work in a data center and we have taken classes re: bending CAT 5, CAT 6, fiber cables.
they teach that bending these cables anywhere close to or past a 90 degree angle could cause data transmission to decrease in speed and even worse, errors, or loss of.
As opposed to an IDE cable, a single Cat5 cable runs great distances, thereby, are prone to dropped packets/data loss. Foil shielding, braided wire shielding offer more protection caused from EMF/RFI in the form of flourescent lights, electric fans, AM and CB radios, as compared to a non shielded pair. Moreover when a hard bend exceeds 90 deg, it induces stretching of the inner shileding causing it to open, and loose its shielding characteristics. Bending is merely a simple precaution.
EMI / RFI stand for ElectroMagnetic Interference and Radio Frequency Interference (RFI is just EMI over the range of 1kHz - 10GHz). The measure of shielding effectiveness is expressed in decibels (dB). The higher the dB rating, the greater it's effectiveness. Silver is the most most conductive of commonly used coatings, has an outstanding conductivity of 0.010 ohms/square/mil provide effective shielding up to 10 GHz, at over 75 dB.
Shielding them from EMI / RFI is done for two primary reasons: (1) to keep your electronic device from interfering with others, or (2) to keep other devices from interfering with yours.
so, i am wondering... always wondering HOW these boutique builders (Voodoo - see their non fan cooled/ heatsink box) tidy up their cabling so well.
then i noticed, that they bend and hide their IDE ribbon (non rounded) in somewhat origami type folding.
why does this not affect the data transmitting through those cables? and if it may cause loss, would their already absurd benchmarks be even greater if not doing this?
How do you measure any significant signal loss from an IDE cable often less than three feet in length? Bending of these flat IDE cable into origami fashion causes internal stress to the wiring and is more a common sense issue, at best. If I gave these cables flat creases at right angles, today, and came back tomorrow to make a flying proa out of them...the more they are bent back and fourth, they will obviously suffer a loss of integrity. I personally don't see any real issues in benchmarking caused by creasing a new cable at right angles along it's entire length (one time). More concern should be placed on how much it's bent back on the same crease, and if it's touching exposed circuits, and in close proximity to fan motors, flourescent ballasts, etc...