-degrees from what a shim would.
iroll
Thanks for the extra information
Noted. Will keep that in mind next time I use this so I call it the right thing. Thanks!
Nice to see a mechanical engi-nerd spreading their beautiful sorcery.
Yeah the 20-80 rule sucks.
I make games. For us the rule is 10-90. :P
Using a circular polarity antenna to pickup wifi signals that are normally vertical or horizontal may prove to be an issue. I think that may be the issue with the higher signal levels when the antenna is parallel to the ground. We run 20+ Mile links using equipment in the uper 5Ghz range. Most of our dual polarity antennas have 20-30dB isolation between vertical and horizontal. With the wrong polarity it can result in a significant reduction in signal.
As for the frequency hopping most router don't change frequencies that often and some are only certified to use certain ranges. If your mapping 2.4 you may be able to get by scanning the center frequency at around 2.445, alternatively you could scan channel 4 and 10 for better accuracy. Scanning 5Ghz would be quite the challenge.
Either way I'm looking forward to the results. Good Luck!
Hang on, can't you derive horizontal and vertical polarisations from a combination of circular ones? If so, then couldn't you add a second coil to the antenna, so you have one clockwise and one anticlockwise, then figure out the overall polarity of the signal in software?
A circularly polarized antenna can receive linearly polarized signals in any rotation, just with some efficiency loss. Also, the signal you were seeing while the antenna was horizontal may be coming from a nearby phone, the near by laptop, or any other source of 2.4 GHz radiation.
That was my point. A random piece of wire can receive can pick up a circular polarized signal but the SNR is going to be pretty poor. The ideal antenna for this project would be either 2 yagis, one horizontal and one vertical or a single x-pole yagi. Since you don't any data you could just use a simple combiner and a single SDR. The reason for yagis is that the beam width is more narrow so you would be able to get a higher resolution. I think the current antenna will act more like a omni pick up signals in a wider beam width, possible lowering the resolution. Either way, I'm curious to see the results.
How can a person start this without any knowledge on radio telescope?
I am always happy when I see you upload. Especially the radio-and programming-oriented videos, keep it up!
I paused at 0:10 to write this comment and tell you this : the video is starting well, I have high hopes for the future.
This project is fascinating, thank you so much for publishing!
That joke with heat death of the universe killed me.
Ive taken a 2.4 and a 5.8 Ghz Omni antenna, strapped them on a pole and wired them to a radio to see noise before. But never automated anything with a pinpoint antenna like this.
I probably would have ended up using an Yagi. But thats only because I have those antennas available to me.
I give you guys props for building the antenna. SUPER AWESOME!
I can't wait until you post the video next week! Great project
This is looking to be good. Hope it works like planned, probably won’t.
Awesome, those final pictures will be so cool!!
Wonderful, looks like good progress!
Lucky you have like a hacker space.
Here in India, we always blame unavailability of stuff to bad projects...
Thanks for these videos! Kind of inspiring!
Lol... best video introduction ever!
cool effort. the play by play is as interesting as the results.
Last 20% being 80% of work.... the truth in that actually makes me cringe xD
Love this project!
I can’t wait to make this
finally, wifi vision will be interesting to see
for the wobble issue, one aspect of things is to reduce the moment of inertia by rotating around the center of gravity. Since you probably don't want to put a metal shaft there, you can google "remote center of rotation".
To simplify everything you could just print gears and the stand...
Ps:. Awesome project !!!
So in the end, that was just 9 minutes and 37 seconds of watching a cardboard nose go up and down and you explaining why you're not the one we should be listening to. Got it.
That's cool as heck! I love the face.
to take the friction away while rotating the whole structure look up a "lazy susan" they are the tray that sits usually in the middle of a dinner table. its a circle that spins in 360 and you can buy just the metal ring that has bearings and mounting holes. they might be at a local hardware store or big name stores
COOOOL!!! Also for WIFI hotspots positioning u can just use 3 (or 2 moving) receivers (eg rpi3) with known GPS coordinates including altitude
I've started work on my own and currently working my way through the design stage quagmire. I've built the attitude component already and I am happy with it. I fully understand why you avoided 3D printing, 3D printing large flat parts SUCKS. Just for the AT part of the mount it took around 15 hours of 3D printing, not including the trial and error testing while I fumble around like a blind man in a bear trap Q&A department trying to work out what the tollerances are of my 3D printer for certain parts.
For the AT I have got a 4:1 gear reduction and for the AZ I am working on a 5:1 reduction with inside-out helical gear (not that I am tempting fate or anything :P) Though mine is going to be used for astrophotography as well as chasing amateur radio satellites.
For the lols I enabled 16th microstepping which equates to something like 12,800 steps per revolution, or 0.028 degrees per step, if my math is right.
edit: thank you for sharing yours though as it motivated me to get off my ass and revisit the one I had sitting in the parts bin for the past 2 years :)
YES A THOUGHT EMPORIUM VIDEO
Awesome video. Imagine what other interesting things you could do with just the mechanical setup like make a security camera that follows people around or water gun to target critters bothering your garden.
love your vids man
Loved it !!
Can't remember if I said in the last video, but you sound too much like 8-Bit Guy for it to be a coincidence :P
Great intro, loved the play-by-play :D You''ll get there! Keep on keeping on :)
awesome bro 👍👍👍
you could use a lazy susan bearing for the bottom / upper assembly... that would minimize friction. :)
I urge you to revisit the question of whether WiFi uses frequency hopping. My impression was that Bluetooth uses FHSS and WiFi uses DSSS. The Wikipedia article on IEEE_802.11[1] says that only the most legacy protocol uses FHSS. Also see this video where this guy detects his phone's AP on different channels/frequencies with an RTLSDR[2]. DSSS uses a pseudorandom sequence of "chips" that modulates the signal, but they don't make the frequency hop around, just spreads the signal in frequency space (they get pretty wide! but centered around a single frequency), see "2. Basic Principle of DSSS and FHSS" in this pdf[3]. I may wrong of course, I'm not an expert :) Love the content btw and good luck!
[1] en.wikipedia.org/wiki/IEEE_802.11#Protocol
[2] ua-cam.com/video/f7mrFNB1D84/v-deo.html
[3] www.sss-mag.com/pdf/Ss_jme_denayer_intro_print.pdf
I apologize for any foul tone on my part. I wouldn't have said anything if I didn't believe I was contributing to seeing some color-coded wifi channels on your final image! Curious what you actually saw.
It's funny to think that hopping is part of a security model that expects an eavesdropping bluetooth device not to know which frequency to listen to, when now it's possible to see the whole range with a SDR :)
You might have been seeing OFDM, but, if I may mention, a personally fascinating thing about DSSS is that several signals can theoretically operate on the exactly same frequency and to recover the desired signal, you sort of multiply the mixed signal with the pseudo-random chip sequence increasing signal-to-noise ratio, "digging out" a signal that was modulated with the same pseudo-random chip sequence.
I liked the video but my suggestion is to better manage the acceleration of the stepper motor. I have swung some amazingly heavy things with simple 3d printed parts, likely weaker than wood, by starting very slow. You should have 100% control over the speed from the gcode. But, maybe you want more speed, so I guess you just need more strength and more torque?
I really know the problem of:” oh this is going to be easy ... no it isn’t “
This is great!!
Yeah....
I need to subscribe now!!!!
A 'Lazy Susan' type bearing would be a cheap way to take the weight and friction off the base mechanism and make it much more solid, and then making the horizontally rotating base plate itself a large cog like in the vertical one would again take all of the weight off and give more mechanical advantage with less parts. Using a mechanism that turns the base one notch every time the vertical movement passes a certain point would get rid of the second motor and associated gearing entirely, and putting a potentiometer on the output shaft for motion feedback, basically turning it into a servo motor, should increase accuracy.
Question. Can Solder wick be used as a helical antennae? The thought occurred to me once upon a time, but never got around to making my own to try it out.
It's because your antenna is directional and the properties of wifi routers. since most wifi signals spread outward, the spikes you're seeing are a result of the antenna picking up wifi at its highest power as well as the fact that since you are close to the router, the wifi sphere is small and has a high factor for curve. Imagine a router as a sphere. When looking towards the center of the sphere, you are seeing through more area whereas if you look up at an angle to a sphere, you see less area and the output should be sinusoidal.
Here's an idea:
What if you injected aerogel before it sets and before it's dried/evaporated with tiny particles of gallium arsenide/silicon germanium at a high rate? Since there would be a gradient of denisty of high refractive index particles, the less of them the further from the surface, upon dyring you could get a very high index gradient lens, since at the surface there would be mostly solid i.e. silicon, it's content continualy decreasing towards the other surface,decreasing therefore the effective refractive index until you get just aerogel which has n=1
You can find nice gears in every laser printer. What type of "hacker space" is that where you made stuff with 3d printer?
How are you detecting the radio signals? What are you using for that?
wow that thing actually looks so funny :D
WiFi router frequency hopping? Many times a second? Is that a thing? I thought this was only used in Bluetooth... Any sources please?
intro 10/10
Damn I wish I could do stuff with you guys...
good luck guys have patience easier said than done
funny that you come up with this now. Well more or less now.
I was thinking about doing something similar, but with a different kind of antenna. (Although mine was aimed at much lower frequencies than yours, so I could find radio transmitters)
That's going to be version 3 actually! we'll be building it so we can sub the antenna out for any other kind and cover a huge swath of spectrum. Will let us do radio astronomy, but it'll be trival to aim it lower and find terrestrial stuff.
The Thought Emporium that will be cool!
The challenge will be to make it portable for me. Its supposed to be a kind of camera for handheld (lower resolution) or tripod (higher resolution)
How can a newbie enter this field? What do you recommend?
I suggest you guys a cloud chamber. That would be awesome!
Mátyás Ludvig if I was still in the US it'd be way easier. Getting dry ice in Canada is an absolute pain
The Thought Emporium I see. First time I thougth you could use a peltier module or a standard cooling unit in this project.
I don't fully understand these two things, but someone not credible informed me on the dangers and negatives on these (probably a chouch warrior of somekind :D).
Peltiers are hard to get cold enough for that to work. It's doable, but difficult.
The Thought Emporium Ok, i see!
I gotta ask - what is that stepper shield shown around 8:00 ? I've been looking for something like that for ages!
Wonder if you considered the camera you were filming the sweeps with might be generating a Wifi signal? I know GoPros and Phones definitely do. Some DSLRs as well.
This is cool
Oh nice, you just went for straight bulk ptfe bushings. That should hold up much nicer.
Bushings! That's the word I was trying to remember. Thanks! Dunno why I called them bearings.
Well, the function is close enough that they're colloquially interchanged. I mixed them up all the time until I almost lost an eye to some bad advice on removing a bushing. There's nothing like core shaking fear and hours of eye flushing with baby shampoo to make you really remember those fine details. On a related note, if you don't have a puller and it's a blind hole, hammer in tortilla or bread to push the bearing out, and don't trust ANY procedure involving grease.
What's the name of receiver module connected with Coax? I need one to measure RSSI values of 2.4 GHz rf waves using a directional antenna like this one.
Thanks.
I have a question.
Can millimeter waves travel through binoculars? Can sound come through binoculars when they're aimed at Street lighting or cell phone antennas?
This is why you plan, design and then build, but you guys did it in reverse
0:01 rick and Morty reference from the last episode of season 2
The Thought Emporium Hi first of all i have to say your projects are amazing especially wifi camera and i am wondering antenna theory but i am a newbie and i dont know how to start can you suggest a book that covers all antenna theory how to built them from scratch filters effects of whether on them THANK you have a nice day!
The steppers should have enough power. Did you tune the current of the driver?
Won't it use less energy and put less wear and tear on the system to rotate the antenna back and forth horizontally rather than vertically? If you keep turning back and forth but step the vertical incline up degree by degree i think that would probably be better.
Jonathan Olson You can see that they added some weights to the back of the vertical assembly to balance it so the motor isn't working against gravity, they also said that the base plate has a lot of friction so that'd take more energy to move.
MadScientist512 that makes sense
Danukeru DESU The higher gearing doesn't necessarily mean that more power gets used if the energy each step uses is less, although in this case, due to the friction, it does use more; they could use a 'lazy Susan' type bearing to get rid of that. Thanks for the interesting info about cogs, I hadn't known that but in hindsight it seems obvious; learn something every day!
Jonathan Olson wow. I watch so many videos like this. And i swear i see your name in the comments on so many for like yrs now xD
LordDecapo oh wow! Really? I didn't think realize that it's been that long and that somebody would notice lol
What did you used for connecting antenna with computer/GNU radio software?
Love this :O
interesting intro and vid lol
Dude its fine. Take your time. I know how you feel... but i have very limited materials...
Nice
could you make a clock like mechanism so you only need one motor? after one pass vertically it rotates to the side one.
i believe you can use an actual clock thinking about it. take a old clock and remove the hands and replace them with gears. second hand drives the up and down, with the minute hand side to side.
It's face looks like a drawing of the Stupeflip singer, King Ju
Remove those spur gears and go with a set of 3d printed herringbone gears (like the ones you found and stuck inside the box) as they have far less back lash and wobble, etc...
The Best Stooge he originally wanted to build it from stuff that was accessible to more people
Shame as that is a pretty big, and bad, compromise.
Odd, since the whole thing works perfectly. Seems like a perfectly fine compromise to me. We have no issues moving it and have tons of fine grain control. And as I said I think 15 times in the video, the point was to use readily availiable materials. Printing out parts means screwing with a 3d printer for hours which was time I didn't feel like spending on this prototype version. For the next iteration everything will be cnc cut or printed, no more hand cut pieces. So maybe for the next model.
Spur gears were never ever meant to go in reverse so a lot of the issues with bouncing and other such nonsense would never have happened, nor the loss of your time plus the aggravation, had you simply used the gears made for the job.
@The Thought Emporium If you have any questions regarding custom 3d printers I should be able to answer them. I have built 2 of them using the same board you used in this video (arduino-mega 2560 with RAMPS 1.4 shield).
WiFi routers most certainly do *not* "hop through the same frequencies many times a second" in the way shown on the diagram. Within channels, yes; across the entire spectrum (11-14 channels depending on regulatory area), no.
What is the purpose of the 250 ms capture time?
I am currently trying to recreate this project but i know nothing of frequencies. And im not sure what you are using to connect the antenna to your laptop. Please help i am stuck
Can you use a satellite dish to detect WiFi?
I have a question or I don't know if I'm understanding correctly (probably both). How are you going to scan all the WiFi channels as a whole?. While it is true that WiFi signals have a frequency of 2.4 GHz , this one is actually the band, hence all channels (if I recall correctly) have a 22 MHz band for them and 10 MHz of separation from both sides. Therefore, for example, you'd have to look for 2.446 GHz to just scan for one channel
The hackrf they are using can scan the entire spectrum fairly quickly, so it will get them all. The latest version of hackrf sweep can scan the whole usable range of the radio in one go, so 1Mhz to 6Ghz in under a second.
Soon as I figure out how to do a frequency scan, I'll have it sweep the whole band, but for now we've got it set to the middle of the band where the most signal is. But as barry said, if I wanted I could scan all 6ghz every second. That'd just be a pain to process all that data so we don't bother. Our biggest issue right now is parsing the data down. But everything works well atm and we see distinct spikes if we point at internet enabled devices, or dead silence if we point away.
I know this is way too late, as usual, but looking at the way You have GnuRadio set up You are basically capturing only a single WiFi channel out of 11. HackRF unfortunately only has max BW of 20MHz, but You seem to be driving it at 10MHz, and only at single channel ? Or are You scanning all the channels for each capture ? I don't see that set up on the GNU Radio setup you show but maybe it's not the final...
it turns like 60s batman. have these guys ever heard of a wormdrive speedreducer? i bet they could make one by cutting a box of hexnuts in half and hotglueing them on the outside of a circle
I'd put money on GnuRadio having a silent G as in Gneiss. So it's "NewRadio". Geddit?
I might be wrong but it fits...
So can wifi signals be read and translated into pictures by computers/AI/algorithms? Can the wifi data be compressed/merged/transform/translated into brain waves? And communicate with our brainas?
Why didn’t you just attach it directly to the stepper motor?
Frequency hopping is not used by modern (non ancient) WiFi. Only oldest 802.11 from 1997 used that. Now it is used by Bluetooth too.
BarsMonster for many of the 5Ghz channels, dynamic frequency selection is mandatory. This is to avoid interfering with various weather radar and military applications.
DFS is very slow, it takes minutes and stays on new frequency. It is not switching continuously like in regular frequency hoping. If it does not detect a radar - it will never switch frequency.
Yes, OFDM modulation is used within 1 selected WiFi channel, not across multiple channels. I've seen WiFi on 50Mhz-wide SDR receiver at both 2.4 and 5Ghz bands - there was no frequency hoping between channels.
Is there any reason to not connect the antenna shaft directly in the motor instead of building this wobling gears?
Hum, I imagined those steppers could hold more power, since DIYers use those on simple CNC machines. Well... the project is amazing nontheless.
Like the support the helical is wrapped around??? The reason is because there's no way we could fit that into the mechanism without the base plate smashing into things and putting metal neat the antenna would screw with it's directionality so you want it away from things. We also built it this way to avoid collisions. Also the gears don't wobble now that we've tuned everything. It all runs really smoothly.
Thanks for the answer! In my mind the design and position of the steppers would be completely different, but I can't assume it would just work without building one like you did. Thanks for sharing this project. Stay strong! You are awesome!
Would worm gearing have given a more accurate control.
yes but that would've required making worm gears which we didn't want to do at the time. We will for the next version though
Where are you located?
Isn't the WiFi is just a vertical polarisation? If so what advantages does it have to use helical antenna? Yagi may be a better option
I think you can derive linear polarisations from a combination of circular ones. Two antennas wounds in opposite directions (one clockwise, one anticlockwise) should then be able to pick up all polarisations, and you should be able to figure out which ones were horizontally and which were vertically polarised while doing the data analysis.
Yagi is a directional antenna (linear polarized) and this of kind is a bit like we can say circular polarized yagi.
Ateeb Tahir The Yagi design is a number of parasitic elements arranged in a row with one driven element and a reflector, this means that one Could make a Yagi with circular polarization; a quick search shows such a thing Does exist and is called a "Cross-Yagi."
Cross yagi !! Hmmm. Instead of single antennas there are crosses (one rod for each axis). Am I right?
Ateeb Tahir Right indeed!
Never uploaded a part 3 ☹️
ua-cam.com/video/g3LT_b6K0Mc/v-deo.html
Maybe the stepper interference is showing up? I know I can't run my radio and 3D printer in the same room at the same time because music changes to something like shshshsshadshdhshsh
Nope we checked last night. Point it at a router and the thing goes crazy, point it away and it's dead silent. We even tried putting a phone infront and when you refresh pages or use the internet we see a spike
its a bushing, not a bearing
I think you need a lazy suzan
This start
Clickspring eat your heart out.
Why aren't you using bearings instead of teflon?
We found the teflon first. Once everything was cut and and mounted we actually found a box of bearings. Ended with us just looking at each other feeling like derps. But teflon worked really well anyway since it's a nice low friction plastic.
Why didn't you just get a lazy susan?
put a yagi on it!
I still dont see why you use horrible toothed wood instead of using air so you could get it smother and more precise
Epoxi validity: end of the universe hahahaha
Dude!!! Thanks your awsome! I was laughing my ass off!!! Reality is a b!+€#! I really understand and will always appreciate MacGyver engineering and repurposing.
Anyway for a possible future coming study, try laser range finding the moon.
And take old ranging data and try correlating it to the Earth's magnetic field shift! Something I'm concerned about. There's a laundry list of issues that come with this idea.
Keep it up!
I Love the Rick & Morty reference!!!
first 5 seconds ,already liked the vid
Hi
Nico Enrique Limonge Colomer hi
Bout this content...instaclick