Baluns and Ground Loop issues

[koolmer 0]

I know there have been threads around about baluns and ground loop problems, but I haven't found any understandable explaination of the problem.

 

I am using baluns on many cameras in 5 different CCTV systems. The ground loop problem (hum bar or just a lot of noise) has only occured on very few cameras. I know that separate power supplies can handle the problem, but in some cases I cannot install 4 power supplies just because I have 4 cameras.

 

Please let me know what you know about this problem and how to avoid it. Most interesting for me is WHY this is happening.

 

koolmer

[alasl01 0]

Between the limiting space in some jobs for cable placing to old failed wiring the ground loop issues are plenty. Depending on the application and distance there are several ways to eliminate the problem.

 

For long distances over 1000 feet to 4,500 feet you can use active transmitters and receivers. Though there are plenty try using NTX or other quality ones.

 

For shorter distances there is a unique solution by Innotech Security. They transmit power, video and OSD controls through CAT5 which eliminates ground loop issues along with giving distances from 0-1,000 feet. The solutions are built into their cameras.

[Soundy 1]

I know there have been threads around about baluns and ground loop problems, but I haven't found any understandable explaination of the problem.

 

I am using baluns on many cameras in 5 different CCTV systems. The ground loop problem (hum bar or just a lot of noise) has only occured on very few cameras. I know that separate power supplies can handle the problem, but in some cases I cannot install 4 power supplies just because I have 4 cameras.

 

Please let me know what you know about this problem and how to avoid it. Most interesting for me is WHY this is happening.

 

koolmer

 

A ground loop of this type occurs when you have two different paths to ground of substantially differing length. Most cheap 12V cameras have a common ground for video and power, so right there you have two paths to ground, one via the power ground, and one via the video shield. Normally, using coax and power, or siamese cable, the two ground paths are essentially the same length, and there's no problem.

 

If you look at the internals of a balun, however, they're effectively inserting a matching transformer inline with the two video wires, thus increasing the total length by the length of the wire in the coil windings, which could be dozens or hundreds of feet.

 

So you end up with one ground path (on the power) that's maybe 30-100 feet long, and another (through *two* baluns) that's 10 times longer, or more.

 

Within a single camera with an isolated power supply, this isn't usually a problem, but when you tie mutiple cameras into a single power supply with a common ground between all of them, and also have a common ground between all of the DVR's inputs, you end up with a mess indeed.

[rory 0]

Buy Distributed Isolated Individually Fused Power Supplies?

Altronix makes some for CCTV.

[koolmer 0]

A ground loop of this type occurs when you have two different paths to ground of substantially differing length. Most cheap 12V cameras have a common ground for video and power, so right there you have two paths to ground, one via the power ground, and one via the video shield. Normally, using coax and power, or siamese cable, the two ground paths are essentially the same length, and there's no problem.

 

If you look at the internals of a balun, however, they're effectively inserting a matching transformer inline with the two video wires, thus increasing the total length by the length of the wire in the coil windings, which could be dozens or hundreds of feet.

 

So you end up with one ground path (on the power) that's maybe 30-100 feet long, and another (through *two* baluns) that's 10 times longer, or more.

 

Within a single camera with an isolated power supply, this isn't usually a problem, but when you tie mutiple cameras into a single power supply with a common ground between all of them, and also have a common ground between all of the DVR's inputs, you end up with a mess indeed.

 

Wow - thanks for the explanation! You seem to know something about that. Any suggestion how to avoid it without multiple power supplies?

 

 

Buy Distributed Isolated Individually Fused Power Supplies?

Altronix makes some for CCTV.

 

Is this proven to work? Do they also sell in Europe for 230V?

[blake 0]

I looked at those Altronix isolated power supplies and i only saw them in 24V applications.

[survtech 0]

12VDC isolated power supplies are quite rare and 230VAC versions are rarer still. Individual supplies for each camera may be your only option.

[Soundy 1]

Wow - thanks for the explanation! You seem to know something about that. Any suggestion how to avoid it without multiple power supplies?

 

Yeah, I've run into it before... spent a lot of time puzzling over it, experimenting with it, asking around as to why it was happening with nobody really having a definitive answer or even a good guess... but coming from the audio world, I'm quite familiar with path-length ground loop problems.

 

Best solution I know of, in this case, is to use 24VAC cameras. Since they naturally have an internal power rectifier, they don't share a common ground between the power and video, and thus don't have two ground paths. The other solution that SHOULD work (although I've not had a chance to test it) is if you have 12VDC cameras with internal regulators - again, something that effective separates the power and video grounds from each other inside the camera.

 

Dual-voltage (12VDC/24VAC) cameras SHOULD also be free of the problem when used with 12VDC power supplies, for the same reason. However, I've found that even having a mix of cameras that include cheap 12VDC-only cameras can cause induced noise on other cameras. The problem is, once the noise gets into the ground plane of the system, it can get into any of the channels.

 

Buy Distributed Isolated Individually Fused Power Supplies?

Altronix makes some for CCTV.

 

Is this proven to work? Do they also sell in Europe for 230V?

 

Essentially what this requires is the power supply to have an individual rectifier and regulator for each output channel - again, something that separates that channel's power ground from the others. You'd end up with something that's effectively one big transformer with a whole bunch of regulator boards inside a single box. "Individually fused" will have no effect on the problem.

 

I had one other idea, that it might help to insert a diode in the power ground, ideally close to or inside the camera, to separate that connection somewhat... haven't had a chance to test that yet, either.

[Soundy 1]

12VDC isolated power supplies are quite rare and 230VAC versions are rarer still. Individual supplies for each camera may be your only option.

 

The good thing is, if you had a 120VAC version, it would work fine with a step-down transformer (such as you might use to run any other North American device on European power). You might be able to just swap out the internal transformer for something appropriate as well (check the existing transformer's secondary voltage and current rating, then find a similar one with a 230V primary).

 

BTW, koolmer, if you do end up having to use individual supplies, see if you can get "switching" types rather than transformers. They're FAR smaller and run a lot cooler, so it's easier and safer to pack them onto a single power bar. Cost may be the only issue, but most of the ones I see are no more expensive.

[Kiwi 0]

I am using baluns on many cameras in 5 different CCTV systems. The ground loop problem (hum bar or just a lot of noise) has only occured on very few cameras. koolmer

 

I see it as more of a camera problem and would suggest it's best to avoid those where it occurs. Any decent modern camera should have isolated power and video to get the best balun performance in any case. Having to use an isolated supply because it's tied to the balanced twisted pair seems a bit of a kludge.

[Soundy 1]

I am using baluns on many cameras in 5 different CCTV systems. The ground loop problem (hum bar or just a lot of noise) has only occured on very few cameras. koolmer

 

I see it as more of a camera problem and would suggest it's best to avoid those where it occurs. Any decent modern camera should have isolated power and video to get the best balun performance in any case. Having to use an isolated supply because it's tied to the balanced twisted pair seems a bit of a kludge.

 

It IS a kludge, but that's a drawback of CHEAP cameras. Save money in one place, spend it in another...

[koolmer 0]

Thanks for all the answers!

 

What soundy suggests with the diode is something I will definately try. Otherwise I could get some rectifiers and connect them to the secondary voltage of my transformers. Would it make a difference where in the power line I place the rectifier? This would definately be cheaper than to get more power supplies.

 

I get the problem mostly with cameras that have build in IR. They only work on 12VDC, but in the future I will see that I only buy dual voltage cameras.

 

koolmer

[Soundy 1]

One thing to keep in mind with diodes and rectifiers: you WILL lose voltage across them. A typical silicon diode will drop about 0.7V; since a bridge rectifier basically inserts two of them in series, that will drop 1.4V or so. If you're using a regulated supply with a solid 12VDC output, you'll be lowering what the camera sees to barely 10.6V... even less with line losses.

 

I'd test it with a single diode, first... if that doesn't help, going to a rectifier probably won't either.

[koolmer 0]

One thing to keep in mind with diodes and rectifiers: you WILL lose voltage across them. A typical silicon diode will drop about 0.7V; since a bridge rectifier basically inserts two of them in series, that will drop 1.4V or so. If you're using a regulated supply with a solid 12VDC output, you'll be lowering what the camera sees to barely 10.6V... even less with line losses.

 

I'd test it with a single diode, first... if that doesn't help, going to a rectifier probably won't either.

 

I will just build a rectifier myself and try it out. My powersupplies are regulated 13.8VDC, so it should be okay to loose 2V.

[survtech 0]

You have to be careful using 13.8VDC power supplies. I have seen a few cameras that don't like that high a voltage. We fried a Dallmeier camera using that type of power supply. It seems the camera wanted 12VDC +/- 5%, which would be 11.4 to 12.6 volts.

 

I agree the design is stupid and told the manufacturer that .

[Soundy 1]

if you use an adjustable regulator, you can just insert your diode or rectifier, put a meter at the camera and tweak your output voltage for the desired results

[Kiwi 0]

Shifting the ground plane with a diode is not going to help the situation. You need to provide fully-isolated power to each of those affected cameras, such as a separate AC adapter.

[koolmer 0]

Shifting the ground plane with a diode is not going to help the situation. You need to provide fully-isolated power to each of those affected cameras, such as a separate AC adapter.

 

Have you tried it? I didn't yet.

[Soundy 1]

I don't have any grand expectation that it's a magic fix, but like I say, it was something I had an idea to experiment with. If it DOES work, a single diode will be just as effective as a bridge rectifier - all the bridge effectively does is add another diode to the power line.

[survtech 0]

I too have my doubts that a diode would work. I suppose it would depend on whether the difference was a positive or negative voltage. You could use a "ground loop isolator" in each camera's video line but those tend to be relatively expensive.

[koolmer 0]

I too have my doubts that a diode would work. I suppose it would depend on whether the difference was a positive or negative voltage. You could use a "ground loop isolator" in each camera's video line but those tend to be relatively expensive.

 

Yeah I've seen these ground loop isolators, but they are really expensive. Does anyone know what they do exactly and how they do it?

[survtech 0]

Yes. I've opened up a Pelco GIT100. It's basically an autotransformer made from very small diameter coax wrapped around an (iron?) form. It won't totally get rid of strong voltage differentials but it will minimize them. Here are the specs:

 

Ground Loop Voltage - Maximum CMV of 10 Vp-p at 60 Hz

Video Performance Characteristics - Equivalent in all respects to approximately 200 feet (60.96 m) of RG59/U coax

 

Table A. Common Mode Rejection (CMR)

 

Length of RG59/U ... Effective CMR at 60 Hz

200 feet (60.96 m) ... 43 dB

400 feet (121.9 m) ... 37 dB

600 feet (182.9 m) ... 33 dB

800 feet (243.8 m) ... 31 dB

1,000 feet (300.8 m) ... 29 dB

 

I'm not even certain it would work on DC ground loops. Transformers typically only work on AC and they do mention 60 Hz. If you try one, make certain you can get a refund if it doesn't work.

 

By the way, at least one company, Speco, makes one for UTP.

[Kiwi 0]

Shifting the ground plane with a diode is not going to help the situation...

Have you tried it? I didn't yet.

No, but to be honest it's just common sense from an electrical engineering point of view. Twisted pair will only work properly if the currents on each side are equal. Any leakage path will upset that and that leakage is extremely unlikely to be corrected by applying a DC offset of the PS (-) to the tune of 0.7 V.

[Soundy 1]

Yes. I've opened up a Pelco GIT100. It's basically an autotransformer made from very small diameter coax wrapped around an (iron?) form. It won't totally get rid of strong voltage differentials but it will minimize them.

 

Sounds more like a basic choke than a "ground loop isolator" - ie. it's not really "isolating" anything.

 

In audio, at least, a ground loop isolator does exactly that: it's a basic matching transformer (1:1 windings) in the audio line that physically breaks the ground path via the audio lines.

 

Of course, "baluns" are a little different design in audio as well, as they too are generally transformers with the input wired to one coil and the output to the other.

[ak357 0]

Shifting the ground plane with a diode is not going to help the situation. You need to provide fully-isolated power to each of those affected cameras, such as a separate AC adapter.

 

Agree with KIWI don't see reason to "rectify" noise"