Sure they may not "capture" plates, but if it's a camera, there's a picture of it and you can just read it. If you don't have a camera in position where it can see the plate, then it's not going to "capture" it either. There'd be no benefit from a plate-capturing system compared to a normal security camera in the case of something being stolen. The only use for something that captures plate numbers is for processing and tracking of them. (I suppose, maybe, the logs would show the plate suspiciously appearing repeatedly. But you could also just see that on a video playback.)
POE inductively coupled through glass?
- Thread starter stevenkan
- Start date
The camera settings to get maximum automated readability are conflicting with the settings to get human readable captures. There are probably settings where systemic reading is superior to human guessing.
The cameras that I use for human monitoring have a 129º FOV. They don't have enough resolution (lens or sensor) to read a plate. Here's the 2MP/60mm license plate camera:
Here's approximately the same region of interest from the 5MP, 129º camera, cropped to match. These two cameras are installed 6" apart from each other:
I still know very, very little about license plate capture/recognition. But what I've learned over the last 48 hours is:
That's why nearly all license-plate cameras are installed side-by-side with a non-LPR camera. One captures the scene; the other captures the plate. There really isn't any way (yet) to do both with one camera.
And it's not just resolution; it's also the optimization of the exposure. Here's those same two cameras again:
If you're optimizing for the plate, you won't see anything else.
Yeah that was my take as well. I thought about it for a bit then just had someone run a cable. Still, I spent a few minutes on Amazon just now and there are some surprisingly interesting wireless power devices:
View: https://www.amazon.com/Taidacent-Distance-Inductive-Charging-Wireless/dp/B07VZZR4ZK/ref=sr_1_5
This (bad!) review mentions:
That is a surprising amount of power through that distanceWe have a fair amount of wildlife (coyotes, foxes, deer) wandering through our yard, it would be neat to be able to clamp a 4k camera on any window for a couple days at a time, but I should probably just buy a battery powered trail cam and stick on a tree.
If you look on google there are whole forums of people who do this kind of stuff, and there are community favorite models for different applications (cars, night surveillance, weather photography, etc). If you want to get license plates it is all about having a very carefully lined up camera and a pretty long focal length lens. I wasn't so interested in that, but I ordered one of these random "empiretech" cameras those forums recommended for low light and it is pretty remarkable. It makes the night mode on my Pixel 8 pro look embarrassingly bad and does so without insanely long exposure times. I think if you really pick the right equipment for the right job you can do pretty well, but theres a lot of expertise involved. I'm not surprised some random sales guys did a shit job for your work.
And I bet if you go to AliExpress you can get it for 1/4-1/3 of the cost including shipping, as long as you're willing to wait a few weeks.
How do you think a 10khz oscillating magnetic field get generated? Hint: same principle as running current through a wire, which generates a field, and electromagetic radiation in a wave form when it varies.
Even spinning a magnet around in free space does that, it turns momentum into a tiny bit of EM.
The simplest induction, 80khz charger I own is from toothbrushes. It has pretty damned good coupling with a peg and hole configuration. Throw a single aluminium foil in between though and it quits, while that does absolutely nothing to a static field and you can stick that with a magnet on a fridge. QED, unless they win a nobel and discover EM that works differently.
Hi All, Inventor of the Power Mole here. Just some notes: the power transfers through low-E windows, this coating does not affect the power transfer. Power transfer is currently 60-90% depending on load and gap size, and I'm also working on improving it. This is not off the shelf Qi charging adaptation (different frequencies, different algos, etc) and is not some foreign rebrand. Built from scratch. Kickstarter was successfully funded and we are shipping out next month. Some details available on youtube (search power mole), and updates can be found by following the Acqua Industries website. Cheers
That is surprising as you note on Kickstarter that metal blocks the transmission and well, my windows are very metallic and severely impacts reception. My accountants office is even worse, triple layer glass and it's a damned faraday's cage with no 3/4/5G reception, phone only connects on 2G/Edge but zero issues once you walk outside.
But hey maybe not all low-E windows are made the same.
Here's the thing...PoE is not just power, but DATA. Data is where you'll have some real challenge. Not saying it can't possibly be done, but I'm going to go with Not Easy(tm).
I think the data would be the easier part. Just flash some LEDs through the glass at a detector on the other side. Doesn’t even have to be Gigabit, as most IP cameras are 8 Mbits or less.
I would have thought gigabit would be the minimum.
And I suspect free space optical, complicated by going through at least 1, possibly multiple, panes of glass is probably pretty hard. Not that it can't be done, but I'm unsure if basic LEDs and sensors would be doable in a very reliable way.
I think the situation some were considering in this thread was only using the LAN port for power, and using Wi-Fi for the data. That's how cheap USB-powered cameras work. So the device would only be needing to transmit the power through the glass and converting it to an RJ45 port on the receiver. If it was used for passive PoE that would be essentially the same as the Power Mole, just a different output connector, because it's always the same power level being output, but would require different voltage and amperage from a USB device so the question is whether enough power could be sent through the glass to the receiver. For standard PoE, you also have to add a level of data communication so that the PoE negotiation can occur. However that could actually be done on the receiver side and not even necessarily need to be transmitted through the wireless, now that I think about it.
A loop antenna has to be on the order of a wavelength (in this case 30km) in order to radiate any significant energy, but ok lets go with your idea and run the numbers.
Efficiency of a loop antenna drops off with the 4th power of circumference/wavelength, so a 30cm loop antenna at 10 KHz would be 10^20 times less efficient than a wavelength scale one. I think it goes up with the square of the number of turns, so lets say 100 turns, that gives 10^20/100^2 = 10^16 less efficient. You have the same losses on the receive end too, so about 10^32 times less power transferred.
So no, in practice the radiated field is nonexistent since the coil is many orders of magnitude too small to function as an antenna and instead all energy is magnetically coupled. Which makes sense. 10 KHz radio stations need antennas that cover kilometers, not centimeters.
Seems petty to imply that his product won't work.
More like a request for clarification, the metal blocks it bit comes from his page and matches my induction charging experiment with a bit of aluminum foil.
I'm still waiting for a source why you think this is magnetic, sticking aluminium with a magnet on a fridge works yes, but that is 0hz. There is no magic frequency when EM changes from magnatic to electromagnetic.
The best explanation I have seen is this: https://electronics.stackexchange.c...hielding-electrically-shield-magnetically-too
35 micron copper shields well against 5Mhz signal, but to shield against 100khz, it needs to be sqrt(5000000/100000)=7x thicker or ~250 micron. So, a sqrt falloff but thick enough foil can kill it.
That also means that a very thin invisible silver layer probably won't block a low frequency EM signal, according to this formula, but it clashes with the previously linked paper stating:
And they were talking about transparant, conductive glass.
Wait, but how does my 10 micron supermarket foil kill the induction charger then? It falls off completely, with air hovering at 1cm, the toothbrush still picks up a bit to light the LED. Foil blocks it completely. Someone is wrong and I think it's the Stack Exhange guy.
Anyway, t.b.d. and maybe hugely depends on glass type and coating. It certainly has an effect and but drops off by sqrt(frequency), so maybe at low enough frequency and super low thickness this doesn't matter. My office has sun blocking coated glass and it's worse than normal glass. Same with window foil that uses silver to reflect / reduce solar irradiation by 80%+.
[edit] Different source that makes magnetic shields from mu-metal
https://mecamagnetic.com/magnetic-shielding/ - They say their stuff works from 0.1 to 10 khz, above that, please just use copper / aluminum. And induction charging is typically above that frequency so ferromagnetic material doesn't work against it. Also note that < 1 khz @ copper/aluminum is just rated "-", not doesn't work, just bad at it relatively.
So evidence is 2-0 so far, EMP paper (20khz+) and Meca Magnetic (10khz+) both claim copper/aluminum works against khz+ fields. This discussion though is separate from if an induction power coupler will work through double glass, more a theoretical one, if a varying magnetic fields are a completely separate thing from EM.
Conductive material having an effect doesn't mean you can't overpower it, especially optically transparent > 90% transmission glass, any coating would be microscopic thin making attenuation maybe low enough to not matter. They did state 60-90% efficiency, maybe lower bound is for coated double glass.
Last edited:
Well, alternatively, it's 2 seconds with a crowbar or can of paint. Any exposed camera is inherently vulnerable, so in our case, I haven't stressed about access to the cables. If you want to spend a few moments unscrewing the coverplates, you can disconnect them. But that's not really the point of cameras IMO.
It's to get notified when someone is fucking around, and ideally have some video to show the cops. (They're on a secluded VLAN so as to prevent access to the network, though).
Of course I'm assuming a residential setting where you probably don't have a security team to react either.
Yes, power only seems possible. You could also just use a PoE extractor to pull the power off to feed it to the input. Although... Why? You have to run a thicker, bigger cable than the basic power cable to it, you still need wires. Only thing I can think of is it allows for switch controlled power cycling. Rather than walking over, unplug it, war a bit, plug it back in. For me, too much of a PITA, for a home user. Maybe there's some edge cases that might beul useful, I can't see it though.
I'm not entirely sure what you're saying, or even whether you're arguing that a Power Mole for PoE would be good or bad. But the point is to be able to put a camera outside where there is no power outlet easily accessible. So you can run a Cat5e cable from the receiver to the camera, without having to drill holes for cabling or route around the entire building or what have you. It's not a matter of whether plugging into a switch would be easier, it's a matter of plugging into a switch not being an option. (If you could run a cable for PoE, there'd be no point to not just usine the PoE for data, too. If you meant injector, not extractor, well, this is for when you can't run cables so an injector still won't work.) It's definitely a very small market. Most home users would be fine with the existing version which just provides low USB power output, but that could be a large enough market to be profitable without it being terribly expensive, if it actually gets traction and news and reviews.
I'm making the argument, that PoE power only, without data, is a bit of a silly thing for a Power Mole style power condutor. Cat5e is bigger and more to work with than a regular 2 conductor 5-48v power cable to go to a standard Power Mole.
If you can get decent data rates and reliability from something like a Power Mole (for which I question being reasonably low cost for, but I could be wrong) then what's the point in running PoE to power it? And if you're using WiFi for data anyway, than doesn't make much sense to use PoE to power a Power Mole.
You seem to not have read the thread completely or aren't understanding. The point is for cameras that only have PoE power capability, where a low-voltage, low-amperage USB connection wouldn't be enough, but also have Wi-Fi as a data option, and where you simply can't get an Ethernet cable outside to the camera. Or the camera may have both PoE and a DC power jack, but you don't have an outside power outlet nearby. The Power Mole transmitter is just plugged into a power outlet on the inside of the building that is near the window. The receiver is on the other side of the window and supplies PoE power to the camera. There is no intent to get data sent through the Power Mole, and the Power Mole is not being powered by PoE. The PoE supplied to the camera is so that it's a simple, standard connection which is very common in such IP cameras, because the Power Mole can't just have one type of DC output that would work for every camera. They don't all use the same voltage or even the same connector.
But, even using PoE to the Power Mole WOULD still be something that some small number of people might be interested in. The entire point of the Power Mole is simply that you CANNOT get a cable all the way to the camera. But if the Power Mole could supply both power and data (and the PoE cable would mean only one wire, rather than needing to have a power cord as well), there would be a use for that. The only question is whether data could be sent through the wireless connection without making it very slow and/or very expensive, and if it would even be profitable to produce such a device given that it would not sell many units.
PoE is an annoying standard though, you would need to boost DC up to 48V at least and then make sure you have about 15W net, after loss, power available.
I would look for a different camera that also accepts 12V DC. I have a Hikvision Wifi cam that does wifi, PoE with data and also plain 12V barrel jack input. Dahua, Axis all sell 12V, 24-40W adapter blocks.
I would look for a different camera that also accepts 12V DC. I have a Hikvision Wifi cam that does wifi, PoE with data and also plain 12V barrel jack input. Dahua, Axis all sell 12V, 24-40W adapter blocks.
Oh. You mean output is PoE? Oh, I thought you meant PoE the entire way/power input, not output outside.
Yeah, that is potentially possible. Most cameras I've seen that have PoE also support 12v in, at least the cheap to decent consumerish cameras. Power Mole would need to be able to do 12v@1a output, haven't looked at that. But that doesn't seem farfetched to be able to do.
The desire is for it to have PoE output, which is standardized voltage and amperage that will work with any device that requires PoE (assuming the af/at/bt standards match) with a standard connector. Having the Power Mole output regular DC would require a jack with a range of cables or adapters in order to be able to connect to different cameras, which could further limit the number that would be sold.
The transmitter just needs to be able to shove a certain wattage through the glass. The receiver can be more complicated, and they could sell multiple models. One for USB output, several for passive PoE which is less common, and one for active PoE. The cost of the transmitter that can send enough power for a 10W USB port would be pretty low. The cost to make it able to push out enough energy for 15.4W PoE or 30W or more may or may not be so different as to make it worth having different models at different prices, and the receiver with passive PoE may be as cheap as USB but would also require multiple models, because all passive PoE devices aren't compatible. For active PoE it will be more expensive as it requires active electronics for negotiation.
Don't forget, this can also power many different devices, which might have higher power requirements. Cameras are just the one that is probably going to be most common.
Sure, but are cameras that do PoE going to generally also have WiFi for the data connection? I haven't done extensive research so I don't know.
You're waiting for a source that transformers use magnetism?
A varying current in any coil of the transformer produces a varying magnetic flux in the transformer's core, which induces a varying electromotive force (EMF) across any other coils wound around the same core.
Transformer - Wikipedia
Ooof you are in deeper than you know.
Not sure how I can make this any more clear, but loses scale with thickness. You've got a piece of metal thats ~100x-1000x thicker than an optical coating, so theres a lot more eddy current and thus a lot more loss. Eventually you get thick enough that essentially all of the energy is lost to resistance in the metal. But when its nanometers thick the coating is essentially transparent because so little current can flow.
"Alternating magnetic fields of sufficient frequency however, will not pass through a metal plate. The alternating field generates an eddy current in the plate which generates a cancelling magnetic field."
So basically, a thin metal coating passes an alternating magnetic field while a thick metal sheet does not, just like I said a couple times now.
"Everyone who disagrees with me is wrong, especially the experts!"
You're trying to google for a physics degree worth of knowledge but starting out with a sub-wikipedia level understanding. You're at the point where you're going to need a couple months of study to even understand how confused you are.
Eh whatever dude. You're just throwing out claims after claims that goes against a published paper using coated, optically clear glass and a company that literally makes mumetal and high field strength magnetic shields. Induction charging isn't a static magnetic field and a varying one at 10+ khz is killed by conductive metal. Induction is in the 100-200 khz range which is in the range of tested frequencies in the earlier linked paper, measuring 40 dB loss.
You have zero citations so far that prove this is false. Just blah and more blah.
Stop just throwing out claims and go look for a link from an actual expert. Published papers are good, Wikipedia has citiations, and companies / people working in the actual field are great sources too. You're 2-0 so far with your "this is magnetic, not EM" claims.
You have zero citations so far that prove this is false. Just blah and more blah.
Stop just throwing out claims and go look for a link from an actual expert. Published papers are good, Wikipedia has citiations, and companies / people working in the actual field are great sources too. You're 2-0 so far with your "this is magnetic, not EM" claims.
The Power Mole is about 82mm in diameter and can do 10W across 30mm. POE at 48V and 15W can be done. The first issue is thermals, efficiency optimization and a larger size for heat dissipation can help with that (and also range). The increased power is main issue. Bumping up voltage is not a huge concern. Power Mole does 12V in, rectifies down to 5V on the receiver side. But this could be scaled with 48V in and 48V out (or similar), and potentially using a voltage boost circuit on the receiver side.
For data, this is not supported. You could add some sort of short range direct WIFI link possibly or Steve's optical method, but that's an extension I haven't looked at.
For low-E: cellular phone frequencies are 600MHz and up. They are propagating EM fields. The power transfer here is magnetic field coupling. E-fields are not strongly generated here. The frequencies used here are sub 1 MHz. I can't post much here since I am new to the forum. I don't think I can add links either, but today I uploaded a video to youtube titled Low-E Glass Windows and The Power Mole, which is on the AcquaIndustries youtube channel.
For data, this is not supported. You could add some sort of short range direct WIFI link possibly or Steve's optical method, but that's an extension I haven't looked at.
For low-E: cellular phone frequencies are 600MHz and up. They are propagating EM fields. The power transfer here is magnetic field coupling. E-fields are not strongly generated here. The frequencies used here are sub 1 MHz. I can't post much here since I am new to the forum. I don't think I can add links either, but today I uploaded a video to youtube titled Low-E Glass Windows and The Power Mole, which is on the AcquaIndustries youtube channel.
You are limited in posting until you hit 5 posts as an anti-spam measure. I've "reported" your post asking for your limitations to be removed early as you are obviously not a spambot and are making extremely valuable contributions to this topic.
Welcome to Ars!
Of course there are always:
My original question was about POE* in and out of some pair of coupled modules, e.g. so that it looks like an extension of a regular Ethernet cables as far as the endpoints are concerned. e.g. the glass and couplers should be . . . transparent to the application
* and the great thing about the POE standard is that there are so many to choose from! For my target application, e.g. IP security cameras, I'm desiring 802.3af/at, which is actively negotiated POE, and NOT passively injected POE.
The vast majority of these cameras will take either 802.3aX on their RJ45 jacks or 12 VDC on their barrel jacks, and I'm pretty sure they all regulate or convert the POE down to 12 V internally, regardless of what the original POE voltage is. I know this because on many of my cameras I've bypassed the diode downstream of the barrel jack, to turn it into an output, to power an external IR illuminator, because spiders:
So a POEwer Mole™
wouldn't have to support every flavor and voltage of POE. Anything that can power a 12 V camera to ~15 W will work for the vast majority of IP cameras on the market. And it's completely reasonable, for example, for a product to require 802.3at in and produce 803.2af output.Some cameras do both, but the target market for IP security cameras has an allergic reaction to WiFi, for fears (real or not) of jamming or snooping. They tend to trust hardwired connections only.
I have a decided dislike more to the sometimes unreliability and crowded spectrum in urban/apartment dwelling areas. So, that's my preference for PoE in this case.
That said, while it'd be really great for something like this to actually work with data, I do wonder if it can be done reliably without getting significantly more expensive.
I haven't done any research, so who knows, maybe there's a couple of modules that are only $30 that can be integrated for not too much that can handle it, given the fairly short distances.
I am an expert and I have written papers on electromagnetics.
Frankly you should apologize to the guy for making him go buy a whole window from home depot just prove what should have been obvious. What a waste of time and money just because you got really upset.