Subscriber Discussion

Ethernet Or Coaxial Cable For Analog Cameras?

UD
Undisclosed Distributor #1
May 31, 2016

Hi everyone,

I had seen in many places, people use Ethernet cable for analog cameras because we can use power and video through single cable, instead of two cables as in the case (coaxial cable & Power cable) as most people use for analog cameras.

So is it better to go for ethernet cable or coaxial cable for analog cameras ? Any advantages or disadvantages ??

Avatar
Mark Kirkeby
May 31, 2016

In my opinion,

As a rule many years ago I quit running coax for cameras, with Cat5 and a dirt cheap balun I haave analog video , and the installation is future proofed that can be upgraded to IP camera without any infrastructure issues.

HD over coax has made it capable now to upgrade on old infrastructure, but unless you are going to base your future business model on HDTVI/HDCVI style architecture there is no reason to run coax.

Ethernet cable is your answer

(2)
Avatar
Gert Molkens
May 31, 2016
IPVMU Certified

Agree with Mark: its easier, cheaper and future proof tu use twisted pair instead of coax

MI
Matt Ion
May 31, 2016

We stopped using coax many years ago as well in favour of Cat5e. The cable itself is cheaper, even if you don't include the cost of separate power wire necessary with analog. It's also more flexible and easier to work with, and typically smaller diameter so you can fit more runs in a give space.

When we were still doing a mix of analog and IP on large sites, we also made use of these - they're a combined unit, rackmount multi-channel balun and 12VDC power source. The RJ45 jacks made it a snap to simply terminate all our runs into a patch panel, then use a regular patch cord to connect to the VPS for analog, or a switch for IP. Upgrading later is as simple as unplugging the patch from the VPS, swapping the camera, then patching into the switch.

Each group of four inputs also includes an RJ45 output carrying all four video signals over a single Cat5e - on a couple of sites we used this feature along with a rackmount multi-channel balun in the rolling rack, to allow us to terminate runs in a wall rack, then cross-connect to the rolling rack with minimal cables; we then had the rackmount baluns on the rear rails directly behind the DVRs, making for easy, direct connections to the DVR input tails.

But enough of that sales pitch (disclaimer: I do not work for Eastern CCTV and never have, but they were our supplier for the VPS units and always gave great pricing and service). We're not installing analog any more, and most sites, when an analog camera fails it gets replaced with IP. As Mark says, having used Cat5e makes the upgrading painless.

Using UTP instead of coax has also allowed me to do some sketchy stuff out of pure necessity... for example, one site was near completion and they wanted us to add a camera to an outdoor location, but all exit points were already sealed... I took the Cat5e going to the nearest outdoor camera and split the pairs off, orange for one camera's video and green for its power; blue for the other camera's video and brown for its power. Fortunately the cameras require very low current and have a wide range of operating voltage, so even with the voltage drop over some 200' of 24ga., both cameras worked just fine. A couple years later one failed and they wanted to replace it with an IP camera... fortunately we were using a good PoE switch that supported mode B PoE (phantom power over the data pairs - most cheap PoE switches don't do this, sending power only over the other two unused pairs), and for the last two, almost three years, I've had one analog and one IP camera running over a single Cat5e. Best practices? Not even close. But it works and that's all that matters.

U
Undisclosed #2
May 31, 2016
IPVMU Certified

Good info, Matt!

Do you ever see any degradation to the video signal due to the baluns?

...a good PoE switch that supported mode B PoE (phantom power over the data pairs...

This might be a typo, since mode B PoE is phantom power over the spare pairs.

MI
Matt Ion
Jun 01, 2016

This might be a typo, since mode B PoE is phantom power over the spare pairs.

Oops, right you are - got my modes reversed. Great explanation here: http://duncansonline.ca/FAQs/WhatisPoEModeAModeB.htm

Do you ever see any degradation to the video signal due to the baluns?

None. In fact, between the twisted pair's inherent noise rejection, and the balanced line, i find it generally cleaner than coax. I also discovered - again, out of sheer necessity - that if you need to split a camera output (for example, to feed a monitor), doing it AFTER the balun results in little or no noticeable loss... vs. a very obvious image degradation caused by putting a T-splitter in a coax line. Example: camera -> balun -> two twisted pairs into balun terminals, one to the balun at the head end, one to a balun at the monitor. I BELIEVE (never really delved into it) this is because the camera is still seeing a proper load, so you're not "bogging down" its output.

I should add one caveat here: there's a substantial risk of ground loops if you're using cheap 12V cameras that internally have a shared power and video ground. This setup creates two vastly different ground paths to the camera: one short, direct one via the power line to a common ground at the power can; and one very long one with high DC resistance, passing through the coils of the two baluns, back to the common signal ground on the DVR. With one camera it's not a problem, but as you start adding more, you get nastiness. The best solution to this is to use dual-voltage (12VDC/24VAC) cameras, or any cameras that have their own internal regulator, which isolates the power ground from the signal ground.

(1)
U
Undisclosed #2
Jun 01, 2016
IPVMU Certified

I should add one caveat here: there's a substantial risk of ground loops if you're using cheap 12V cameras that internally have a shared power and video ground. This setup creates two vastly different ground paths to the camera: one short, direct one via the power line to a common ground at the power can; and one very long one with high DC resistance, passing through the coils of the two baluns, back to the common signal ground on the DVR.

Coils don't create ground loops, they eliminate them. However, what you describe is correct, the coils create a very high DC resistance to the power supplies ground current thru the transmission line. This is what you want.

What you don't want is when there is a *low* DC resistance ground path to the camera thru the transmission line, for instance without a balun, where the chassis ground of the camera is locally earthed.

The fact that you don't see the effect with one camera is indicative of this not being a ground loop issue. IMHO, the increasing 'nastiness' you see as you add more cameras is more likely the result of unintentional radiation from the coils at the source, due to their proximity, which is an additive effect.

MI
Matt Ion
Jun 01, 2016

The problem is that you get a number of different ground paths at widely varying resistance, that just get worse as you add cameras. Keep in mind that all the signal grounds for all the cameras connect together at the DVR, and all the power grounds are tied together at the power supply.

Using cameras with a shared ground (as most cheap 12VDC cameras are), with one camera, the two grounds are separated at the other end, and each has its own path. Add a second camera, and each now has a path to the DVR via its own signal line, as well as via its power line, to the power supply, back to the second camera, and then back to the DVR via that camera's signal ground. So now each camera's signal has a ground path of camera -> DVR, as well as camera -> PSU -> other camera -> DVR.

With coax, the resistance in the signal ground is extremely low, and doesn't (usually) create a problem, at least not as long as you're using good cable with braided copper shield. With baluns, you're adding up the DC resistance of baluns at both ends of every camera run, and the more cameras you add, the worse it gets.

If the camera has an internal regulator or rectifier (a necessity in dual-voltage models), the power and signal grounds are separated at the camera, so you don't get the multiple paths.

If it was just a matter of multiple baluns inducing EMI in each other where they're clustered behind the DVR, you'd get bad noise even on AC-powered cameras.

U
Undisclosed #2
Jun 01, 2016
IPVMU Certified

Just to be clear, I'm not saying that multiple devices with different ground potentials isn't problematic, it is.

I'm saying that the addition of baluns, which among other things act like isolation transformers, does not cause (or make worse) ground loops.

With coax, the resistance in the signal ground is extremely low, and doesn't (usually) create a problem, at least not as long as you're using good cable with braided copper shield. With baluns, you're adding up the DC resistance of baluns at both ends of every camera run, and the more cameras you add, the worse it gets.

Look, in a perfect world, you DON'T want DC from the power supply traveling in the coax. It COULD be ok, but only if the grounds are truly the same, which the more physically separated they are, the more likely they are not.

Its the DC in the coax that gives you the artifacting you see in the screen.

You are literally looking at the current flowing between the two different ground potentials.

If what you were saying was true then there would not be any ground loops with one camera and recorder.

Also, what do you think isolation transformers do then?

MI
Matt Ion
Jun 03, 2016

Also, what do you think isolation transformers do then?

Video baluns are not the same as audio baluns. Audio baluns ARE isolation transformers. Your typical video balun is not.

Typical audio balun schematic (though they have the impedences reversed - the RCA side is hi-Z while the XLR side is 600 ohm):

Typical video balun schematics:

 

If what you were saying was true then there would not be any ground loops with one camera and recorder.

There aren't... at least, not due to the use of video baluns (there are plenty of other possible causes). As I already stated, one camera - no problem. As you add cameras, the problem becomes worse. And it ONLY happens (in my experience) on cheap cameras with a shared video and audio ground, and a shared power supply. Dual-voltage cameras, no problem. AC-only cameras, no problem. 12VDC cameras with regulators, no problem. Anything that isolates your power connection's ground from the video connection's ground, eliminates this issue.

See, I came from an audio background, live and studio. I was very familiar with audio "isolation-type" baluns. I've also used isolation transformers in car audio to break ground loops.

So when I discovered this problem on a system using baluns and a buttload of cheap cameras, I couldn't figure out why it was happening, because I figured video baluns were isolation transformers too.  But I also noticed, on this particular site, that when I took cameras off the power can and gave them their own 12V wall wart, they cleaned right up.  Cameras still on the can - noise; cameras on separate power, no noise. Two cameras on a wall wart gave a little noise.  Three got even worse. In short, any time multiple cameras shared a power connection, they got noise. Anything not on that same power was fine.

Later I bought a couple dozen super-cheap "baluns" my buddy came across... they didn't work, so I took one apart to see how they were wired: it essentially matched the first diagram I have under video baluns. I metered it... yup, resistance from one balanced pin to the coax center pin, but infinite resistance measured between pins, and between center pin and shield on the coax side - not what you would expect from an isolation transformer.  So I looked up video balun schematics, and there they were... 

Look, in a perfect world, you DON'T want DC from the power supply traveling in the coax.

Of course you don't. But when the camera uses the same ground for power and video, that's what happens, whether you want it or not. And that's why CHEAP cameras have this problem.

Its the DC in the coax that gives you the artifacting you see in the screen.

You are literally looking at the current flowing between the two different ground potentials.

Yes you are. Differing potentials caused by the (relatively) massive DC resistance introduced by the balun designs shown above, where a coil is inserted in-line with the signal. 

U
Undisclosed #2
Jun 03, 2016
IPVMU Certified

That's a lot of information there! Most of it I either don't have a problem with, or don't know enough about it to have a problem with it. Thanks, it was fun reading.

What I do still have a problem with is this original statement:

This setup creates two vastly different ground paths to the camera: one short, direct one via the power line to a common ground at the power can; and one very long one with high DC resistance, passing through the coils of the two baluns, back to the common signal ground on the DVR.

Just to be clear, "vastly different grounds paths" aren't enough to cause ground loops.

For instance, there are an infinite number of ground paths from signal to ground at any moment, thru the air, thru insulation etc.

They are vastly different, but they don't cause any problem because they don't present a low enough impeadance option to cause any significant current to flow.

So when you say that the coax has a very high DC resistance, that means very little signal distortion, because very little stray current is mixed with the signal.

Would you continue to say if the DC resistance was higher that the problem would be worse?

Don't you think that sounds a little fishy?

Avatar
Chuck Janzer
Jun 05, 2016
IPVMU Certified

Similar to Matt, we stopped installing coax in about 2005, unless really pushed.

The story I have.... We were doing a larger new construction project and installed all Cat5e with analog cameras. We sold it to the customer saying that IP cameras are coming, and this will future proof the cable investment. It worked well. We ran the Cat5 to each data closet and then a 25 pair telephone trunk cable from each closet to the head end. We added cameras over the following few years using the same concept.

We lost the customer to a competitor when the security director changed. The new company installed six cameras using coax - camera to head end, losing the concept of data closets and using the UTP for analog.

We got the customer back two years ago. We then changed the head end to hybrid and have been installing only IP cameras. They really wanted to change those analog coax cameras to high res IP. In each case we evaluate if it will be more effective to use coax-IP converters or install Cat5. It was just foolish that they didn't evaluate the design concept in the building before running coax for hundreds of feet when it wasn't necessary.

Cat5!

Avatar
Chris Daniels
Jun 05, 2016
IPVMU Certified

Ethernet is the only cable we specify for Video Surveillance due t

o flexibility of applications and very detailed installation standards. (TIA/EIA BUILDING TELECOMMUNICATIONS WIRING STANDARDS & BICSI TDMM). Power over Coax is rather proprietary, termination and cbe certification practices are not as uniform.

UM
Undisclosed Manufacturer #3
Jun 05, 2016

You should NEVER use a pair in the cat5 cable to power your Analog or HD Analog cameras. If you really want to future proof you rumen both cat5 and coax cables! You can get greater distance in HD analog with coax.

Avatar
Chuck Janzer
Jun 05, 2016
IPVMU Certified

UD3 - Why do you say that? Just use service loops on each end, depending on your termination design, so that someday you can properly terminate and test to Cat standards.

(1)
New discussion

Ask questions and get answers to your physical security questions from IPVM team members and fellow subscribers.

Newest discussions