As 'marce' and 'DF96' wrote, coax is the best choice for an unbalance interconnect cable. For short (1 or 2 meter) cables any shielded cable will work just fine. For long cables (well more than 3 meters) a coax cable with a low end-to-end resistance of the shield should be used. Look up 'Common Impedance Coupling'.
As for a shield connected at only one end. It will function as a shield (but there is fine print attached.
Multiple coax signal cables will not form ground loops, if you route them all along the same path.
As for a shield connected at only one end. It will function as a shield (but there is fine print attached.
Multiple coax signal cables will not form ground loops, if you route them all along the same path.
The main purpose of an audio screen is to act as an electrostatic shield, keeping hum away from the signal. It acts in conjunction with the low output impedance of the source and the relatively high impedance of stray capacitance.AndrewT said:
2GHz can be filtered out. 50Hz can't be filtered without losing some signal.
Sorry about that. This does:
https://www.fairchildsemi.com/application-notes/AN/AN-42036.pdf
How to have low noise and a radiotransmitter in the same device.
The TI article has a chapter about PCB trace antennas, one of the reason for stargrounding is to avoid those.
The whole idea boils down to ground near the signal and return + no loops that can make an antenna.
It also mentions:
"This shield is called a “Faraday Cage” and must be
carefully designed not to allow frequencies that are causing
the problem to enter the circuit. This means that the
shield must have no holes or slots larger that 1/20 the
wavelength of the offending frequency."
So 1 cm hole is letting FM signals inn.
So audio frequencies can be protected by a screen with small holes in it. And the higher frequencies I would filter out as most amps and loudspeakers do.
I dont think cables distinquish themself from the rest og the audioequipment. The screen is the case or ground plane. Signal is the signal and the returnpath is the extra connector in a two connector cable.
To save the extra connector is to connect the returnpath to the groundplane.
Last edited:
DC DC converters totally irrelevant to low level analogue layout. The terminology is a ground star point, you create a separate low impedance ground for the converter switching currents loops and converter noise loops (linear and LDO's also have noise as well as SMPS's) prefferably on the same layer as the components and connect this noisy plane to the main ground plane at a single star point so these unwanted currents do not have loops on the main ground plane. I can provide more on this if you want I do power supply layout for many critical designs.
Totally wrong and misguided, using star ground is more likely to make the problem worse.... If you look at the diagram regarding antennas (Figure 6 in the Ti document) you will notice that the full ground plane gives the best results, separate ground routing as traces not a ground plane is worse. This can be critical with low level analogue layout as the traces can pick up noise that is superimposed on the signal as extra noise. This is more of a concern when the analogue circuitry is spread out instead of crammed together to minimise trace lengths and thus any loop areas that may act as antennas.
I would suggest you look further into shielding (2.4 and 5GHz is not uncommon in the domestic environment) these days it is down to slots and seams, use as a search with EMC and you'll get plenty of results.
This may be a more relevant paper to start of with...
http://www.x2y.com/filters/TechDay0...log_Designs_Demand_GoodPCBLayouts _JohnWu.pdf
I have many more such documents covering all aspects of critical layout and interconnection.
To me a loop is a loop is an antenna....
I like to learn from people that have to deal with harsh enwiroment to see what they do. The 2.4 and 5gHz is even inside the IPad.
From my knowledge you have to have something that fold the noise down in the audioband i can't see that in a cable if you have a normal god solderings, ground close to signal, normal screen and one ground connection (not many) between the components.
About stargrounding, thank you for your insigt I cant see the contradictions. The stereo components are the system components that should be connected together in preferably a star ground configuration as you dont have a goundplane round your hifirack
I like to learn from people that have to deal with harsh enwiroment to see what they do. The 2.4 and 5gHz is even inside the IPad.
From my knowledge you have to have something that fold the noise down in the audioband i can't see that in a cable if you have a normal god solderings, ground close to signal, normal screen and one ground connection (not many) between the components.
About stargrounding, thank you for your insigt I cant see the contradictions. The stereo components are the system components that should be connected together in preferably a star ground configuration as you dont have a goundplane round your hifirack
I do PCB layout for mission/life critical designs and have done for 30+ years.... Working environments don't get much harsher than where some boards I design go.
So I do know what I am talking about here...😉
Ground planes are for PCBs...
Again look at the Ti document you posted its all about minimising loop area and minimising the noise pick-up. I would also suggest you read "Dipoles for Dummies" Henry Ott another good read.
I was pointing out the frequencies of RFI you will encounter in a normal domestic environment... What has an IPad to do with that....
The cable was to be designed for ipad.
Here is an output of a lowest of lowend hifi cable connected to an ipad. Looked at with a 100Mhz oscilloscope.
The ipad sends a lot of HF noise, that is stronger than the cable HF pickup noise.
And that noise is is not strong enough to make any degradation of the sound or couple into the amp at any degree more than normal noise sources in the amp itself. This is just at my workbench result so take it as is.
Here is an output of a lowest of lowend hifi cable connected to an ipad. Looked at with a 100Mhz oscilloscope.
The ipad sends a lot of HF noise, that is stronger than the cable HF pickup noise.
And that noise is is not strong enough to make any degradation of the sound or couple into the amp at any degree more than normal noise sources in the amp itself. This is just at my workbench result so take it as is.
Tor, you are producing sufficient evidence to show you should be putting the ipad in the bin and not into your audio system.
Can we assume that the ipad output is actually a tiny (and probably filterless) Class D amp for driving headphones? If so, worrying about cable quality is a complete waste of time. Any old rubbish cable will do, as almost anything except the weirdest DIY design will pick up less RF than the ipad is already putting out. What is needed is a low pass filter, to keep all the ultrasonic stuff away from the amp input.
From other handheld designs I have seen and some of the IC solutions to save space the headphones are often a part of the filter, in fact I believe a thread a few months ago discussed something similar, try looking for filter less class D or similar.
Here is one I have seen on a board, minimal extra components:
http://www.ti.com/lit/ds/symlink/tpa6130a2.pdf
In these very small handheld devices there is not a lot of room or scope for really isolating sections of circuitry.
Here is one I have seen on a board, minimal extra components:
http://www.ti.com/lit/ds/symlink/tpa6130a2.pdf
In these very small handheld devices there is not a lot of room or scope for really isolating sections of circuitry.
So maybe the best cable will have a integrated transformers in it (the headphone driver should be able to drive low impedance). Then the output can even be balanced🙂
Waste of money, but fun.
Regards Torgeir
Waste of money, but fun.
Regards Torgeir
- Status
- Not open for further replies.