Me and Eok were talking, sure we lead ears can't hear the difference between cables, but we surely do agree that from an engineering standpoint, superior cabling can be made with enough knowledge. I sure can't hear or can't see the point of building a BMW M6 to drive around the jammed streets of Kuala Lumpur....butbutbut.....it's there and it's a dream.
The price point is high since i've relegated this to be my best work with no real expenses spared. I'll be building building a 3 meter IC, finished depending on my finances to the highest of standards. So here's a list to design goals.
- Minimum Resistance (but enough to kill off reflections)
- Zero Inductive Coupling
- Minimum Impedence mismatch
- Minimum Reflection
- Optimized Conductor use
- Maximum RFI and EMI rejection
- Minimum Inherent body noise
- Minimal handling
- 80dB of 100khz Attenuation
- Intererence free dedicated ground
- Minimized temperature coupling (this part i seriously don't know how to achieve, besides putting heat sinks on the damned thing)
The base core idea is simple, 2 cores of silver wire (best material for low resistance) within clear teflon insulation given the maximum twist ratio without resorting to flat cabling (that becomes a big ass antenna), which then are then twisted with another pair of silver wires in teflon.
If possible here i will pad it with some sort of polyeurathane foam, but so far i have been unable to find a source. It reduces handling noise and increase temp insulation, keeping the wire from heating up due to other sources that can cause thermal agitation and noise. But we'll skip this.
Ok, now for the noise shielding. I seriously have no source on this, but for filtering out high frequency RFI, this is a must due to it's lack of holes like braid (anything with a high enough lamda can slip through a braid shield). So i devised a plan of using either bulk foil with a polymer backing as a shield, or good old cooking aluminium foil cut into strips and then slotted (following the Belden Method) and then wrapped around the two conductors, and then sealed with a proper epoxy.
Secondary noise shielding will be provided by 95 percent coverage copper braid from Alpha wire, this is to absorb EMI and due to it's low resistance it does have very good qualities even when it comes to higher frequencies.
Tertiary noise shielding is provided by a ferrite core, specifically selected to filter out the frequencies that i have calculated to lack in attentuation from the all the above interference rejection methods. The ferrite core has a curved attenuation ratio depending on frequencies, so the ferrite acts as a spot cleaner, patching up attenuation gaps that i might have missed and/or augmenting the other shields.
Finally added on top is plain strong high AWG PVC jacket. The boundaries between the RCA pluges and the stripped shields after the single cable splits into an L and R plug each is to be shielded with 3M copper embossed tape and the material of each plug is to be as closely matched to the wire as possible (WBT silvers
)
A point i'd like to make, it's simple why people put connectors with resistance into their ICs, to reduce reflections from the impedence mismatch, depending setup, i don't plan to offer any resistance on the cable itself, although i might add a resistor in series for it. It's trade off between both sides, resistance increases noise, but so does reflections, so i'll let a sampling decide instead. All connections are made with eutectic solder
Lastly the shield ground will be a post meant to be connected to chassis grounded device. This is to make sure no ground contamination occurs. And in theory the noise level should even plunge further.
So i guess that's it. All of the above are generally made with science, matlab (for materials analysis) , mouser, digikey, hakko, qualitek, alpha and 100 percent overkill.
If the first batch goes well, i'll start selling outrageously overpriced ICs to rich goons which keeping them cheap for hobbists! YAY for Capitalism!
This post has been edited by empire23: Mar 28 2007, 06:53 AM