Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[dave w]

 

[dave w]

MD Corie
Testing, one, two, three. Tap, tap , tap...is this thing working? I must not be communicating well because this is getting too complex.

Take a male and female RCA connector and tie the commons togeather. Then wire the center conductors in series with the N.O contacts of the relay. Drive the coil of the relay from the camera powersupply. You could strip back 1/4 inch of insulation from the power leads and solder + wire to one terminal of the coil and the -  wire to the other terminal of the coil. OR if you wanted to modularize to make it easy to repalce a defective rely, you could make  a second short jumper for power,  using proper size male and female coaxial power connectors with the jumper wires also connected to the relay coil terminals. Now insert the coax power connector jumper in series with the X10 power supply. Insert the  RCA jumper  between the X10 adapter cable video line male and your monitor or video input female connector. These "relay modules" would be connected at each camera addressable power supply end. The 16 "relay modules" are simply opening every one of the  video paths, excepting the camera that is powered.    

[MD Corie]

dave w
I'm pretty sure that I understand your proposal, but let me paraphrase to be certain:

Your scheme is to create a one-of-16 "multiplexer" using individual mechanical relays, each activated by a respective camera power supply, so that the video line from only one camera gets connected through to the monitor at a time - well, assuming only one camera gets turned on at a time, of course.  Your intent is to multiplex all the camera lines back near the monitor with these relays.

If this is correct, then I understand (and agree with the theory of) this scheme, but my "problems" arise due to the perhaps special requirements of my application, as well as the layout of my existing cabling installation:

First off, several groups of my cameras are located 100'-300' from the monitor, so they have their signals joined together out there near the cameras, and the merged signals are sent back to the area of the monitor via single lines in Cat-5 extension cables (and this cabling is likely where the reflections occur, that causes the video "ringing").  The upshot of having this layout is that any relays for the individual cameras in each remote group would need to be out there where the signals are joined before entering the extension cable.  This isn't a big hairy deal as far as installation, because the PCBs of the junction boxes could be modified fairly easily (I think) in order to insert some small PCB-mountable relays.  The problem comes in due to the fact that these junction boxes are located where they are exposed to ambient outdoor conditions of temperature and humidity, leading to my concerns over reliability of mechanical relays under the extremes of temperature that may be encountered there.  Consequently some sort of solid-state multiplexing chip(s) might be better suited to these conditions.

The second issue is that even if a multiplexing scheme like this were to be to be employed, it still would not address the potential problem of impedence mis-match in the Cat-5 extension cables, which may be the cause of the video "ringing".  Because some of the solutions for the impedence-matching issue would also eliminate the need for actively "multiplexing" the camera signals, it may be that a relay scheme wouldn't even be needed.  So, the impression I get is that it would be better to tackle the impedence problem first (since it has to be cured in either case), and then add multiplexing if any other video problems still exist after the reflections are subdued.

Are we on the same page yet? 

[dave w]

Yes, I have a better picture now. My mental image was your custom CAT cables brought your 16 video feeds back to a central location where you had all your camera power supplies and where you "Y"ed your video feeds.

So you must be sending X10 PLC commands long distances?

Not sure if a bunch of in-line composite video amps would help or not.

As far as relays, there are RF relays using magnetic reed switches which are very rugged. But their cost (even surplus) would be high and I am sure they are current hogs. I guess solid state mux'es may be the answer but with a lot more engineering. I doubt you can drive a mux off the X10 addressable power supply with out additional regulation and filtering.

Good luck

[MD Corie]

Yes, I have a better picture now. My mental image was your custom CAT cables brought your 16 video feeds back to a central location where you had all your camera power supplies and where you "Y"ed your video feeds.

It might very well be that the system would be "cleaner", signal-wise, if all camera feeds came back separately, but that would require a whole lot more cable runs... on the other hand, having even more un-matched transmission lines that way might serve to compound the reflections problems.  Sort of picking one's poison, I guess.

So you must be sending X10 PLC commands long distances?

Generally, the camera control PLC commands are not long-distance, since all the camera power supplies are located back near the monitor.  The long distance PLC commands are more limited to those from motion sensors and to the lighting that are located out in the remote locations.

Not sure if a bunch of in-line composite video amps would help or not.

I don't know about video amps, but I'm thinking some Cat-5 cable transmitter/receiver sets might help.

As far as relays, there are RF relays using magnetic reed switches which are very rugged. But their cost (even surplus) would be high and I am sure they are current hogs. I guess solid state mux'es may be the answer but with a lot more engineering. I doubt you can drive a mux off the X10 addressable power supply with out additional regulation and filtering.

There used to be some solid-state "relay" chips that were for low-power application; perhaps something like that would work, although the RF issue may come up.  Depending on the specs of muxes, it may be practical to drive them with a simple voltage divider or perhaps a buffer chip, if need be.  That is, a digital gate chip that can be hard-wired for its ON state, and then provide its power supply via the camera's switched power.

Good luck

Thanks... looks like I'll need it!

Going back to my original question, I'm wondering whether there is a simple, "passive" way to provide enough impedence matching to suppress the reflections, "ringing", etc.  The cable-matching chip sets may be the "best" solution, performance-wise, but are kind of problematic due to power source needs, etc. and baluns (that I've found) are all intended for "external" insertion at a "typical" connector - like a BNC or RCA - which isn't much help for a system that uses modular connectors;  I also found none for PCBs.  I'm not knowedgable enough to design a "ballpark" matching solution using passive components, but if something like that is possible, it is probably worth a try here, before getting too fancy.

[dave w]

Generally, the camera control PLC commands are not long-distance, since all the camera power supplies are located back near the monitor. 

OK I'm confused again. If your camera power supplies are at one location near the monitor then your video breakouts are located there also and all your CAT 5cables. Otherwise you can't be turning on your wired cameras individually. You are controlling your cameras individually right?

Going back to my original question, I'm wondering whether there is a simple, "passive" way to provide enough impedence matching to suppress the reflections, "ringing", etc. 

Going back to my original answer. No. No amount of impedance matching will address the fact that you have the outputs of all 16 cameras wired in parallel through all your "Y"s.

 You have 16, 75 ohm resistors in all in parallel. Your monitor is looking at a whats left of a video signal across 5 ohms instead of 75 ohms. You need to isolate those 16 lines OR seriously boost the signal.

[MD Corie]

Generally, the camera control PLC commands are not long-distance, since all the camera power supplies are located back near the monitor. 

OK I'm confused again. If your camera power supplies are at one location near the monitor then your video breakouts are located there also and all your CAT 5cables. Otherwise you can't be turning on your wired cameras individually. You are controlling your cameras individually right?

I'm confused at your confusion...   Maybe I'm misunderstanding the term, "video breakouts", but let me try to clarify what I think you are asking:  The camera power supplies are all located near the same physical spot where all the "local" camera cables and the "near end" of the Cat-5 extension cables for the "remote" groups of cameras come together and are merged -via several "breakout boxes"(*)- into a single video and single audio output feed to the TV monitor.  So, most of the signal joining takes place at this "central office".  However, the Cat-5 extension cables serve three to five remote cameras each, and so the signals from the remote cameras in each group are joined together at the remote location (i.e. - at the other end of the respective Cat-5 cable) via "breakout boxes"(*) into single video and audio lines coming back over the Cat-5 cables.  And yes, all individual cameras are controlled individually.

(*) Note:  "Breakout boxes" would more correctly be called "merging boxes" as they function in this layout.

Going back to my original question, I'm wondering whether there is a simple, "passive" way to provide enough impedance matching to suppress the reflections, "ringing", etc. 

Going back to my original answer. No. No amount of impedance matching will address the fact that you have the outputs of all 16 cameras wired in parallel through all your "Y"s.

 You have 16, 75 ohm resistors in all in parallel. Your monitor is looking at a whats left of a video signal across 5 ohms instead of 75 ohms. You need to isolate those 16 lines OR seriously boost the signal.

OK, I'm no RF engineer (hence all my questions about impedance matching).  But, I'm not sure we're talking about the same issue here.  My understanding -such as it is- is that video "ringing" is caused by "standing waves" on "transmission lines", that occur as a result of portions of the video signal being reflected (perhaps repeatedly) due to the incomplete transfer of signal from the transmission line to the load - which, in turn, is caused by the load impedance not matching the "characteristic impedance" of the transmission line.  This is the impedance mismatch that I'm seeking to address.  I understand that traditonally this matching is accomplished by balun devices at one or both ends of the transmission line (and at each merged input, in the case of signal-joining).  Because I'm not finding any suitable balun devices, and the "active" Cat-5 matching chips or compensator chips would require some sort of power supply, I'm hoping there might be some other passive component(s) that would work instead.

That said, I want to emphasize that the problem I'm seeing is with video "ringing";  I don't see any other overt evidence of signal loss - such as weak/"snowy" pictues - that would indicate any need for signal boosting, per se, because the images with all cameras merged is comparable to the image from any camera when it is directly connected by itself - except for the presence of the "ringing".  That's why I'm uncertain that there is any need for signal multiplexing with the intent of avoiding direct signal loss.

So, that's my "take" on the situation, FWIW...

[dave w]

The camera power supplies are all located near the same physical spot where all the "local" camera cables and the "near end" of the Cat-5 extension cables for the "remote" groups of cameras come together and are merged -via several "breakout boxes"(*)- into a single video and single audio output feed to the TV monitor.  So, most of the signal joining takes place at this "central office". 

You just stated all camera power supplies are located in one area, so no, we actually were not on the same page in previous post.

There is a simple test to see if 16 relays to open the video paths will clear your picture. In your breakout boxes, disconnect 15 camera RCA video connectors from your merging "Y" cables. Leave ONLY one camera connected, preferably the one with the longest CAT 5 cable. Now turn that camera on .

Is the monitor picture clearer???

If so, the 16 relays will fix your poor picture problem. They should be inserted between the male RCA connector on the X10 breakout cable and the female RCA connection on the "Y" cable (in the breakout boxes?). Each relay N.O. contacts opens or connects the video path from one camera. The relay coil is powered by the respective X10 camera power supply. So if only one camera is on, then the monitor will only be coupled to one camera, not 16 as you have it now. The remaining 15 "Y"s are all open when only one camera is powered so their infleunce on the video signal is minimal.

[MD Corie]

You just stated all camera power supplies are located in one area, so no, we actually were not on the same page in previous post.

Clearly, something is getting lost in the translation... although it is unclear to me just what that may be.  As I mentioned, all of the camera power supplies are located together at the monitor end, while some of the camera signals are merged at the far end of the extension cables (and, consequently, any relays for those cameras would have to be inserted at that remote end of the cable).  I guess I'm not seeing the problem...

There is a simple test to see if 16 relays to open the video paths will clear your picture. In your breakout boxes, disconnect 15 camera RCA video connectors from your merging "Y" cables. Leave ONLY one camera connected, preferably the one with the longest CAT 5 cable. Now turn that camera on .

Is the monitor picture clearer???

I thought that I already mentioned that the image is slightly - but insignificantly - better when a given camera is connected directly to the monitor (with no other cameras or cables involved)... and the main difference between a single camera being connected directly versus having a Cat-5 extension cable in its individual path is that video "ringing" is present with the Cat-5 extension cable in place - as is the case when the given camera is installed in its operational location.

If so, the 16 relays will fix your poor picture problem. They should be inserted between the male RCA connector on the X10 breakout cable and the female RCA connection on the "Y" cable (in the breakout boxes?). Each relay N.O. contacts opens or connects the video path from one camera. The relay coil is powered by the respective X10 camera power supply. So if only one camera is on, then the monitor will only be coupled to one camera, not 16 as you have it now. The remaining 15 "Y"s are all open when only one camera is powered so their infleunce on the video signal is minimal.

I understand how multiplexing relays would need to be connected in my system, and just to avoid further confusion over a non-existent wiring situation, please be aware that the only RCA connectors that exist in my system are where the monitor connects to the final junction box... and there are no RCA Y-cables at all.  The signal merging is done within the junction boxes, and the cameras connect to the junction boxes via their modular 6P6C connectors, and the extension cables connect to the junction boxes via the usual LAN-type 8P8C modular connectors.  The camera power supplies plug into barrel jacks on the junction boxes that are located at the monitor end of the network - although the junction boxes that are used at the remote end also can accept power supply inputs - just in case circumstances force the power supplies for those remote cameras to be placed at the remote location.  (So far, that remote placement of the camera power supplies has not been necesary).

Bottom line is that I know how to incorporate multiplexing relays into my system if needed, but I'm not yet seeing evidence that they are really needed.

[Noam]

I've been following this thread, and I'm pretty confused myself.
From what I gather, you created your own wiring to extend the wired camera signals, as well as to combine the video feeds from 16 cameras into one. You did this by joining them together at several places, so that the single combined feed comes together at one point near where the power supplies connect.

How long is the run to the farthest camera?
You indicated (unless I read it incorrectly), that the picture is better when you don't use your home-made extension cables.

Perhaps the signal degradation is due to a combination of the distance, and interference / crosstalk within your wiring?

I had a similar situation a few years ago when I tried to extend a wired camera (not X10 - some no-name from Home Depot) using CAT5.
Even though the instructions said you could use standard quad phone wiring to extend the signals, I found that once I did that (even by 20 feet), the signal degraded heavily. Using CAT5 for the extension didn't help any.
I ended up re-routing the wiring for one camera so that it only used the original 50-foot cable (I had to cut it and crimp on a new connector to run it, extending that with a short barrel connector was okay), and I was unable to use the second camera (the run was just too long).

[dave w]

I thought that I already mentioned that the image is slightly - but insignificantly - better when a given camera is connected directly to the monitor (with no other cameras or cables involved)... Bottom line is that I know how to incorporate multiplexing relays into my system if needed, but I'm not yet seeing evidence that they are really needed.

I must have missed where you said you had only one camera plugged in to the monitor with the rest of the system completely disconnected, and the picture was no better. That definately says relays to isolate lines will not help, but I don't understand what the problem is. As long as you are using twisted pair CAT5 and not the flat junk, it should be able to go 100 feet with out appreciable signal loss, although it won't have the interference immunity that coax might. But you are not describing interference. Baluns might help with the ringing, but "foggy" low contrast picture is a "weak signal" description to me. I doubt the X10 power supply has the oomph to drive a video amp at the head end either. I'm at a loss. Sorry.

[MD Corie]

I've been following this thread, and I'm pretty confused myself. ...

Here are some answers and perhaps some clarifications:

I have a combination of cameras having only their own pre-connected 60' cables, and two Cat-5 extension cable runs (well, actually three if I count a short 5' extension that was needed to make a pair of "local" cameras reach the "central office").  The camera cables all plug into custom junction boxes, that are located either in the "central office" or in the case of the extension cable runs, at the remote ends of the respective runs.  Each junction box serves both to combine the video (and audio) signals from the cameras and to route the camera power supplies to the respective cameras.

I don't know exactly how far the total run is to the farthest camera, but it would be approximately 150' of extension cable plus the 60' of camera cable.

The video image is "better" when a camera is connected individually to the monitor than when it is combined with other cameras... and it is better than when an extension cable is added (although the extension cable more adds "ringing" effects than otherwise degades the picture).  To try to put these situations into perspective, on a scale of 0 to 10, where zero is no video signal and 10 is a crystal clear image, the cameras provide an image of about 5 when connected individually and directly to the monitor (or to the PC); when they are connected with other cameras, but no extension cables, the rating would be about a 4; when extension cables are added to the mix, the rating would be about 3.5 (because the basic image quality does not degrade, but the effects of "ringing" make the picture appear to be a bit more "out of focus".  The point is, the image is poor to begin with, so any added degradation due to "Y-ing" or extending the cables is not really noticable except upon direct comparison (i.e. - plug/un-plug while watching).

I am confident that the added degradation is due to the combination of several factors, as you mentioned.

For reference, I tested a couple of cameras on the full spool of Cat-5e cable (1000') before I ever installed anything, and found that the image quality was only slightly diminished by the insertion of the 1000' of spooled cable... so I don't think that extension cable lengths of "only" 150' are likely to be causing the bulk of the problems.

[MD Corie]

I thought that I already mentioned that the image is slightly - but insignificantly - better when a given camera is connected directly to the monitor (with no other cameras or cables involved)... Bottom line is that I know how to incorporate multiplexing relays into my system if needed, but I'm not yet seeing evidence that they are really needed.

I must have missed where you said you had only one camera plugged in to the monitor with the rest of the system completely disconnected, and the picture was no better. That definately says relays to isolate lines will not help, but I don't understand what the problem is. As long as you are using twisted pair CAT5 and not the flat junk, it should be able to go 100 feet with out appreciable signal loss, although it won't have the interference immunity that coax might. But you are not describing interference. Baluns might help with the ringing, but "foggy" low contrast picture is a "weak signal" description to me. I doubt the X10 power supply has the oomph to drive a video amp at the head end either. I'm at a loss. Sorry.

The camera images are only a little bit better when the cameras are connected individually to the monitor than when they are connected as part of the networked system; they are clearer, and don't have the ringing effect.  However, even when the cameras are connected individually to the monitor of PC, the images seem to be out-of-focus and/or "foggy", so as-is, I'd have to say the cameras themselves contribute the bulk of the problem.  Nevertheless, the "ringing" causes enough of a detriment that I think it would help a lot if I could eliminate it... hence my question about ways to address that problem.

FWIW, I have tested the camera supply voltage with the camera connected locally versus remotely, and it appears there is only about a 0.2v - 0.35v drop when the camera is attached at the other end of the extension cable.  I don't know what the voltage tolerance is for the cameras, nor how much current they draw, but I would suspect that loading the (unregulated) camera supplies with other devices might cause some image degradation - or even loss of image.

[Noam]

I'm no expert here, but I'm wondering if part of the signal problems (degradation and/or "ringing") might be due to crosstalk from the power supplies over a long cable run. Since the cables are not shielded, other electrical noise along the way can also degrade or interfere with the camera image. I had this happen myself. Using a stock wired camera (not X10 - some other brand from Home Depot), the image was perfect when I tested it. However, when  I fished the wire through the walls for the final installation, the image degraded severely. Along the way, it passed near electrical wiring, copper pipes, ductwork, and several florescent lights. The image was unusable for me, so I had to give up on that camera. I pulled it out, and tested it again (with the wire running loose along the floor in the hallway), and the image was clear again (so I know I didn't break anything when I installed it).

If the "stock" camera, connected directly to the monitor (with NO home-made cabling or junction boxes) does not produce a perfect image, then perhaps there is something wrong with the camera. Have you done this test with more than one of your cameras?

I don't know if it is the same for the wired cameras, but I seem to recall with my wireless ones (which I sold, so I can't check them anymore) that the lens rotated to help focus the image. Perhaps yours are actually out of focus, and that might be adding to the problem.

It sounds to me like you probably have a number of factors working together to give you a less-than-stellar image at the end. There may be no single fix that will take care of it all.

[dave w]

Noam's quick fingers beat me to this and we are  thinking along the same lines, so I will post anyway

The ringing is likely caused by multiple RJ45 connections. Each conection can cause a reflection, especially if there is a layer of oxidation on the pins. The "foggyness" is loss of high frequency content in the video signal (edge definition). Also the flat cable is much more prone to interference and cross talk than twisted pair. If your flat cables are run by themselves and not near the other flat cables, Romex, or flourescent lighting fixtures, etc. you probably do not have much induced interference. Long runs of flat cable is not good for video, but it is cheap.