The subject of gnd, shield, RF drain seems to have many opinions, I reached out to Garth Powell for his take:
"For audio, if it’s any variant on twin or Tri-axial (the shield is not carrying primary signal), you telescope for induced RF noise ALWAYS! One side only is terminated. The shield in this instance is a RF drain, and you will have circulating induced RF noise with both sides terminated. Not to mention it would make the shield a primary signal carrying portion of the cable for a single-ended RCA circuit. Not desirable.
"you want the termination on the load for everything but AC cords or power supply cords. The reason is that you want to drain towards the highest signal level, where induced low-level RF noise will do the least damage. If they were referencing DC power supply cords, then they are correct, shield terminated to source or supply chassis."
I have respectfully asked if he has measurements to support this and where does the noise drain to?
"For audio, if it’s any variant on twin or Tri-axial (the shield is not carrying primary signal), you telescope for induced RF noise ALWAYS! One side only is terminated. The shield in this instance is a RF drain, and you will have circulating induced RF noise with both sides terminated. Not to mention it would make the shield a primary signal carrying portion of the cable for a single-ended RCA circuit. Not desirable.
"you want the termination on the load for everything but AC cords or power supply cords. The reason is that you want to drain towards the highest signal level, where induced low-level RF noise will do the least damage. If they were referencing DC power supply cords, then they are correct, shield terminated to source or supply chassis."
I have respectfully asked if he has measurements to support this and where does the noise drain to?
I used to work in communications with rack wiring. Our audio cables, twisted pairs most often, had the shield grounded at one end only at the load end as stated by Mr. Powell. The cables had a drain wire and that was terminated to a multipoint signal ground in the rack. The audio cable grounding on one end really helped keep RF interference down especially near microwave and radar signals. No measurement, but at a radar facility, the drain wire would draw an arc when removed though it was connected only on one end. It carried a potential just being in the microwave field.
The power ground went to a separate ground. Power cables shield or conduit were terminated on both ends for grounding of faults.
With access to scopes (with storage of data) and RF spectrum analyzers, I was able to see the RF get reduced on the audio signal when the drain wire was connected. It was a substantial improvement. Sorry, no data capture or screen grabs. I did that work before we had cell phones and wifi.
The power ground went to a separate ground. Power cables shield or conduit were terminated on both ends for grounding of faults.
With access to scopes (with storage of data) and RF spectrum analyzers, I was able to see the RF get reduced on the audio signal when the drain wire was connected. It was a substantial improvement. Sorry, no data capture or screen grabs. I did that work before we had cell phones and wifi.
100% agreed with runclesid. Did also wirings in cabinet, 1000’s of connections, going everywhere, servo motor drive signals, with a lot of RF, spikes, name it, high power radar, the way he described the wirings, better than I could do, was the way to do it.
Used to worked in TACAN transmitters site, with huge transmitter tube. There was a neon tube on the transmitter cage, connected to nothing, but that was glowing when the transmitter was on… if you get into the transmitter cage when the transmitter was on, you could be fried. I know, the tech before me was killed when he bypassed the kill switch to troubleshoot the circuit…
That was the kind of high RF energy environment where perfect shielding and grounding was ’very important’, if not critical.
Our simple home environment audio setup are not that critical, to say the least.
I once built a phono preamp for a friend. He called me back complaining about the noise. I asked him did you place the phono supply away from the phono preamp, he said yes. Had to drive back to his place, the supply was indeed on the ground away from the phono, that was sitting on the huge power amplifier, with maybe a 800-1000VA power transformer, go figure he had induced noise in his phono preamp…
SB
Used to worked in TACAN transmitters site, with huge transmitter tube. There was a neon tube on the transmitter cage, connected to nothing, but that was glowing when the transmitter was on… if you get into the transmitter cage when the transmitter was on, you could be fried. I know, the tech before me was killed when he bypassed the kill switch to troubleshoot the circuit…
That was the kind of high RF energy environment where perfect shielding and grounding was ’very important’, if not critical.
Our simple home environment audio setup are not that critical, to say the least.
I once built a phono preamp for a friend. He called me back complaining about the noise. I asked him did you place the phono supply away from the phono preamp, he said yes. Had to drive back to his place, the supply was indeed on the ground away from the phono, that was sitting on the huge power amplifier, with maybe a 800-1000VA power transformer, go figure he had induced noise in his phono preamp…
SB
My latest version of the Pearl preamp has a ground lift built-in (10R, 0.1uf, 0805 smd parts built-in)
you can use it or not, your choice…
sB
you can use it or not, your choice…
sB
For me I think of GND as a place to grow potatoes
😉In an amplifier, it helps me to think of signal return, power return and safety earth instead.
https://web.mst.edu/~jfan/slides/Archambeault2.pdf
But for a power cable connection between to chassis then the shield can act as a "continuation" of the chassis (safety earth) between the two chassis. That was my understanding for how Passlabs use the power cable shield when they have spilt chassis with power supply in one chassis and amplifier in the other chassis.
It is not the same application as signal cables like e.g. RCA cables.
It is not the same application as signal cables like e.g. RCA cables.
Have a Pearl 3 coming 
, chassis from @Gianluca and am part of the group buy CRCRC. From what I can tell I will need to source some parts for the umbilical cable, got a lot of scraps of cables at work and would like to know what I should be looking for, other than shielded 2 conductor?
, chassis from @Gianluca and am part of the group buy CRCRC. From what I can tell I will need to source some parts for the umbilical cable, got a lot of scraps of cables at work and would like to know what I should be looking for, other than shielded 2 conductor?The P3 power cord between chassis (a Canare star quad kind of thing) has 4 wires plus a shield. The red and pink wires are connected in parallel. The white and translucent wires are also connected in parallel. One of the parallel pairs of insulated wires is positive and the other pair is negative. The "shield" is a grounded current carrying conductor. The negative voltage supply carries more current than the positive supply.
So no the shield is not an extension of the green wire or safety ground.
Thanks DT
OK!
I have found a 3-wire (0.5mm2) + shield (4. wire) I will use for the power connection. Then I can use the shield for the chassis (safety ground).
I have found a 3-wire (0.5mm2) + shield (4. wire) I will use for the power connection. Then I can use the shield for the chassis (safety ground).
I read this description by Wayne (now many posts back) as a recommendation:
"The shield in current Pass Labs pre products is used as a safety ground and must pass at least 20 amps and the cable be approved by the appropriate agency."
"The shield in current Pass Labs pre products is used as a safety ground and must pass at least 20 amps and the cable be approved by the appropriate agency."
The shield goes to gnd, this is for safety? Based upon what Garth has said doesn't that inject noise into the signal?
I read it as shield is "just" the yellow/green wire that has chassis connection. How Passlabs products connect audio gnd to chassis I don't know if it is a direct connection or via a resistor, NTC, bridge, capacitor etc. The NTC approach has worked for me so far.
My Lazy Singing Bush mono block amps have external PSUs. Also here the PSU cable has yellow/green wire in cable to power amp and in power amp audio gnd is connected to chassis ground point via a NTC. This works perfect and why not do the same for the P3. Then there is no connection from audio gnd to chassis in PSU chassis (only yellow/green wire connected to chassis ground point).
Hello,
I suspect that the 20 amp requirement is for ground fault conditions not for normal operation.
Thanks DT
Yes, that is "just" to be certified. Don't know how many amps the shield+wire can take in the microphone cable. With 0.5 mm2 5-6A should be possible and peak current probably more. I know from my 2000W water heater that from 230 VAC and 0.75mm2 mains cable it gets a bit......not hot but just that you can feel some heat. I think it is 0.75 and not 1.5.
A few observations:
1) If there is a continuous conducting shield there can be no electric field inside, grounding in whatever form is irrelevant. Skin depth complicates this a little, but for 1MHz in copper the skin depth is 65µ, at 10kHz 650µ.
2) Grounding is important if you are concerned about static charging or static dissipation.
3) Shielding/grounding does essentially nothing for magnetic fields. Most of the EMI from frequencies in the 100's or Hz of less is magnetic.
4) The only practical effective means of attenuating magnetic fields are high permeability metals like mu-metal or physical separation distance.
5) Large rf fields can induce large skin currents. If it is a continuous conducting surface these will only be on the outside. They will have no effect on anything internal to the surface. Dealing with large induced skin currents is an engineering problem depend on the specific circumstances.
6) Holes that ventilate an exclosure also compromise the electric field integrity. A rule of thumb is to get good rf attenuation the aspect ratio, length to diameter, should be 3x or longer.
7) Safety grounds are an engineering solution to a different set of requirements.
Now go design a solution for your particular circumstances.
1) If there is a continuous conducting shield there can be no electric field inside, grounding in whatever form is irrelevant. Skin depth complicates this a little, but for 1MHz in copper the skin depth is 65µ, at 10kHz 650µ.
2) Grounding is important if you are concerned about static charging or static dissipation.
3) Shielding/grounding does essentially nothing for magnetic fields. Most of the EMI from frequencies in the 100's or Hz of less is magnetic.
4) The only practical effective means of attenuating magnetic fields are high permeability metals like mu-metal or physical separation distance.
5) Large rf fields can induce large skin currents. If it is a continuous conducting surface these will only be on the outside. They will have no effect on anything internal to the surface. Dealing with large induced skin currents is an engineering problem depend on the specific circumstances.
6) Holes that ventilate an exclosure also compromise the electric field integrity. A rule of thumb is to get good rf attenuation the aspect ratio, length to diameter, should be 3x or longer.
7) Safety grounds are an engineering solution to a different set of requirements.
Now go design a solution for your particular circumstances.
If there is a continuous shield that connects the two boxes then the circuit as shown shorts out the ground breaker in the PS chassis, hard connection of RIAA board GND through RIAA chassis through the shield to PSU chassis to power line earth.
The power supply has three voltages, V+, V-, and V0. There is also VE, the potential of the third pin of the power plug. Using the canonical ground breaking scheme, V0 is within a diode drop of VE, but it probably isn’t equal to VE. It isn’t immediately clear how much current flows over V0. If the circuit is symmetric, very little, but if it isn’t. But if there is a shield effectively in parallel with V0 then in gets complicated to ensure no current through the shield.
So there are starting to accumulate a lot of unknowns.
I happen to know that my cartridge shell and tone arm are essentially floating metal. So I would start by tying those to shield and finally back directly to earth. Now any static buildup will be earthed. The RIAA ground is connected to V0 which ties back to earth via the diode bridge and possibly/probably a parallel 1-10M Ohm resistor to bleed V0 back to VE.
That and keep both the RIAA PS and the RIAA amp physically separated from the power amplifiers and the RIAA separated from its power supoly. But since the currents here are only 10s of mA the physical separation needed is not as much.
You are welcome!
Last edited: