Post by thread:[EE] Video over long distance

List, I've got a relative who needs to get video (color PAL) over akilometer however only on a single twisted pair. I know this isprobably beyond my electronics ability so an off the shelf styleapproach is probably needed. He has been talking to an "expert" who has suggested a solution I amhaving trouble with…. They have been told to: Use a small PC to acceptup to 4 video signals and act as a streaming web server to stream outof a 10mbps Ethernet. Then use a dongle which will convert Ethernet to12mbps over a single twisted pair and back. Considering they want to record this to DVD for later viewing I can'tsee this will work too well. All of the camera to Ethernet I have seenoperate like web cams sending jpegs to a web browser. To convert thisto a DVD will take a fair bit of processing on the behalf of a laptop. The video is for remote diagnostics and so production quality is notnecessary. Also they are locked into some twisted pair they alreadyhave. Anyone got ideas ?? Liaaa

-----Original Message----- From: spam_OUTpiclist-bouncesTakeThisOuTmit.edu [.....piclist-bouncesKILLspam@spam@mit.edu] Sent: 21 October 2005 11:18 To: Microcontroller discussion list - Public. Subject: [EE] Video over long distance List, I've got a relative who needs to get video (color PAL) over akilometer however only on a single twisted pair. I know this isprobably beyond my electronics ability so an off the shelf styleapproach is probably needed. He has been talking to an "expert" who has suggested a solution I amhaving trouble with…. They have been told to: Use a small PC to acceptup to 4 video signals and act as a streaming web server to stream outof a 10mbps Ethernet. Then use a dongle which will convert Ethernet to12mbps over a single twisted pair and back. Considering they want to record this to DVD for later viewing I can'tsee this will work too well. All of the camera to Ethernet I have seenoperate like web cams sending jpegs to a web browser. To convert thisto a DVD will take a fair bit of processing on the behalf of a laptop. The video is for remote diagnostics and so production quality is notnecessary. Also they are locked into some twisted pair they alreadyhave. Anyone got ideas ?? Liaaam Do they actually need four separate video streams at once or just one? There are lots of commercialy available solutions for video over twisted pair, but they are generaly limited to 300meters or so. I think you are going to struggle to get useable quality from a purely analog solution over 1km. If it's just for diagnostics, could the video not be monochrome? That would significantly reduce the bandwidth requirements. The Ethernet idea sounds feasable. Running ethernet over a single twisted pair implies only half duplex operation but for streaming video using UDP this shouldn't be too much of a handicap. A 10Mbit link should be able to support very good quality video, especialy with Mpeg4 Divx/Xvid type encoding. Regards Mike ======================================================================= This e-mail is intended for the person it is addressed to only. The information contained in it may be confidential and/or protected by law. If you are not the intended recipient of this message, you must not make any use of this information, or copy or show it to any person. Please contact us immediately to tell us that you have received this e-mail, and return the original to us. Any use, forwarding, printing or copying of this message is strictly prohibited. No part of this message can be considered a request for goods or services. =======================================================================

If you really need full-motion 30 (or so) fps video, than some digital solution is probably the way to go. Although, considering ethernet over twisted pair actually works, I wonder if modulating the video at a relatively low frequency, say 3-5 MHz, wouldn't help it survive the trip. If you don't really need full motion, then some slow scan technique should be a piece of cake over a km. There are relatively low cost RF solutions that will work over a km. Depends on laws, though. Might not be a choice in all jurisdictions. --McD {Original Message removed}

{Quote hidden}> -----Original Message----- > From: piclist-bouncesKILLspammit.edu [.....piclist-bouncesKILLspam.....mit.edu]On Behalf > Of liam . > Sent: 21 October 2005 11:18 > To: Microcontroller discussion list - Public. > Subject: [EE] Video over long distance > > > List, > I've got a relative who needs to get video (color PAL) over > akilometer however only on a single twisted pair. I know this > isprobably beyond my electronics ability so an off the shelf > styleapproach is probably needed. > He has been talking to an "expert" who has suggested a solution I > amhaving trouble with…. They have been told to: Use a small PC to > acceptup to 4 video signals and act as a streaming web server to > stream outof a 10mbps Ethernet. Then use a dongle which will > convert Ethernet to12mbps over a single twisted pair and back. > Considering they want to record this to DVD for later viewing I > can'tsee this will work too well. All of the camera to Ethernet I > have seenoperate like web cams sending jpegs to a web browser. To > convert thisto a DVD will take a fair bit of processing on the > behalf of a laptop. > The video is for remote diagnostics and so production quality is > notnecessary. Also they are locked into some twisted pair they > alreadyhave. > Anyone got ideas ?? > Liaaam You might be able to do this using 10baseT cable (which will give you 4 twisted pairs) and use a video amplifier as a ballancing amp at the sending end and then another one at the recieving end to unbalance it again. You will definately need some (a lot of) equalisation (i.e you will have to give more gain to the higher frequency components to conteract the frequency response of the cable) to get reasonable quality out of a system like this. I would think it should be doable. CCTV security camera systems use this sort of cable for transmitting video as standard, but I don't know how far they usually get it to go. Someone else suggested modulating at a lowish frequency. Modulate at 50Mhz or so, and put the signal through a balun at each end. That will almost certainly work, and quality will depend on the modulator/demodulator rather than the response of the cable. Good luck. Mike.

>>Assuming you're outside and in a non-city environment (likely, by >>the description) you might even be able to use a cantenna type >>setup with 802.11 wireless networking. Because of the industrial nature of the environment wireless just will not work. >>You cannot send ethernet over 1km This is the impression I had which caused doubt in the proposed idea of 12mbps ethernet over a single twisted pair, even if not duplex. > > > > not electrically, but surely fiber should do the trick? > Fibre was my first idea however this needs to "fit in" with an already existing loom of wire in which a single twisted pair has been alocated. I was familure with the idea in security camera dongles for 300m or so but not for over a kilometer. Does anyone have any experiances with them ? I've done some more googling and found most described as passive but the odd one is active and some do quote longer lengths http://www.123securityproducts.com/viovtwpa.html I think an analogue solution is desired since then a standard DVD recorder can be used to "live" burn disks. >>Maxim and IIRC Linear Technology provide chips for doing this. I've had a look at the maxim site however havent been able to see any chips along these lines. Perhaps I am looking in the wrong area, do you have a part number or know of an application note ? Thanks, Liaaam

On Oct 21, 2005, at 1:55 PM, Wouter van Ooijen wrote: >> You cannot send ethernet over 1km > > not electrically, but surely fiber should do the trick? > The main reason for not allowing more than 1km of ethernet has to do with retaining the ability to detect collisions in a timely manner. For something one-directional like UDP video, you MIGHT be able to get away with longer distances (but electrical issues will show up sooner or later; I'm pretty sure that to go even 1km you're supposed to invoke several "repeaters" (which never caught on much, thankfully!) If you have power along the way, you can insert arbitrary numbers of bridge/switch type boxes... BillW

William Chops Westfield wrote: > On Oct 21, 2005, at 1:55 PM, Wouter van Ooijen wrote: > >>> You cannot send ethernet over 1km >> >> >> not electrically, but surely fiber should do the trick? >> > The main reason for not allowing more than 1km of ethernet has to > do with retaining the ability to detect collisions in a timely > manner. For something one-directional like UDP video, you MIGHT > be able to get away with longer distances (but electrical issues > will show up sooner or later; I'm pretty sure that to go even 1km > you're supposed to invoke several "repeaters" (which never caught > on much, thankfully!) If you have power along the way, you can > insert arbitrary numbers of bridge/switch type boxes... You can now get POE (power over ethernet) switches (which are basically repeaters when only two ports are connected). R

> >> You cannot send ethernet over 1km > > > > not electrically, but surely fiber should do the trick? > > > The main reason for not allowing more than 1km of ethernet has to > do with retaining the ability to detect collisions in a timely > manner. Indeed. That's why I suggested fiber, in which light travels with the speed of light :) Just don't try this on the disk. Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu

On Oct 21, 2005, at 11:51 PM, Robert Rolf wrote: > You can now get POE (power over ethernet) switches (which are > basically repeaters when only two ports are connected). > They're basically bridges rather than repeaters (a repeated is a level 1 signal relay, whereas bridges and switches are level 2 packet devices that effectively create a separate ethernet with a common level3+ address structure.) And I'll bet POE has more trouble sending POWER over 1km than a typical ethernet interface has sending signal over that distance! BillW

Bill's right on this one. Except in an exceptional vacuum (deep space), the speed of light is always less than the official "Speed of light". To my knowledge, nobody on earth has moved anything at the speed of light yet...but they have moved a few things very close to it... in a well- prepared lab. Light in a fibre optic cable constantly bounces off of each inner side of the fibre, so travels much farther than the distance of the cable. --Bob William Chops Westfield wrote: {Quote hidden}> On Oct 22, 2005, at 12:54 AM, Wouter van Ooijen wrote: > >> >> That's why I suggested fiber, in which light travels >> with the speed of light... > > > But it doesn't. In fact, IIRC, an ethernet signal on coax (sorry; old > knowledge) travels at about .77c while light down a glass fiber only > travels at about .67c > > (index of refraction is ration of speed of light in substance to speed of > light in vacuum. For glass it's about 1.5) > > BillW -- Note: To protect our network, attachments must be sent to EraseMEattachspam_OUTTakeThisOuTengineer.cotse.net . 1-520-777-7606 USA/Canada http://bem.to/azengineer

Wouter van Ooijen wrote: >>>>You cannot send ethernet over 1km >>>>not electrically, but surely fiber should do the trick? >>> >> >>The main reason for not allowing more than 1km of ethernet has to >>do with retaining the ability to detect collisions in a timely >>manner. > > > Indeed. That's why I suggested fiber, in which light travels with the > speed of light :) Just don't try this on the disk. And how does that compare to the speed of an electric signal in a copper wire? As far as I know only about 50% faster. Rob. -- Rob Hamerling, Vianen, NL phone +31-347-322822 homepage: http://www.robh.nl/

On 10/21/05, Mike Hord <mike.hordspam_OUTgmail.com> wrote: > > There are relatively low cost RF solutions that will work over a km. > > Depends on laws, though. Might not be a choice in all jurisdictions. > > Assuming you're outside and in a non-city environment (likely, by > the description) you might even be able to use a cantenna type > setup with 802.11 wireless networking. > > I've heard of sufficiently excited people getting distances in excess > of 100km with 802.11, Show me two of those excited. They need cold water. There are a few problems with transmission: - for 100km it must be stright line, direct visibility and good wheather = almost impossible, now you have Wilma's interferencies. Vasile

On 10/21/05, Peter <@spam@plpKILLspamactcom.co.il> wrote: > > Buy a RF modulator and build an adapter to the impedance of the cable. > With 10mW in it should be more than enough to go 1km even if the cable > is lousy. Joking as usualy... An impedance adapter will kill -1dBm at least from the output power if it's perfect. One km of shielded cable will have so high parasitically capacitance that you'll see only green stars (or white snow ) on the cable's end... Take the best double shielded cable and compute the attenuation after 1Km. Take a twisted cable and see the high noise versus smaller capacitance. cheers, Vasile

On Mon, 24 Oct 2005, Vasile Surducan wrote: > On 10/21/05, Peter <KILLspamplpKILLspamactcom.co.il> wrote: >> >> Buy a RF modulator and build an adapter to the impedance of the cable. >> With 10mW in it should be more than enough to go 1km even if the cable >> is lousy. > > Joking as usualy... > An impedance adapter will kill -1dBm at least from the output power if > it's perfect. One km of shielded cable will have so high parasitically > capacitance that you'll see only green stars (or white snow ) on the > cable's end... Take the best double shielded cable and compute the > attenuation after 1Km. Take a twisted cable and see the high noise > versus smaller capacitance. I don't really care about the attenuation because I know I can get a good picture with -40dBm at the receiver (~=100uW, ~2mVrms, 50Ohms). 10mW in = +10dBm (~0.7Vrms). 50dB is the approximate attenuation for coax over 1km at 50MHz (VHF modulator). BUT I agree about the signal to noise ratio at the receiver. It will be bad and it will probably doom the project. He could increase transmitter power until the cable is no longer necessary ;-) On the other hand, sending 802.x digital signals without a repeater over the same cable will also have about 40dB attenuation and the ethernet receivers are *much* less well prepared to deal with a low level broadband signal than tuned amplifiers for RF. Peter

The centre refractive index for fibre is about 1.48. giving a velocity factor of 0.672 which is very similar to coaxial cable with solid dielectric. Twisted pair cable is likely to have a slightly higher velocity factor as the effective dielectric constant will be lower - therefore the fibre will probably work out slightly "slower" than the cable. RP On 23/10/05, Rob Hamerling <RemoveMEr.hamerlingTakeThisOuThccnet.nl> wrote: {Quote hidden}> > > Wouter van Ooijen wrote: > >>>>You cannot send ethernet over 1km > >>>>not electrically, but surely fiber should do the trick? > >>> > >> > >>The main reason for not allowing more than 1km of ethernet has to > >>do with retaining the ability to detect collisions in a timely > >>manner. > > > > > > Indeed. That's why I suggested fiber, in which light travels with the > > speed of light :) Just don't try this on the disk. > > And how does that compare to the speed of an electric signal in a copper > wire? As far as I know only about 50% faster. > > Rob. > > -- > Rob Hamerling, Vianen, NL phone +31-347-322822 > homepage: http://www.robh.nl/ > -

>> I've heard of sufficiently excited people getting distances in >> excess >> of 100km with 802.11, > Show me two of those excited. They need cold water. > There are a few problems with transmission: > - for 100km it must be stright line, direct visibility and good > wheather = almost impossible, now you have Wilma's interferencies. What he says. And, to get range you need good antenna height at both ends and significant clearance in between. This is to provide that line of sight path above the curvature of the earth and to avoid "Fresnel Zone" problems due to interference with phase shifted reflected signals. (Google knows). Ranges of 100 km are technically achievable but in addition to the above you'll want an excellent antenna at each end (probably a largish parabolic), as much power as you are (possibly not) allowed (although aerial considerations usually swamps power limitations), low loss and short feed lines and good equipment. Alignment becomes crucial with high gain aerials. Then pray for no rain or operate in eg Arizona - there's an excellent reason that microwave ovens share the same frequency band :-). RM

Intersil has chips to transmit video over CAT-5 This might be of interest:www.intersil.com/products/deviceresults.asp?i=13295 my 2¢ Regards, On 10/21/05, liam . <spamBeGoneliaaam2spamBeGonegmail.com> wrote:> List,> I've got a relative who needs to get video (color PAL) over akilometer however only on a single twisted pair. I know this isprobably beyond my electronics ability so an off the shelf styleapproach is probably needed.> He has been talking to an "expert" who has suggested a solution I amhaving trouble with…. They have been told to: Use a small PC to acceptup to 4 video signals and act as a streaming web server to stream outof a 10mbps Ethernet. Then use a dongle which will convert Ethernet to12mbps over a single twisted pair and back.> Considering they want to record this to DVD for later viewing I can'tsee this will work too well. All of the camera to Ethernet I have seenoperate like web cams sending jpegs to a web browser. To convert thisto a DVD will take a fair bit of processing on the behalf of a laptop.> The video is for remote diagnostics and so production quality is notnecessary. Also they are locked into some twisted pair they alreadyhave.> Anyone got ideas ??> Liaaam --Chetan BhargavaWeb: http://www.bhargavaz.netBlog: http://microz.blogspot.co

> Intersil has chips to transmit video over CAT-5 > This might be of interest:http://www.intersil.com/products/deviceresults.asp?i=13295 www.analog.com/en/subCat/0,2879,770%255F842%255F0%255F%255F0%255F,00. html "Whether you are sending a signal across the board or pushing it through 1000 meters of twisted pair cable, the signal you send will be the signal you receive using ADI differential drivers and receivers" My neighbour and I used to network over 70m of CAT5. He gave me access to his broadband, I gave him acces to my subscription TV. Which was sent down a spare pair. I know it was done with AD chips (a differential amp and a driver) but can't remember which, and the board eludes me at the moment This looks intriguing - network through house mains. Something I might consider as a Christmas gift to a couple of deserving boys http://www.siliconchip.com.au/cms/A_105317/article.html http://www.netcomm.com.au/index.php