Video flicker problem with lots of scope data

In the image below, when the switch is closed (and when it's opened again), the video has a severe flicker. It only does this when the switch is changing states. While the load is on, the video returns to normal. I've been trying for a week to figure out how to get rid of it. Here is some scope data that may help someone more experienced than I.. Any help would be greatly appreciated

Can you give the 4A device its own pair of wires from the battery?

it has it's own positive lead coming from the battery but it shares a ground near the camera. (I realize that is different from what the wiring diagram shows.)

Your picture only shows a single connection for the video, so presumably it's using the common earth. That suggests that the two wires carrying the video signal may not be close together. You might try taking the video and earth to the camera via a shielded twisted pair cable (with the shield grounded at one end) and see if that reduces the amount of EMF picked up on the video channel.

They are in their stock housing coming out of the camera, then they are apart for about half an inch as they go into a cat5 breakout board, then they are in a twisted cat5 pair for 8ft or so, then the reverse happens as they go into the LCD screen at the other end.

I have an idea of what might be happening, if thought of like a Y connection in a water pipe, when the the high amp load is turned on all of the water (amperage) flows into it temporarily depriving the camera and causing the brief loss of video. Once the motor's on, the current flow stabilizes and the video is clear again while the motor is powered. Is that what's happening? if so, how do you make sure the camera will always have the current and voltage it needs to operate smoothly? Use an inductor?

If you provide the 4A load with its own supply (with common earth) does this affect the problem?

David82:
They are in their stock housing coming out of the camera, then they are apart for about half an inch as they go into a cat5 breakout board, then they are in a twisted cat5 pair for 8ft or so, then the reverse happens as they go into the LCD screen at the other end.

Wait - you're not using impedance matching baluns, similar to these:

http://www.active-vision.com/Passive-CCTV-Balun-p/abl-1p03-c.htm

???

...also - video shield/signal "ground" should be kept separate from chassis/power ground, IIRC...

PeterH:
If you provide the 4A load with its own supply (with common earth) does this affect the problem?

It's still the same. That was one of the first things I tried.

cr0sh:
Wait - you're not using impedance matching baluns

\

What specifically do those do besides provide an easy breakout for video? I don't have any issues with video over longer distances (only tested 50 ft). Just when the switch is cycled.

cr0sh:
...also - video shield/signal "ground" should be kept separate from chassis/power ground, IIRC...

I can't. All 8 ports of the cat5 breakout board are being used. consolidating all the grounds together was what enabled this to work with a cat5 cable in the first place.

What size wire is used between the 12v battery and the 4a load, and how long is that wire?

zoomkat:
What size wire is used between the 12v battery and the 4a load, and how long is that wire?

It is just whatever thickness is usually used in PC power supplies. Then it goes into a solid copper cat5 cable and back to pc power supply wire. From there it is soldered to the thicker pigtails wires that came with the 4amp motor.

Then it goes into a solid copper cat5 cable and back to pc

Cat5 cables have 8 conductors, how many of these did you use for the power +- power wires? You also seemed to missed the "how long" part of the question, especially the cat5 section.

just one is ground, one is 12v+ that powers the camera, one is the power for the 4amp load, one is +video, and the other four are going to two bi-directional motors which, by the way, have no effect at all on the video when used.

These graphs are before and after adding a cap across ground and load+. I also tried other cameras, and powering those other cameras with a completely independent power supply (same results). Maybe the cap increases the current draw because the power supply has to suddenly charge the cap and power the load.

For the final time, how long is the cat5 cable run? :roll_eyes:

8ft like it says in the wiring illustration and again a few replies back.

Is this a repost of the same problem you asked about in 'How do you make the current draw from a device gradual instead of instant?', and again in 'How do I prevent video interference caused by electric motors', and also in 'Simple video interference problem. I'll pay to have it resolved.'? The people following this thread are in danger of wasting a lot of time and effort coming up with suggestions that have already been made.

I'm trying to provide the information necessary for someone experienced enough to actually explain why the problem exists and exactly, with certainty, what will solve it.

That sounds like a 'yes'.