Serious apparatus having plastic boxes shoud have, at least, shielding metallic sheets used as chassis inside.
Imagine two coxial cables with their shields bonded all the way : no loop.
It seems to me that problems relative to hum and problems relative to RFI should be and can be prevented in an ideal manner ideal at the same time.
For sure, the idea must already have been tested but I don't know of such a cable.
Not too difficult to build one, a bit more to make a protocol for compare it with more standard configurations. When time permits...
"Imagine two coxial cables with their shields bonded all the way : no loop."
You now have too loops. The original loop is bisected down the middle. Whether the noise is less or more is besides the point - the same EMF generation and the same noise reduction techniques still apply.
You can buy twin core shielded cable which could be used to interconnect equipment - although I have not seen this used in practice.
"Serious apparatus having plastic boxes shoud have, at least, shielding metallic sheets used as chassis inside. "
They are audio signal generators - 3Hz to 3 MHz. They have no shielding and the BNC connectors are not connected to a metal chassis. At RF, this may be required to pass EMI regulations. The point is, you cannot generalize about connectors having to be directly connected to a metal chassis.
"It seems to me that problems relative to hum and problems relative to RFI should be and can be prevented in an ideal manner ideal at the same time. "
I think a better way to understand it is to separate the two issues and solve each using best practice. If there is any duplication once that is done (which I have not found in practice) then you can make any necessary adjustments.
You read many articles where people conflate problems at HF (10's MHz to GHz) with problems at audio. You don't need to put audio circuits into metal boxes to necessarily get good hum reduction - there are plenty of audio products in plastic housings that prove this. Manufacturers use metal for audio because it is cheap and easy to manufacture with e.g. a pressed and punched metal chassis for an AV receiver or a CD player costs <2 $. Another example: you need screened cable interconnects to prevent hum. You don't - a screen has no effect at 50Hz and even less so because the screen is almost always copper - a non-ferrous metal. Twisted wires that minimize loop area of the interconnect are just as good. However, screened cables are necessary to reduce or prevent the ingress of RF. The screen should be electrically connected to the chassis at RF frequencies. A 1nF cap at the connector entry point from the screen to the chassis will do this.
You now have too loops. The original loop is bisected down the middle. Whether the noise is less or more is besides the point - the same EMF generation and the same noise reduction techniques still apply.
You can buy twin core shielded cable which could be used to interconnect equipment - although I have not seen this used in practice.
"Serious apparatus having plastic boxes shoud have, at least, shielding metallic sheets used as chassis inside. "
They are audio signal generators - 3Hz to 3 MHz. They have no shielding and the BNC connectors are not connected to a metal chassis. At RF, this may be required to pass EMI regulations. The point is, you cannot generalize about connectors having to be directly connected to a metal chassis.
"It seems to me that problems relative to hum and problems relative to RFI should be and can be prevented in an ideal manner ideal at the same time. "
I think a better way to understand it is to separate the two issues and solve each using best practice. If there is any duplication once that is done (which I have not found in practice) then you can make any necessary adjustments.
You read many articles where people conflate problems at HF (10's MHz to GHz) with problems at audio. You don't need to put audio circuits into metal boxes to necessarily get good hum reduction - there are plenty of audio products in plastic housings that prove this. Manufacturers use metal for audio because it is cheap and easy to manufacture with e.g. a pressed and punched metal chassis for an AV receiver or a CD player costs <2 $. Another example: you need screened cable interconnects to prevent hum. You don't - a screen has no effect at 50Hz and even less so because the screen is almost always copper - a non-ferrous metal. Twisted wires that minimize loop area of the interconnect are just as good. However, screened cables are necessary to reduce or prevent the ingress of RF. The screen should be electrically connected to the chassis at RF frequencies. A 1nF cap at the connector entry point from the screen to the chassis will do this.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.