Help on horizontal lines/ghosting

[RoyB 0]

I am at a total loss here. My background is in commercial electrical but have installed 3 camera systems so far.

 

My current issue is on a new install, 44 cameras. I have 3 power supplies (altronix sav18d) which are designed for CCTV, and 3 DVRs. Wiring is all cat 5, minus 2 outdoor cameras. We ran 2 cameras on 1 cat 5. Done it in the past without issues. 1 pair for power, 1 for video for each camera.

 

The problem: horizontal lines.

Solution 1: twisting 2 power pairs together for the cctv power supply.

Solution 2: using a 12 volt computer monitor power supply on one pair

 

I have not needed to do this before. All runs are less than 400 feet, some less than 200 and I have issues. I have tried EVERYTHING. Grounding is spot on, I've tried different power sources unhooking other cameras, etc etc. The only solution I have found that works is twisting two pairs together. That is fine, but building is complete and not easy to run 20 extra cables now.

 

One thing I tried before twisting 2 pairs together was a different 12V source- I used a 12VDC video monitor power supply and tried it instead of the CCTV power and the picture works fine...with ONE PAIR for the power. How can I use 1 pair on that power supply, but a dedicated CCTV power supply I need to twist the pairs together? I would say it was a bad CCTV supply, but all 3 do it. I brought in an old power supply and had similar issues with wavy lines. Could the problem be the power supply in the end? The sav18D doesn't look like a real great unit, but is designed for this application.

 

Would appreciate any insight. Thanks for the help.

[RoyB 0]

I should note, these are basic Dome cameras, not real fancy. Here is a video I took of the problem:

 

http://www.youtube.com/watch?v=RMDYz78S1_0&feature=youtu.be

[tomcctv 190]

can you list how you terminated each end .... how many paires for video how many for power. which baluns used and how have you grounded everything ??

 

12v at 400ft .... can you list the spec of the cameras used

[RoyB 0]

can you list how you terminated each end .... how many paires for video how many for power. which baluns used and how have you grounded everything ??

 

12v at 400ft .... can you list the spec of the cameras used

 

The power and video feeds were split at the head end/camera closet. Within one cat5e cable there are 2 power pairs being used and 2 video pairs being used. Roughly 16 cameras from each dvr/power supply.

 

Power pairs are ran into the power supplies, and into the camera on the cam end.

 

The video pairs have a Balun on camera side and DVR side. Baluns are models: (gem) BLN-TL and COMELIT-CNTB. It shouldn't matter if different baluns are used, correct?

 

Grounding- All power is in conduit here. Originally I had everything bonded to the box/conduit system. Then I tried building steel. Last resort was running a ground wire back to the panel- no differences were seen in either. I am getting <.2 ohms from conduit ground to steel and <.1 ohms when I tested to building water supply nearby. IMO, grounds are solid and not the issue...especially with ground wire to panel run.

 

I am pretty sure none of the runs are even at 400 feet. Most are <200 feet. Voltage at the camera UNDER LOAD is about 11.5 Volts. It drops 2 volts under load, without camera it is 13.5 volts. With the monitor power supply it drops 3 volts- 12 out, 9.5 under load and picture is clear. When I turn voltage down on the CCTV power supply to 12V out, the picture gets worse and turns off on a few. This to me makes zero sense.

 

Cameras are capture, I believe these are the specs:

SENSOR

Image Device 1/3-inch SONY CCD

Resolution 420 TV Lines

Lens Fixed 3.6 mm

Picture Elements NTSC: 510 (H) x 494 (V)

PAL: 500 (H) x 582 (V)

Minimum

Illumination

0.1 Lux / F1.2

Video Output 1Vp-p (75 ohms, Composite)

S/N Ratio 48dB (AGC off)

Shutter Speed NTSC: 1/60 – 1/100,000 sec

PAL: 1/50 – 1/100,000 sec

White Balance Auto

AGC Auto

BLC Auto

GENERAL

Input Voltage 12 VDC – a plug-in regulated

750mA power supply is

recommended (not included)

Consumption 120mA

Operating

Temperature

14 to 122 °F; -10°C to 50°C

RH 95% MAX (Non

Condensing)

[RoyB 0]

Couple pics of the setup. Not sure why I never get a real closet

 

 

[Soundy 1]

Within one cat5e cable there are 2 power pairs being used and 2 video pairs being used.

This is your problem - use only one pair for video, otherwise you don't get the noise rejection and balanced line function working properly.

[RoyB 0]

Within one cat5e cable there are 2 power pairs being used and 2 video pairs being used.

This is your problem - use only one pair for video, otherwise you don't get the noise rejection and balanced line function working properly.

 

I am not sure if it was clear- but I am only using one pair (power/video) for each camera. There are two power/video pairs within 1 cat5 because I am powering 2 cameras. I am not using 2 pairs for one video feed which is what it looks like you are saying?

 

One a side note, within testing, I have disconnected the second camera (on the cat5 line) from the power, dvr, and camera end completely and the problem persists. Basically eliminating any possible cross-talk/interference with the second camera.

[tomcctv 190]

are both cameras in the same location ??? 1 pair for power will not be enough. have you tried 2 pair for power with just 1 camera ??

 

also have you put a temp power at camera end to see what you have.

[Soundy 1]

Within one cat5e cable there are 2 power pairs being used and 2 video pairs being used.

This is your problem - use only one pair for video, otherwise you don't get the noise rejection and balanced line function working properly.

 

I am not sure if it was clear- but I am only using one pair (power/video) for each camera. There are two power/video pairs within 1 cat5 because I am powering 2 cameras. I am not using 2 pairs for one video feed which is what it looks like you are saying?

 

One a side note, within testing, I have disconnected the second camera (on the cat5 line) from the power, dvr, and camera end completely and the problem persists. Basically eliminating any possible cross-talk/interference with the second camera.

Okay, guess I missed that part.

 

Ummm... at a guess, I would suspect ground loops, which is common when using baluns with cheaper 12VDC cameras that share a video and power ground, and a central power supply. It's a little work, but one way you can test this is to use completely separate pairs for each camera, one video and one power, and then power each camera with a separate wall-wart power supply via the Cat5.

 

Powering each camera separately at the camera end will accomplish this same end, but that won't differentiate whether it's a ground loop issue or a problem with voltage loss over the long run. Assuming that powering the cameras directly clears up the problem, giving each camera a separate power supply over its own pair *at the head end* will determine whether it's a ground loop issue, or a voltage-drop issue.

[RoyB 0]

are both cameras in the same location ??? 1 pair for power will not be enough. have you tried 2 pair for power with just 1 camera ??

 

also have you put a temp power at camera end to see what you have.

 

Yes, I have tried 2 pair for power, 1 pair for video for 1 camera and it corrects the problem (in my first post). Everything I read online says 1 pair is enough for power though? I don't have enough cable for 2 pairs to each camera though. Two cameras on one cat 5 are roughly 30-50 feet apart on average. I also don't understand how a single 12volt power supply with one pair for the power corrects the problem as well. If I needed 2 pairs for power it shouldn't (IMO) matter what the 12VDC source was, correct?

[tomcctv 190]

hi. yes your problem is amps you need 2 pair.

[Soundy 1]

are both cameras in the same location ??? 1 pair for power will not be enough. have you tried 2 pair for power with just 1 camera ??

 

also have you put a temp power at camera end to see what you have.

 

Yes, I have tried 2 pair for power, 1 pair for video for 1 camera and it corrects the problem (in my first post). Everything I read online says 1 pair is enough for power though?

You'll always get voltage loss over a wire run. The longer the run, the smaller the wire, and the higher the current draw, the greater the loss.

 

For example, a camera that draws 300mA, over a distance of 100', will see a loss of about 1.54V over a single pair, or 0.77V with two pair. So if the power supply outputting a regulated 12VDC, then at the camera end, you're only getting 10.46V with one pair.

 

Keep in mind that doubling the cameras on a run also doubles the current, so with two cameras on the above example, even with the power wire doubled up, you're getting 600mA draw, which brings you back to 1.54V drop.

 

Now if the power supply is unregulated, it's probably pushing out around 15-16 volts with no load, so the loss may not be affecting it. You're using Altronix CCTV power cans though, so I'd guess the chances it's regulated are about 99.9999%.

 

I also don't understand how a single 12volt power supply with one pair for the power corrects the problem as well. If I needed 2 pairs for power it shouldn't (IMO) matter what the 12VDC source was, correct?

A quick test whether it's voltage or a ground loop might be to leave the second camera connected to power, but disconnect its video feed. That would maintain the current draw (and subsequent voltage loss), while breaking the ground loop. If that eliminates the problem, then that's your answer.

 

The best solution to a ground loop problem like this is to replace the cameras with cameras that have an internal regulator, which usually means 12VDC/4VAC dual-voltage cameras. The workarounds are a separate power supply for every camera, or ground-loop isolators.

[RoyB 0]

Okay, guess I missed that part.

 

Ummm... at a guess, I would suspect ground loops, which is common when using baluns with cheaper 12VDC cameras that share a video and power ground, and a central power supply. It's a little work, but one way you can test this is to use completely separate pairs for each camera, one video and one power, and then power each camera with a separate wall-wart power supply via the Cat5.

 

Powering each camera separately at the camera end will accomplish this same end, but that won't differentiate whether it's a ground loop issue or a problem with voltage loss over the long run. Assuming that powering the cameras directly clears up the problem, giving each camera a separate power supply over its own pair *at the head end* will determine whether it's a ground loop issue, or a voltage-drop issue.

 

I am not sure I am understanding correctly. Are you saying to use a separate power supply for EACH camera? or just a separate pair to power each camera? It is currently setup for 1 power pair to each camera now.

IE:

Blue pair= Video Cam 1

Orange pair= Video Cam 2

Green pair= Power Cam 1

Brown pair= Power Cam 2

 

In troubleshooting this is what fixes a trouble camera

Blue pair= Video Cam 1

Orange pair=unhooked

Green pair=Power Cam 1

Brown pair=Power Cam 1

 

OR this works too

 

Blue pair= Video Cam 1

Orange pair= Video Cam 2

Green pair= Power Cam 1 from single power source by itself

Brown pair=Power Cam 2 from CCTV power source with 14 other cameras.

[RoyB 0]

I also don't understand how a single 12volt power supply with one pair for the power corrects the problem as well. If I needed 2 pairs for power it shouldn't (IMO) matter what the 12VDC source was, correct?

A quick test whether it's voltage or a ground loop might be to leave the second camera connected to power, but disconnect its video feed. That would maintain the current draw (and subsequent voltage loss), while breaking the ground loop. If that eliminates the problem, then that's your answer.

 

The best solution to a ground loop problem like this is to replace the cameras with cameras that have an internal regulator, which usually means 12VDC/4VAC dual-voltage cameras. The workarounds are a separate power supply for every camera, or ground-loop isolators.

 

I will try this. So I understand correctly, you are saying to keep both cameras powered through the CCTV supply. Keep troubled camera plugged into DVR and unhook second camera on the cat 5 from camera end?

 

ie:

Blue=Video Cam 1 to DVR connected

Orange=Video Cam 2 to DVR UNHOOKED at camera end

Green= Power Cam 1 from CCTV power connected

Brown=Power Cam 2 from CCTV power connected

 

If problem goes away after unhooking video feed from second camera this tells me there is a ground loop issue and not voltage/power supply issue, correct? I am almost positive I unhooked the second camera from the DVR side, but not the camera side.

[RoyB 0]

Problem persists with the following done:

 

Blue=Video Cam 1 to DVR connected

Orange=Video Cam 2 to DVR UNHOOKED at camera end

Green= Power Cam 1 from CCTV power connected

Brown=Power Cam 2 from CCTV power connected or unhooked

 

I think the only solution is to run more cables. I just can't comprehend why it's fine with an external power source at camera end or head end.

[Soundy 1]

I've more than once run power and two video over a single Cat5 with no problem - these are CNB VCM-24VF cameras with about 185mA current draw at 12V. The difference is, these are dual-voltage cameras with internal regulators that separate the power and video grounds, so I don't get two different ground-path lengths. It's hard to be certain without knowing the make and model of your cameras, but I'd still be willing the bet the problem is a ground loop caused the cameras using a common video/power ground. This is what clinches it for me:

 

this works too

 

Blue pair= Video Cam 1

Orange pair= Video Cam 2

Green pair= Power Cam 1 from single power source by itself

Brown pair=Power Cam 2 from CCTV power source with 14 other cameras.

[PaulsonLaw 0]

Not that this is related, but our problem was the cable tv signal coax, the rg6, not being grounded by the installers back when. So this was grounding thru the tv to the rca input & all the way back thru the dvr....

Grounded nicely & all is good....

[RoyB 0]

I've more than once run power and two video over a single Cat5 with no problem - these are CNB VCM-24VF cameras with about 185mA current draw at 12V. The difference is, these are dual-voltage cameras with internal regulators that separate the power and video grounds, so I don't get two different ground-path lengths. It's hard to be certain without knowing the make and model of your cameras, but I'd still be willing the bet the problem is a ground loop caused the cameras using a common video/power ground. This is what clinches it for me:

 

this works too

 

Blue pair= Video Cam 1

Orange pair= Video Cam 2

Green pair= Power Cam 1 from single power source by itself

Brown pair=Power Cam 2 from CCTV power source with 14 other cameras.

 

how come when I unhook the other pair isolating it to just one pair for one camera power and video, the problem still exists though? Maybe I don't understand ground loop fully?

 

Will running 2 pair of power for each camera help the situation? That is my next current step as I don't know what to do otherwise.

[Horizon 0]

Hi RoyB.

 

You could have two problems there - both the ground loop and low voltage.

The ground loop is when you have two or more ground lines between the camera and DVR. In your case it is probably the power supply ground and the video ground. The twisted pair relies on the video signal being balanced, with the same signal strength on each half of the pair. The ground loop sends part of the video signal around a different path (the power ground), making the twisted pair unbalanced, and vulnerable to noise and interference. The ground loop also acts as an antenna, which increases it's ability to collect interference. The way to get around this is to break the loop, using either separate floating power supplies (floating in that the negative terminal is not connected to ground), or to fit isolating baluns at either the camera end, or at the DVR. Separate supplies wouldn't be practical for a 44 camera install.

Whether a camera is affected by ground loops depends on it's design. Some use isolated power supplies inside the camera - they chop the 12vDC into AC, then feed it through a transformer, then convert it to regulated DC. The transformer isolates the power supply ground from the video ground. Susceptable cameras don't use a transformer, they either use a linear regulator, or switching regulator and inductor. You can check cameras by testing for continuity between the camera BNC shield and the power supply negative pin. If there is continuity, the camera could have issues with ground loops.

 

Low voltage problem: Cameras don't always draw a constant current - sometimes there are surges, which combined with the resistance of the CAT5 cable will cause the voltage to sag at the camera. The multimeter will be averaging out the voltage that you've measured - it could be briefly dropping low, causing the camera electronics to glitch. You could try adding a (at a guess) 1000uF 25volt capacitor across the power supply terminals at the camera, and see if there is any improvement. Remember to connect the capacitor around the right way - it's polarised! This is a bit of a bodge - it would be better to run heavier wire.

 

Anyway,

Good luck!

[RoyB 0]

Hi RoyB.

 

You could have two problems there - both the ground loop and low voltage.

The ground loop is when you have two or more ground lines between the camera and DVR. In your case it is probably the power supply ground and the video ground. The twisted pair relies on the video signal being balanced, with the same signal strength on each half of the pair. The ground loop sends part of the video signal around a different path (the power ground), making the twisted pair unbalanced, and vulnerable to noise and interference. The ground loop also acts as an antenna, which increases it's ability to collect interference. The way to get around this is to break the loop, using either separate floating power supplies (floating in that the negative terminal is not connected to ground), or to fit isolating baluns at either the camera end, or at the DVR. Separate supplies wouldn't be practical for a 44 camera install.

Whether a camera is affected by ground loops depends on it's design. Some use isolated power supplies inside the camera - they chop the 12vDC into AC, then feed it through a transformer, then convert it to regulated DC. The transformer isolates the power supply ground from the video ground. Susceptable cameras don't use a transformer, they either use a linear regulator, or switching regulator and inductor. You can check cameras by testing for continuity between the camera BNC shield and the power supply negative pin. If there is continuity, the camera could have issues with ground loops.

 

Low voltage problem: Cameras don't always draw a constant current - sometimes there are surges, which combined with the resistance of the CAT5 cable will cause the voltage to sag at the camera. The multimeter will be averaging out the voltage that you've measured - it could be briefly dropping low, causing the camera electronics to glitch. You could try adding a (at a guess) 1000uF 25volt capacitor across the power supply terminals at the camera, and see if there is any improvement. Remember to connect the capacitor around the right way - it's polarised! This is a bit of a bodge - it would be better to run heavier wire.

 

Anyway,

Good luck!

 

 

This was spot on- It ended up being a ground-loop problem. I tested continuity on the cheaper cameras inside, and sure enough, Both - had continuity. The better outdoor cameras did not.

 

We ran extra cables for parallel power to the 3 floors. The paralleling worked on floor 1 and 2. For some reason when we added extra power to the bad looking cameras on floor 3, the ones that looked good before now looked bad. WTF? I started unhooking the new twisted power cables that I ran. When I unhooked the Negatives on the power supply, the cameras were still on! Whoa. I was able to to unhook 4 Negative sides, leaving only the positives attached and I STILL had video on these things! Wowww. That really told me there was a ground loop issue. I had a hard time understanding it until I did this and saw first hand how much power was going back into the power side of the camera. Enough to give me picture! I hooked up one of the bad cameras to a separate power supply and it was crystal clear. It also makes sense when you check continuity how you could be getting "cross talk".

 

The solution we are going to do for the last floor is put a couple plug-in single transformers in the ceiling for the 4 bad cameras. Hopefully that clears up the rest of the issues we have.

 

This post was a real eye opener. I really appreciate the help and learned so much by troubleshooting this issue. From here on out I will make sure the boss gets better cameras, even if they cost an extra $20 to make sure they have a way to isolate video and power. Also I will never run 2 cameras off 1 cat5. We did it because that is what the boss wanted, even though we have done it in the past with success. I think he just learned that going cheap doesn't always work out well in the end. I suggested calling our supplier and asking why they would sell us cameras, dvrs, and power supplies that don't all work together.

 

Appreciate the help. Just not sure why paralleling some of the cables on floors 1 and 2 fixed the problems and not the third floor. hmmmmm