Max, Video cables come in at least 3 flavors.
1. High Density 15 Pin VGA to High Density 15 Pin VGA 2. High Density 15 Pin VGA to BNC 3. 13W3 to 13W3 (e.g. Sun Microsystem's video cables) For the types 1 and 3, there is typically an outer shield that is connected to the D-shell connector at each end and is separate from the R/G/B returns on the outside of the system enclosure. These tend to perform much better than type 2 above. I have run into a lot of EMI problems with cables that are D-shell at one end and BNC at the other. This is because the BNC's must break out of the D-shell shield near the monitor end. This is improper termination of the outer shield from the D-shell. Also, the BNC coaxes do not have an ideal termination at the D-shell connector. They can't possibly due to the non-coaxial geometry of the D-shell's signal pins. This typically results in excessive RF currents on the shields of this type of cable. I am not saying that they can't pass B limits, just that they often don't. In theory, if the R/G/B cables were BNC to BNC and there was very low leakage through the outer coax braid then they would be good for EMI suppression as the signal and return currents would be on the interior surface of the coax conductors due to the symetrical field distrubution (equal and opposite signal and retune currents within the coax). But the BNC to VGA construction is far from this theoretical ideal. Finally, if you disassemble most "low end" color displays (esp 15" and smaller and cheap 17" models), you will find that they have all but eliminated shielding of the video amp circuitry. They tend to have very poor shield termination of the D-shell at the monitor end and very inductive connections between the various pieces of "grounded" metal within the plastic enclosures. So even if you use properly designed cables, the poor EMC design of the monitor kills you if you are using a high bandwidth video card at high resolutions. I can only suspect that a lot of the cheap class B video displays out there are qualified with video cards having very compromised output bandwidth because they sure don't seem to pass with higher performance graphics cards. Regards, [email protected] ---------- From: Max To: Steve Chin Cc: Jon D Curtis; mkelson; comp_lab; EMC-PSTC; Tony Fredriksson Subject: Re: Immunity/Cables List-Post: [email protected] Date: Friday, February 14, 1997 9:03AM Steve, %>Max, the amount of immunity a cable has is dependent upon the type of %>shielding it has and how well it is connected to the case ground. I have gone %>through a whole bunch of gyrations with the cable vendors I work with (my %>other responsibility at this company is high-speed signal cable development) %>in order to build EMC-tight cables. %> %>If the cable is built and grounded well, then the likelyhood of a failure due %>to the effects of outside radiation is very low. %> %>Steve Chin %>StreamLogic Corp. %>Menlo Park, CA, USA %> That's been my experience also with PC's, workstations. I haven't ever tested anything to a severe industrial level, though. However, I assume there wouldn't be a problem if the cables were well shielded and connected to the chassis. This raises another interesting question concerning the connection of the return to the chassis. Years ago I did a survey of a variety of monitor manufacturers to see if any of them used differential input. All of them said no. In addition, I have checked PC's and workstations from many manufacturers and never found any that didn't have the (video) cable return connected to the case. I assume, however, that without this connection of return to the cabinet, a system would probably flunk emission tests as well as immunity tests. Has anyone had any experiences in this area? Thanks, Max Max Kelson [email protected]
