VFD Interference with CCTV Cameras

[bfletcher]

I've got an installation where I'm feeding (16) geothermal wells located under a parking lot. Also in the parking lot are some CCTV cameras mounted on light poles. The geothermal wells are controlled by VFD's located in the building. Each geothermal well feed from the VFD is in it's own 1" pvc conduit from the building to the well location. The feeds for the cameras are located 20' (or so the electrical contractor tells me) from the geothermal well power feeds. The problem is that the CCTV camera pictures have distortion. A lot of trouble shooting has been done (length of cable, shielding, etc.) and the problem has been traced to the geothermal well power feeds. Whent he wells are shut off, the CCTV pictures are not distorted. The CCTV cameras are fed with coax. We maintained recommended seperation distances and we still get distorted pictures. Any ideas?

 

[Cold Fusion]

Not my area of expertise - so I lkely won't be much help. Mostly I'm curious as to how it gets solved. Keep us posted.

I'd see if I could figure where the distortion is comming in. Figuring the VFDs put out quite a bit of harmonics onto the AC line, start with the power supply.

Out at the camera pole, feed a camera with a clean source of power. Either batteries to a DC input, or batteries and a sine wave inverter. If that does it, perhaps filters on the AC inputs to the VFDs will help

Or maybe the interference is radiated from the unshielded power conductors from the VFDs to the well pump motors. Maybe look for a high pass filter for the coax line. I have no idea what a CCTV video signal looks like. So I wouldn't have a clue at to which band of frequencies you are looking to block. I've got a pretty good Tek2024 scope that will do frequence domain plots, so you could see the change in frequencies - pumps on, pumps off. perhaps something like that wold help.

cf

 

[JWCELECTRIC]

BF,

Is the CCTV equipment located in the same electrical room? The VFD's could be interfering with the equipment not the cameras on the poles.

Start turning one one VFD at a time to see if it is only a particular one causing the problem.

Change the camera wire to a Heavy duty sheilded cable that reflects unwanted signals.

 

[bfletcher]

The VFDs and camera equipment are in two seperate rooms/ We did everything possible to keep these systems away from each other, or at least we thought we did.

 

[drbond24]

I think I understand the OP to be saying that the VFD cable is shielded. If not, installed shielded cable.

If it is shielded, is the shield smooth and continuous? In other words, when the shielded VFD cable was installed was it yanked around and bent such that the shield could be kinked up? A wrinkle or tear in the shield can act as an RF transmitter which could cause the problems you are describing.

 

[nhfire77]

Also, You can use an isolated AC power supply for the cameras. They also can change the coax to twisted pair (CAT 5) using baluns.

 

[zbang]

Also, You can use an isolated AC power supply for the cameras. They also can change the coax to twisted pair (CAT 5) using baluns.

Without knowing what "distorted" means in this case, I'd try the baluns first. Lots of video issues are either snow in the picture (usually induced HF noise) or wavy-ness (low freq, often cause by a ground loop).

You could try the baluns without even pulling new cable, just insert a back-to-back set at the camera to isolate the feed.

 

[steve066]

Are you sure its the camera's picking up the noise, and not the monitors?

My first guess would be noise from the VFD coming in through the line cord on either the monitors, the recorder, or the power supply for the cameras. Any chance you can get a small portable UPS and plug all of the above into it and run it off the battery (don't plug in the UPS)?

Steve

 

[johnny watt]

Try Balun, 1:1 video isolation transformer.
When you say distorted it sounds like a sync issue. I would describe other problems as snow or confetti.
Make sure camera signal is high enough (1V p-p nominal.) may be affected by auto iris settings
Make sure coax is 75 ohm and is terminated at 75 ohm at the end of line at the monitor.
Try to run all cameras and monitors on the same leg.
Then try using line lock and see if that helps. It will sacrifice color quality but improve immunity to ground loops.
Try Ferrite beads on power wires feeding VFD
Replace pump wires with shielded
Replace coax with twinaxial cable
Dig up PVC and replace with rigid
Try Balun, 1:1 video isolation transformer.
Did someone already suggest that?

 

[Jon456]

Have you tried using isolated grounds on the power supply circuits to the CCTV cameras and recording/monitoring systems?

 

[quogueelectric]

This is what you are seeking.............http://www.hdtvmagazine.com/forum/v...ghlight=&sid=f8203efd8056051f7b8673d563663658

 

[dbuckley]

Have you tried using isolated grounds on the power supply circuits to the CCTV cameras and recording/monitoring systems?

Close

The issue is almost certainly that the interfering potential is getting across the coax screen, through ground paths. Put a humbug on each of the camera feeds which will isolate the ground. There are other ways to get the interference off the screen of the coax, but they are all more work

 

[zbang]

Make sure camera signal is high enough (1V p-p nominal.) may be affected by auto iris settings

But you can't measure that with a normal meter, you need a o'scope or waveform monitor. I wouldn't bother.

Try to run all cameras and monitors on the same leg.

IMHO, that's voodoo. TV studios don't bother with it and I've never bothered with it. Haven't had problems, either.

Replace coax with twinaxial cable

Uh, no. Twinaxe is a shielded twisted pair, and is going to be either 78 or 95 ohm. And that's even if you can get the adapters to make the connection. Are you thinking Triax? I wouldn't bother with that either.

When you get right down to it, without knowing just what's wrong with the picture, most the suggestions are wild guesses, albeit ones that solve common problems. It's like saying "that motor makes a funny noise, what's wrong?"

First thing to do is carry the monitor to the camera and see what the picture looks like there. Maybe post a picture of the problem. Then we can make concrete suggestions.

 

[Jraef]

Think about the output of a PWM VFD, which is now virtually all VFDs. Pulse Width Modulation of Frequency, i.e. Frequency Modulation, i.e. FM! The output cables of VFDs are great FM radio transmitters and depending on the size, relatively powerful ones as well. If the "shielding" you are referring to is only on the CCTV signals, that may not have been enough. VFD output cables must be in metallic conduit, or if not, the conductors must be run using specially designed shielded VFD cable. Then when installed, you must ground BOTH ends of the shields, unlike in signal conductors where you only ground one end.

http://www.belden.com/pdfs/Techpprs/Evaluating VFD.pdf

 

[drbond24]

And don't buy the stuff that is shielded with aluminum mesh, either. Get continuous copper shielding.

 

[dbuckley]

Talking of single phase power

IMHO, that's voodoo. TV studios don't bother with it

Certainly was standard BBC practice to use single phase power, and my still be, or not. But it was.

 

[glene77is]

Cold,
Good idea about the Clean Power Source.
About the power source feeds:
(1) Well filtered AC, through a pi-network.
(2) AC fed through a coax set having the (Faraday) shields
grounded at both ends (Inductive mode), maybe passing through Ferrite Cores.
(3) D.C. supply located at the CCTV itself.
(4) combinations from above.

I designed / built medical instrumentation, once upon a time,
and one device picked up A.M. radio stations.
I could tune between three stations by adjusting the grounding / shielding,
by altering the capacitance and inductance accumulating in these sub-systems.

For the video coax,
the Video signal itself tops out at 4 MHz,
and the channel 4 band is about 60 MHz.
So, a high pass filter looks good for the video signal.
(Faraday) sheilding grounded at one end (Capacitive mode) helps.
Pi-network filtering is good (in hi-pass mode).

Ferrite Cores are Not recommended (Inductive Impedance).
Xl = (2 pi F L ) and
Xc = 1 / ( 2 pi F C ) I think,
so do not add inductance in the video system.
Or so my memory goes.
:smile:

 

[glene77is]

Are you sure its the camera's picking up the noise, and not the monitors?
My first guess would be noise from the VFD coming in through the line cord on either the monitors, the recorder, or the power supply for the cameras. Any chance you can get a small portable UPS and plug all of the above into it and run it off the battery (don't plug in the UPS)?
Steve

Steve,
That's a good thought.
And running from an "un-plugged" UPS is a SUPER idea
for isolating the CCTV system and/or Monitor systems
from each other and from the Utility AC power.
:smile:

 

[steve066]

Steve,
That's a good thought.
And running from an "un-plugged" UPS is a SUPER idea
for isolating the CCTV system and/or Monitor systems
from each other and from the Utility AC power.
:smile:

Thanks. The way I see it, the noise is either being radiated and picked up by the equipment, or its coming in directly via the 120V wiring.

I'm not sure which is more likely, but its much easier to eliminate the direct wiring with a UPS or something similar. And it that is the cause, its much easier to fix.

Steve

 

[johnny watt]

Uh, no. Twinaxe is a shielded twisted pair, and is going to be either 78 or 95 ohm. And that's even if you can get the adapters to make the connection. Are you thinking Triax? I wouldn't bother with that either.

I?m sorry to throw in that unlikely solution. I was just regurgitating stuff I have seen before. I was talking about 128 ohm balanced twinaxial cable such as I have seen at several power plants for long distance composite video.

If the "shielding" you are referring to is only on the CCTV signals, that may not have been enough. VFD output cables must be in metallic conduit, or if not, the conductors must be run using specially designed shielded VFD cable. Then when installed, you must ground BOTH ends of the shields, unlike in signal conductors where you only ground one end.

Interesting.
Everything I have read says that a faraday shield should only be grounded at a single point.
Is there an explanation as to why the shield should be grounded at both points in this application?