Specific guidance is really what I need most at this stage, so thank you. I am also trying to gain greater understanding of the various aspects of design etc. How is a non-connected shield a floating gnd? Wouldn't it work a bit like an antenna?
Yeah, what he said. Oh well experimentation is often the first step to understanding.
Yeah, what he said. Oh well experimentation is often the first step to understanding.
Hopefully I'm not steering this too far a field, but the whole subject of noise in audio circuits is pretty interesting. Logically and experientially I have that the earlier in the chain that it can be reduced the bigger the pay-off, and in a phono pre-amp with such low signal and high gain well...I'm pretty certain WC did his job of minimizing it where possible which of course leaves the PS.
Which is quite important as it is about as early in the chain as it gets and carries a large voltage in comparison to the millivolts from a cartridge.
Which is quite important as it is about as early in the chain as it gets and carries a large voltage in comparison to the millivolts from a cartridge.
As I was an early builder before the chassis were made, and am a bit on the cheap side, I didn't pull the trigger on the beautiful chassis set, but made up my own. My intent was to evoke but not copy the design. Unfortunately, it was too difficult to set up my lathe/milling machine to cut a groove in one pass, so that didn't happen. 🙁
However, it came out great, is much more compact than before, and looks really nice to me. The power supply is hidden deep in a lower shelf, so the appearance wasn't as critical to me, and I needed to make it very small since I'm running out of room. Photos are before I put labels on the rear panel.
However, it came out great, is much more compact than before, and looks really nice to me. The power supply is hidden deep in a lower shelf, so the appearance wasn't as critical to me, and I needed to make it very small since I'm running out of room. Photos are before I put labels on the rear panel.
Current induced along the length of the conductor?
I have used this approach on many power supplies, both AC and DC, and results have been very good so far. The shield is grounded at earth potential at the PS end, and the intent, at least, is that the shield serves as a bit of a Faraday cage.
This is the same approach I use for all of my interconnects; I know it's not the same as a power supply cable, but it works very well for my interconnects, and my system does not have hum.
My concern is the effect of connecting the preamp chassis to the shield ground that is traveling 6' to the PS box.
This shield at one end only approach is used pretty extensively in high-gain guitar amps, and since that's where I started, it seems to be a holdover that is working for me. Of course, I could always try to measure hum with the shield configured as-is, then connect it at the phono chassis and see if I can measure the difference. It is certainly easy enough to wire up one way, test, then wire up another way and test.
In my experience, the biggest impact on hum is eliminating ground loops.
Last edited:
In the retail world this is convential wisdom, which is not always wise, and a lot of co-ax digital cables are designed this way.
Is hum the only potential contaminant, reducing EMI/RFI can help lower the noise floor?
I know that I don't have anything to truly contribute to this discussion so will sit back and pay attention.
Old Counterpoint, I imagine you've been inside that, very nice units but not always that stable.
Good point.
At this time, beyond reporting how quiet this is, I don't have any measurements to compare connecting the shield at both ends, or only the PS end as I've done.
It has been very effective at eliminating hum, EMI/RFI in my entire system, but again, this is with signal interconnects, not the PS umbilical we seem to be obsessing over here.
My approach with interconnects is to use quad-core cables, either Mogami or Canare, and use two connectors each for positive and negative, and the shield grounded at the source only. Whether is conventional wisdom or not - I haven't seen that to be the case - most have shields connected at both ends or neither in my experience. Arriving at this method has been an iterative process for me, as I have worked to make the background noise and interference minimal or inaudible.
It is so easy to try different approaches, so it should be encouraged. However, there are so many interactions with different systems and components that one method may not work for everyone. Just try it and see, is my thought. For my own build, I'm happy to experiment, but when it sounds so good, I'm not in a hurry to make changes.
@Toys4Boys - that has been repaired, restored, and upgraded by our own Anatech. He does a lot of work that reduces the noise floor and increases stability. I installed a circular connector for the remote power supply, which, by the way, is putting out high voltage as you know, and has no shield around the numerous conductors.
When I think about cable shield grounded at one end I agree with Jim. Where does the current flow? How many one leaded semiconductors exist?
To shield at low frequencies would require very thick metal and not some thin shielding. Soldering at one end prevents ground loops but doesn't offer much shielding. You can solder the shield at one end and connect the other side with a small film cap to get shielding at higher frequencies and avoid DC loops.
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.
To shield at low frequencies would require very thick metal and not some thin shielding. Soldering at one end prevents ground loops but doesn't offer much shielding. You can solder the shield at one end and connect the other side with a small film cap to get shielding at higher frequencies and avoid DC loops.
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.
Outstanding! Thank you so much, Wayne!
If I use a 4-pin connector, I could connect the shield to the small film cap as you advise, not to the shell of the connector, correct?
I'm pleased to learn a better approach.
If I use a 4-pin connector, I could connect the shield to the small film cap as you advise, not to the shell of the connector, correct?
I'm pleased to learn a better approach.
To be clear, are you suggesting that this is something worth doing, or just thinking about it, not that they are mutually exclusive.
@wayne a friend of mine once designed a circuit to turn a MC to a totally capacitive load, I think that I still have it in storage. He used to say that if you hooked it up to a light bulb you could use it as a very sensitive mic. Just wondering if this is something that you are familiar with?
I have expressed it elsewhere but would like to personally thank you for this wonderful Xmas gift that you have given the community!
I have expressed it elsewhere but would like to personally thank you for this wonderful Xmas gift that you have given the community!
Last edited:
I make my cable shielding out of solder braid. I’ve never measured it, but it probably gives a good bit of attenuation from a few kHz up. Below that the coupling is probably magnetic anyway, which shielding doesn’t do much for, but twisting does. So twisted conductors inside a braid. This also easily meets Wayne’s 20A criteria. The connector backshells are conducting as well, so there is one continuous shield. I also use at least one connector pin to carry chassis ground because connector plating can make a low impedance joint somewhat problematic. I just run a bare wire between the chassis pins and it contacts the braid along its length. The shield does not need to be grounded to serve as a Faraday cage, only continuous. The reason to ground it is to dissipate any static buildup. If you walk across your wool carpet in the winter and touch your preamp, or anything else for that matter, you want those electrons to go down the shield and into the ground, not find their way somewhere else.
@ARGOS
Your cable design makes sense.
However, if
"...you want those electrons to go down the shield and into the ground, not find their way somewhere else."
and the ground post is part of that shield system, why does it make sense to connect the circuit board grounds to that "shield" terminal rather than run them back to the signal side of the bridge in the PSU? Or ground via the power supply?
If this was a one box design wouldn't they be wired directly to the bridge while the post would connect to the earth/chassis side of the bridge?
Your cable design makes sense.
However, if
"...you want those electrons to go down the shield and into the ground, not find their way somewhere else."
and the ground post is part of that shield system, why does it make sense to connect the circuit board grounds to that "shield" terminal rather than run them back to the signal side of the bridge in the PSU? Or ground via the power supply?
If this was a one box design wouldn't they be wired directly to the bridge while the post would connect to the earth/chassis side of the bridge?
The RCA cables which has shield only connected in one end (and therefor has a direction indicator) has already a twisted pair wires inside where the signal gnd is connected to the shield in one end. The reason for that is not to introduce ground loops or "destroy" group delay is my guess. And my guess is that the shield in this case should protect against incoming RF only. The RCA cables I have that is made like this the shield is not metal but a kind of conduction foil not made for heavy currents. But that is a signal cable and not a power cable.
I can see P3 kits still available. So I could have gone to bed early and still able to get a kit this morning.
I can see P3 kits still available. So I could have gone to bed early and still able to get a kit this morning.