Soundy

Simple: a cost-saving scheme. It's cheaper to make two different cards using the same basic hardware (board, connectors, etc.) than to have two different designs. Either a second 8-channel card, or upgrade to a 16-channel card...

If the DVR has RS485 connections, then yes.

It shouldn't cost any more to reinstall the software - normally you'd fill out the form on the GV website, provide the serial number of the card, and they'll give you a link to download the software.

Either - both are powering all the cameras off a common power and ground rail. Having all the camera power grounds tied together like that can create ground loop problems in certain circumstances. It makes for a cleaner install... that's about it. Read this thread for more info on the ground loop issue: viewtopic.php?f=6&t=30871&hilit=+ground+loop

You could do that, but using multiple pairs should alleviate the problem. Using better cameras would as well, addressing both this issue AND the problem with the waving lines (sounds like a ground loop, common when using low-grade cameras with baluns and a shared power supply).

As far as low-cost equipment and standalone DVRs go, Dahua makes some really good stuff. Their DVRs have an easy-to-use interface, they have a good web client as well as a "thick" client, and the mobile apps work quite well. They're sold under a variety of different brand names in North America; there's a VERY partial list here: viewtopic.php?f=53&t=31629 Also suggested for cameras are CNB's "Monalisa" series - they work great at night with very low light. Be sure to check CNB's website (www.cnbusa.com) for a list of unauthorized resellers, though - there are a number of online stores that are not authorized to sell them and buying from those will void any factory warranty coverage.

That must be with the IR off - that's only 200mA. Most IR cameras will pull anywhere between 500mA to a full amp when the IRs come on. That's likely your problem. Assuming a camera pulls 500mA with IR on, at 90m (about 300'), you're going to see a loss of about 3.8V(!!!), so assuming a regulated 12V supply, the camera is only getting 8.2V. Even at 200mA, that's 1.5V loss, which may be borderline. Step it up to use two pairs for power, and your loss at 90m and 200mA is only 0.7V... 3 pairs, and the loss is only 0.3V.

How many pairs of the Cat5 are used for power, and how much current do the cameras draw? Sounds typical of voltage loss caused by too much current draw over too-small wire. If that's the case, replacing the power supply won't fix it, as it's not a matter of how much current is available at the source.

As Shockwave indicates, it will really vary depending on your actual needs, but anticipate $100-$150 per camera, to start with... a "decent" four-channel DVR can run anywhere from $100 to $400 depending on what features and performance you need. Any *good* online store will have a phone number, email address, or some other sort of contact that will let you communicate with a real person. If it doesn't avoid it. If it does and you get no response from the contact info, avoid it. Even $550 for all that indicates low-quality equipment. You should expect to pay $400 or more JUST for a decent PTZ, never mind the rest of it. (Hmmm, looking up that unit... it's not a PTZ, just a PT - fixed lens. Not terribly useful. The specs also list D1 recording, and 240fps total on another line... but doesn't specifically state that it does both at the same time, which given the cost would lead me to think it probably only does 240fps at CIF resolution (352x240), or 7fps at D1.

As long as the same camera connects to the same input... yes. There's no FUNCTIONAL difference between this and individual coax runs. As far as use with video baluns... none. Baluns don't REQUIRE UTP; I've used them with alarm wire, speaker wire, microphone wire... four-pair UTP like Cat5e and Cat6 is more commonly used just because they're a more common wire, and because it allows for easy upgrade to network later. The twists in Cat3 and other such cables are also designed to reject outside interference... the specific twists in Cat5e and up - each pair different from the others - is designed to reduce cross-talk between pairs (signal induced in one pair from its adjacent pairs). Either type will work fine for video over baluns.

I don't think there's one central site for pricing these things... you could try eBay to see what similar items are selling for, but you'll need to know the brand and model number of everything.

Uhh... copper?

Ah, that makes sense. Well, hard to say without seeing the actual circuit boards, but yeah, it sounds like that cable is the main connection between this interface board and the camera, and the black wire would most likely be something power-related, whether general camera power or just for the fan. Small fan is PROBABLY just a cooling fan, rather than something for circulating air to keep the dome clear. Either way, there should be no difference in the ratings you need to look for - you want a transformer that's 24VAC output, minimum 40VA output capacity. Only difference between you and someone in North America is that you'll be looking for something with a 240VAC input, rather than 120VAC.

I'm running two MP cameras over a WiFi link - one IQ511 (1.3MP) and one IQ755 (5MP), using a pair of ASUS routers and a WDS link courtesy of DD-WRT. It's relatively stable, although I don't know if I'd use it for a customer's site, at least not as such. And of course, there are plenty of HTPCs and streaming media players that regularly stream HD video content via WiFi. I think the warning against "HD and WiFi" is probably more related to cameras with weak built-in WiFi.

This is exactly the problem. The best solution is to tap the video signal BEFORE the modulator (assuming it's not built into the camera, which is unlikely). For that price, you could have just replaced the camera with something that gives you a standard composite-video output, or with an IP camera that you could access directly from your smartphone. For that matter, for the cost of the Axis encoder and the demodulator, you probably could have just added a megapixel IP camera to access directly from your smartphone.

The labelling on the unit is pretty clear: the red connector is for 24VAC power, the black connector for the serial connection. As to why there are two wires on each of the red terminals, I'd guess that's just done to double the current capacity, since that appears to be fairly small wire. Id' say your power supply should be at least a 40VA rating, more if the PTZ has a heater/blower. What's odd is the connection to the camera: a single pair to carry video, power and serial??

If it's a real GeoVision card, you can get the software from GV directly, but you have to provide the serial number that will be on a sticker on the card.

Something like this? Again, if you're going to continue using the card with it, you'll have to make sure there are Win7 drivers available.

You're right, doesn't sound like this guy has a clue. He is right on one point: a lot of cards, especially cheap cards, don't have proper Win7 driver support, so they won't work properly with anything but XP. If there are Win7 drivers though, there's no reason they shouldn't work as well. As for memory, there's ZERO reason it should be more or less stable just because you have 4GB vs 2GB memory. Instability COULD be caused by bad or faulty RAM, but not by the amount itself. What card and software are you using? Symptoms like these are common when using cloned/grey-market GeoVision cards, since the drivers and software must be hacked to work with the fake cards. I expect it's a common occurrence with other illegitimate cards as well. He's out to lunch. GeoVision is a decent DIY system, but beware of cheap cards from online/eBay sellers - there are a LOT of fake ones out there. A legitimate seller should be able to provide you a serial number of the card you're buying. If you look up "GV800" on eBay, for example, you'll find tons of listings for <$150, and several for around $500-$600... pretty much guarantee the $150 ones are fake and will very likely give you problems.

That doesn't mean it's a camera... my bet would be on some sort of PIR motion detector.

What makes you think it's a real camera?

Do you have the ability to test an Axis P3354/64 in that scene? It's a dome, not WDR but just as light sensitive as the Q1604 and produces a very sharp image. If you can provide one, they'd probably be glad to have me test it - they know they have harsh lighting to deal with, so they like to know what their available options are. Light sensitivity is only half the equation though - as you see from the clips, dynamic range is a major issue on these sites.

Thanks Matt, but I was more looking towards the WDR performance in a hostel foyer. It will be pointed straight at a door to outside. I do understand that CCD > CMOS for WDR but I hope to use a MP solution rather than analog. As indicated by the title of that thread... three of the four cameras ARE megapixel, including the Axis that blows the others away. It's not "backlit" like you'd have in a doorway, but you get a good idea of how it handles the extreme dynamic range of that situation: dark floor, dark fixtures, little pin-spots above each of the tables... from a "sensor" standpoint, there's little difference between the lounge scene and a backlit doorway.

Optocoupler sample: http://www.technogumbo.com/projects/Learning-to-Use-an-Optocoupler-or-OptoIsolator/