Coil Cable Help

[KingJL]

Background: I have built a SuperD coil assembly ( TX coil 315uH, 0.6 ohm, 955 KHz SRF; each RX coil 376 uH, 21 ohm, 1.39 MHz SRF ). I am using the coil in conjunction with the moodz alt2h-bridge TX and a fully differential RX (coil fully floating (i.e. no coil connection to GND ), connected to a THAT1512). Initially using NEGLEX 2534 as a coil cable.

The problem: I find that there is WAY, WAY too much capacitive coupling between the conductors (RX to shield 525 pF; RX to TX 1.2 nF). The TX-RX coupling results in an 80 V pulse from the flyback being transferred to the RX input (am using clipper diodes to limit the input to the THAT 1512, and since the result is identical on the + and - THAT inputs, they cancel out differentially). Also the huge capacitance causes the flyback duration to increase from 1 usec to 1.75 usec.

I need to find a more suitable cable for my configuration. I have used Mogami W2947 dual coax S-video cable in the past but the center conductor is too thin for this application (AWG 27... need AWG24 or heavier). So far the best I can find is dual RG-6. I do not like the idea of the solid center conductor as it is a failure point. Also I would prefer to find something with a 5th conductor to use to connect the coil shield. Any suggestions would be much appreciated!

[waltr]

What about Cat5 cable? This might be ok for the RX but is it thick enough for TX.
Low C shielded cable is tough to find.

[Teleno]

Speaking from theory, the capacitance of the shielding cable could be eliminated by keeping the shielding at the same voltage level as the transient by means of a low impedance buffer. Hardly feasable if at all given the voltages involved.

Alternatively, the shielding could be disconnected from GND (high impedance state) during the transient and then reconnected to GND once the voltage level drops below a few volts. This could be realized by some sort of voltage dependent resistor, or perhaps even an inductor in series with the shield

Just an idea, I haven't tried or simulated any of this.

[Carl-NC]

The switch would need to be inside the coil, otherwise the coil still sees the cable.
The S-Video might be your best bet. If you can find 4-way S-Video then you could use 3 cables for TX and 1 for RX.

[KingJL]

The switch would need to be inside the coil, otherwise the coil still sees the cable.
The S-Video might be your best bet. If you can find 4-way S-Video then you could use 3 cables for TX and 1 for RX.

I'll search around and see what I can find in this... it just might solve my problem.

[Altra]

I have had good results with L-COM s-video cables. On ebay you can get a 50ft cable assembly for $19 free shipping. Only thing i dont like is the copper is untinned. It's cheap and can be cut into 10 5ft lengths. On my phone don't know how to cut and paste. Search ebay for
L-COM CCD234MM-50 male male 50ft molded s-video cable.

[KingJL]

... The S-Video might be your best bet. If you can find 4-way S-Video then you could use 3 cables for TX and 1 for RX.

Not yet located 4-way S-Video. But this... what would be your assessment?

[Carl-NC]

What more could you ask for? Capacitance looks good, and it's double twisted-pair-shielded with an over-shielding. I'm always leery of foil shielding but it's worth a try.
Bonus: for CCPI you really want to use shielded twisted pair for the TX. Probably on the differential RX as well.

[KingJL]

What more could you ask for? Capacitance looks good, and it's double twisted-pair-shielded with an over-shielding. I'm always leery of foil shielding but it's worth a try.
Bonus: for CCPI you really want to use shielded twisted pair for the TX. Probably on the differential RX as well.

Concurrent with my thoughts... only concern might be the voltage rating. I am going to place an order and give it a try. Will follow up with the results.

[Tinkerer]

Speaking from theory, the capacitance of the shielding cable could be eliminated by keeping the shielding at the same voltage level as the transient by means of a low impedance buffer. Hardly feasable if at all given the voltages involved.

Alternatively, the shielding could be disconnected from GND (high impedance state) during the transient and then reconnected to GND once the voltage level drops below a few volts. This could be realized by some sort of voltage dependent resistor, or perhaps even an inductor in series with the shield

Just an idea, I haven't tried or simulated any of this.

I obtained good results by critical damping the transient in the shield.

[Teleno]

I obtained good results by critical damping the transient in the shield.

Interesting. I guess you placed a resistor between the two ends, correct?

[Tinkerer]

Interesting. I guess you placed a resistor between the two ends, correct?

No, the resistor, actually a 10 turn trimpot goes between the cable shield and the TX or RX ground. Using twisted pairs (2 or3) plus shield.

[Teleno]

No, the resistor, actually a 10 turn trimpot goes between the cable shield and the TX or RX ground. Using twisted pairs (2 or3) plus shield.

Basically what you do is to increase the isolation between the shield and GND.

I believe it makes sense because a Faraday cage doesn't need to be attached to any potential in order to work. Or so I believe. What would happen if the shield is left floating? There's a patent claiming such a shield https://patents.google.com/patent/US20210375505A1/en

See also this discussion:

"You do not need to be connected to ground to shield something and this is a huge misconception among many people. Enclosing something in metal will stop all EM radiation from getting through

"I would agree with that because the test results for a floating shield still showed attenuation as I suspected."

[Teleno]

And there's more

"Some cables are built with a floating shield, which has no direct electrical connection to ground. In these applications the shield layer just covers the length of the cable and is cut or trimmed adjacent to the cable’s end. While in theory one would electrically tie the shield to ground, some cables don’t always need this to function and actually work better with this floating ground."

[Tinkerer]

Basically what you do is to increase the isolation between the shield and GND.

I believe it makes sense because a Faraday cage doesn't need to be attached to any potential in order to work. Or so I believe. What would happen if the shield is left floating? There's a patent claiming such a shield https://patents.google.com/patent/US20210375505A1/en

See also this discussion:

"You do not need to be connected to ground to shield something and this is a huge misconception among many people. Enclosing something in metal will stop all EM radiation from getting through

"I would agree with that because the test results for a floating shield still showed attenuation as I suspected."

Interesting.
However, we also use the cable shield to discharge any electrostatic charge of the coil. Sometimes the electrostatic charge builds up on a detector and then discharges when touching the ground or even grass. I usually test the coil shielding by touching a PVC pipe that I previously electrostatically charge to several thousand volts to the coil. If I don't notice any reaction by the detector, I consider the coil shielding to be good enough.

Does anybody have a better suggestion?