Where abouts it mentions RF immunity being increase with a low impedance path to earth, I couldn't find it in the text.
Or the opposite.
Many may agree that unbalanced systems are helped by the fact that the chassis are normally not earth grounded. This allows an entire unbalanced system to float with respect to earth ground. This eliminates the potential for multiple return paths for the audio grounding system, since there is not a second path (ground loop) through the
earth ground conductor.
Many may agree that unbalanced systems are helped by the fact that the chassis are normally not earth grounded. This allows an entire unbalanced system to float with respect to earth ground. This eliminates the potential for multiple return paths for the audio grounding system, since there is not a second path (ground loop) through the
earth ground conductor.
This makes sense to me, even if there is RF in the chassis the fact that 0V is connected to it means the system as a whole (including the chassis) is floating? And thus immune? The chassis could still all be connected to PE I don't see a problem with that, easy enough to avoid a loop there surely?
Last edited:
I just took a quick look at the eleven page 1994 AES paper by Stephen Macatee that the Rane Note is based on. My take is that 'earth ground' refers to the AC power Safety Ground/Protective Earth system. Not the connection to Planet Earth.
Will reread the entire ASE paper and the Rane Note this evening.(maybe)
Will reread the entire ASE paper and the Rane Note this evening.(maybe)
I am getting 95 dB S/N (measured with a QA401) without a ground lifter, single ended input. The chassis, power earth and the system 0V are all connected at a single point. The power transformer is 1200 VA and the associated smoothing capacitors are 88 000 uF - there are huge charging currents and lots of scope for problems.
I think the point made by others earlier that the chassis ground is not part of the audio ground is an important one - AndrewT has summed it up nicely.
I would stress that any kind of conductive metal enclosure, provided the incoming interconnect shields are connected to it at the point of entry with a small capacitor, offers important RF immunity. Use a BW limiting filter on the amplifier input (I usually use 250-500 kHz) and you should have the RF thing covered.
If you use a ground lifter, you must decouple the lifter bridge with a small capacitor (a few nF ceramic).
Most hum issues in DIY gear arise from:-
1. excessive loop areas - especially where high currents are involved - for example the connections between the transformer > Bridge > smoothing cap. Minimize the loop areas by tightly twisting the two connections together
2. Common impedance coupling issues. Make sure to 'T' off from the central ground where the smoothing caps are joined
3. Inter-equipment earth loops caused by stray transformer fields and a failure to ensure these are routed away from amplifier inputs and forced to flow in the ground connections (hint: always use a 'hum breaking resistor')
4. Very important point wrt amplifier PCB design: Connections for power rails, ground, speaker output and speaker return must be tightly bunched together. This minimized loop area. Never run a ground loop around you PCB - use a star connection for ground on the PCB. Failure to do this will almost invariable lead to hum issues
I think the point made by others earlier that the chassis ground is not part of the audio ground is an important one - AndrewT has summed it up nicely.
I would stress that any kind of conductive metal enclosure, provided the incoming interconnect shields are connected to it at the point of entry with a small capacitor, offers important RF immunity. Use a BW limiting filter on the amplifier input (I usually use 250-500 kHz) and you should have the RF thing covered.
If you use a ground lifter, you must decouple the lifter bridge with a small capacitor (a few nF ceramic).
Most hum issues in DIY gear arise from:-
1. excessive loop areas - especially where high currents are involved - for example the connections between the transformer > Bridge > smoothing cap. Minimize the loop areas by tightly twisting the two connections together
2. Common impedance coupling issues. Make sure to 'T' off from the central ground where the smoothing caps are joined
3. Inter-equipment earth loops caused by stray transformer fields and a failure to ensure these are routed away from amplifier inputs and forced to flow in the ground connections (hint: always use a 'hum breaking resistor')
4. Very important point wrt amplifier PCB design: Connections for power rails, ground, speaker output and speaker return must be tightly bunched together. This minimized loop area. Never run a ground loop around you PCB - use a star connection for ground on the PCB. Failure to do this will almost invariable lead to hum issues
Last edited:
I disagree.
I intentionaly use multiple connections to the chassis which is the audio ground. I use ground plane on pcb connected to chassis on multiple points with no problem of hum, even with 2*0.1F reservoir capacitors.
This point is dogmatic. There is no technical reason for not using the chassis as ground.
How will this not hurt the common impedance coupling problem of currents flowing down the interconnect cable shield/screen?
These currents will have to go through the audio circuit common to get to the chassis.
There is a lot of information in http://www.diyaudio.com/forums/diya...udio-component-grounding-interconnection.html which I posted at the start of the thread. I'm pretty sure it was there that I read of very low impedance connections between the separate chassis' and that parallel connections are not necessarily loops
Do the most sensitive apparatuses for noise and distortion measurements respect this ?
Occam's Razor. You are making things difficult for yourself - and there is a ton of evidence produced by all sorts of experts that make it clear this is not the optimum way to do things at audio frequencies. But, if you are happy with your methodology, then go ahead.
Last edited:
Keep the audio signal loop areas small and use a HBR to keep ground loop currents out of the shield.
Common impedance coupling typically arises when you make a connection (in this case ground) where charging or signal related currents cause an error signal. This is normally a problem inside the equipment
You are creating a loop forr. I've ended up 'breaking' it on a few PCB's and getting noise improvements. You also set yourself up for common impedance problems doing this.
I'm pretty sure Occam's razor would lead to a solution where the connectors ground would simply connect to the chassis rather than insulate them and then install a capacitor for a less effective RF protection !
Authority argument. I know a lot of "experts" who work in electronics where frequencies include audio spectrum, none of them wouldn't use the chassis as ground, and none of them would disconnect ground connector to the chassis.
I have good reason to be happy, it is simpler and it works fine.
Last edited:
I don't think I've heard a reason yet why signal return shouldn't be referenced to chassis.
Last edited:
i'm finding this thread great 🙂
might have my wires crossed on the above statement but, the Earth wire in household wiring goes to a big copper/brass stake that has been driven into the ground somewhere in your house. so yes it is connected to planet earth.
Andrew keeps bashing on about PE but he does so for a good reason, you cant listen to music if your dead.
something a lot of audiophiles and indeed us morales with no cash forget is that your consumer unit needs servicing every now and again as the wiring becomes loose on the bus bars. before i had a new one fitted mine was buzzing like mad and the electrician said that a few of the wires were actually arcing. that cant be good for hifi.
- Status
- Not open for further replies.