Soundy

Yeah, managed switches are great - we've used several of the SFE1000P, which is similar to the SRW2008P. Not only can you monitor the power usage on each camera, you can toggle the power remotely, in case a troublesome cam needs a quick power cycle

Why are you using an S-video cable? That camera doesn't output S-video (Y/C), only composite (CVBS). What capture device are you using to plug it into your laptop?

It's really nothing magical or super-hi-tech. In its simplest form, it's just using the two otherwise-unused pairs in the UTP cable (standard 10/100 ethernet uses only the orange and green pairs). Spec says 44VDC but devices must be able to accept anything up to 48V or more. In fancier incarnations, the power source (switch or injector) can detect whether there's a PoE load attached and turn the PoE on when appropriate, and beyond that, some devices will report how much power they need, allowing a similarly compliant switch to manage how much power is allocated to that port. *Complete* PoE spec also allows "phantom power" to be sent over the data pairs (also necessary for use with gigabit, which uses all four pairs itself), which means load devices must be able to ACCEPT power on any of the pairs. There really a lot of safety built into the spec to protect devices. 12W *should* be referring to the *maximum* required. Panasonic being a reputable brand, this should apply. Consider your average non-IR dome camera only draws 2-4W. Yes, although you could only add one more PTZ like that. 802.3af specifies that ports should be able to provide up to 15.4V; however, it doesn't say that they have to ALL do that ALL the time. In this case, you have a maximum of 30W available to all the PoE ports. Of course. Most of them simply use an external adapter that outputs 48VDC. That's only because the filtering in a surge suppressor or UPS will interfere with the data signal that it's modulating on the power line. That's not a factor with the switch.

No reason it shouldn't work fine, just remember that the PoE injector goes between the remote powerline adapter, and the camera. As far as switches and injectors, I've used a fairly wide range with no problems, from cheap $15 DealExtreme injectors to enterprise-grade switches.

That's possible as well, if the picture is still bad with only one of the affected cameras on the DVR.

The iris, specifically, is the set of blades that forms an adjustable opening. That opening, the "hole" inside the lens formed by the iris, is the aperture. If it's a fixed-aperture lens (like most M12 board lenses), then yes. If it's a manual-iris lens, you'll need to make sure it's wide-open, but that's done with a ring on the lens (usually marked "open" and "closed"). Generally, no, you won't need to. The exception might be if the camera's ESC (electronic shutter) or gain control can't compensate for a very bright scene - for example, if the image is washed out with the iris wide open. HOWEVER... if you need to close down the iris to compensate for that in a normal view, then you'll probably need to LEAVE the iris at that partially-closed position for regular operation, and being a manual iris, you don't have to worry about it opening up again on its own and thus affecting the DOF. Summarizing: this problem is specific to auto-iris lenses being partially closed when focusing, and then losing DOF when they open again in low light. The ND filter is intended to reduce the light and force the auto iris to open fully thus giving you the least DOF while focusing, knowing that the iris closing down later will only increase the DOF. Manual-iris and fixed-aperture lenses should not require ND filters to set up.

Laser pointer DIY would probably be your best bet. Most sensors use simple IR LED setups with a simple lens to tighten the beam to cover 15-20'. You're talking about probably 80-100' by the time you get across the road and back. Try this page for a start: http://www.electro-tech-online.com/electronic-projects-design-ideas-reviews/33205-laser-beam-break-detector.html

I don't think you'll have a lot of luck here - any such adapter would probably need to be an optical type (like, with a lens of its own) as the physical structure of such a thing would move the lens too far from the sensor to be able to focus. It would probably be easier to simply find a fixed-iris longer board lens - I have both 16mm and 25mm lenses in use on board cameras. It looks like this is in a shaded area so you should be able to get away with not having auto-iris.

The baluns are not "at fault", really... the ground loop is created because each camera has multiple signal-ground paths through the other cameras' power grounds, and they're different DC resistances because the baluns are essentially coils inline with the other signal paths. You don't get problems on the TV because you're only connecting one camera to the TV, so that camera only has a single signal-ground path. If you disconnect two of the three cameras from the DVR, the picture on the remaining one should be crystal-clear as well. Reconnect a second one, you'll probably see a bit of noise; reconnect the third and the noise becomes worse still. I've been meaning for months to sketch up a diagram of what's happening to cause this... really should get that done. Ultimately, the "fault" lies in a cheap camera design that uses a shared ground for video and power. Cameras that physically separate the power ground via transformer or regulator don't have this problem.

Ground loop caused by using baluns with cheap 12V cameras that have common video and power grounds. Put each camera on a separate power adapter and the problem will go away.

I can't see any TECHNICAL reason that wouldn't work. You might need to have it trip a relay, and have that switch the alarm inputs on the DVR... From a logistical angle, keep in mind that most of these actually have a beam on one side and sensor on the other, so you'd need to get a wire to the far side of the road... plus, the beam might become too diffused at that distance, especially if you're bouncing it back from the other side of the road. There ARE designs that would work, designed for longer distances, for this but I don't think a standard off-the-shelf garage door sensor is suitable.

Your RG59 should not terminate with a balun. It should terminate with a BNC that connects to the balun on the UTP... like this: Make sure you're only using a single pair of the Cat5 into the baluns, too - for example, white/blue to +, blue to - .

You're just begging for impedence-related problems doing that. The RG59 will still be 75 ohms, while the TP side of the baluns are at 100 ohms (if I understand correctly here, you're talking about connecting the RG59 to the TP side of the baluns).

The idea of an ND filter is to force an auto-iris camera to open the iris all the way; otherwise, a partially-closed iris (due to bright daytime light) will cause a greater depth-of-field, and if your focus point isn't set right, your image will go out of focus when it gets dark and the iris opens up again. With your setup, it shouldn't be necessary: just set the manual iris to full-open when doing your initial focusing, then if you want, try closing it down a bit to increase the DOF and sharpness. If you do use one, you'll normally just hold it in front of the lens while focusing. Some guys use actual photographic ND filters; some use a piece of welder's glass (#4 or #5 shade). Ultimately, you could do it with a pair of sunglasses, in a pinch - the purpose is to reduce the light and force the iris to open fully.

With an older camera, it's also possible that the CCD is, for lack of a better term, worn out... and it's possible the lens itself has become defective over time (excessive heat causing plastic parts to soften, moisture or contaminants getting inside the lens, etc.). One thing you can do is take the lens off and ensure both front and rear elements are clean (because the rear element is so near the focal point, the smallest imperfection there can cause a significant image problem). Look at the sensor as well, and make sure it's clean (I've seen them muddled by fingerprints, dust, even some kind of substance that seemed to be vegetable oil). If that doesn't help, try a different lens, to narrow down whether the problem is the camera, or the lens itself.

I'd probably recommend the analog+encoder route using coax in the traveler, especially if that's what you're already familiar with, so you know the right type of cable to use for longevity, the maintenance issues, etc... Then in the future, if they do decide they want MP cameras in the elevators, you can easily retrofit ethernet-over-coax adapters without having to worry about the distance. You'll already have a network drop at the other end for the encoders, so it should be a simple matter to swap over.

Are you sure they're baluns and not just adapters? We've seen this question a few times recently here, straight-through adapters being sold as baluns, but without that balun function, you don't get the proper line balancing or noise rejection. Is the coax proper solid-copper-core, braided copper shield?

You need some sort of receiver or converter that can downsample the signal. Don't expect it to be particularly cheap. Does the camera not also output regular composite video?

^Yes, but so is an encoder, so... no loss there.

If you're talking about the UTP cable... you're right, it LOOKS like a balun, but it's unclear if that's actually the case. It's optional, either way, so you won't find it bundled with the camera. Anyway, in terms of buying one, how about one from this page or this page? Is there a better place to buy from? Never dealt with either of those sellers, so I can't give you any direction there... however, we use this style (GEM branded) with the VCM-24VFs, and they do fit perfectly in the back-box; in fact, the balun and the BNC connector just wedge nice and snug in the back-box to hold the wires in place while you mount it.

They're actually not that bad if you flush-mount them...

You're not going to get wider than 90 degrees with most cameras without going to specialized lenses - 2.8mm on a 1/3" sensor will get you about 88-89 degrees on the horizontal. With careful positioning though, you can probably get by with that. Just use the same VCM-24VFs indoors.

Zero chance of problem. And the others are correct, the cameras DO NOT come with baluns... fortunately there are lots of options out there, but as we've seen in other recent threads, there are some shady online sellers hawking straight-through connectors as baluns, so be sure to check in here before you actually buy something if you're not sure

Analog and encoders will probably be your best bet, especially if you're already familiar with using analog cams in elevators. There are several options out there - Avigilon has a four-channel encoder; Axis makes models ranging from single-channel to 1U 16-channel units to rack-mountable blade frames for plug-in blade modules: http://www.axis.com/products/video/video_server/index.htm As far as wireless, I think reliability is a far greater concern than security (be it hacking OR jamming). There was a discussion on this several months ago, and one of the more promising methods mentioned was a laser-based system... but ultimately, ALL elevator cars will already have cables going to them, and frankly, you don't even need coax - you can use pretty much any available pair of wires with passive baluns for an analog signal.

Welcome!