PS/Amp isolation in 1 box ?

[jimbo1968]

How important is electric/magnetic isolation and how is it achieved?

Which are the parts that need to be isolated in order of importance? 240V AC in, transformer, rectifier, amplifiers.

Least good solution? has everything in a common space. Physical separation, shielded wires being the only tools available?

Best one box solution? has AC input, transformer, rectifier in 1 compartment, amplifiers in another compartment.

If the psu is in a steel compartment, how sealed does it need to be? There are obviously holes to let the DC power out, so do more holes/gaps just reduce the isolation. Does the RF get absorbed by the steel walls or does it bounce around the compartment until it finds a gap to escape through?

[Nuuk]

While it is sensible to try and keep items apart that may interfere with one another, in practise, it doesn't seem to matter as much as we may think.

To 'block' the magnetic effects of a transformer you need a very thick piece of metal and putting a traffo in a 'sealed (thin) metal box' won't do.

Try and keep your AC part of the circuit away from the DC and particularly the signal carrying parts and you should be fine.

[jimbo1968]

With my split level design (3 stereo units for driving 3way speakers) everything can be separated just fine but it makes assembly/disassembly more tedious.

Does the bolt through the transformer carry any fields? ie if I bolt the tranformer to the wall that separates amp and psu compartment, will that be worse than bolting it to an external wall?

[Nuuk]

Does the bolt through the transformer carry any fields? ie if I bolt the tranformer to the wall that separates amp and psu compartment, will that be worse than bolting it to an external wall?

I don't think that it would make a difference but you could 'play safe' and use a nylon bolt (or cable ties)!

[jcx]

you need a really thick layer of conductive metal (Al or Cu) to absorb/reflect line frequency (and rectified harmonics) but if you use iron (or low carbon steel) which is a good magnetic flux conductor at these frequencies you can easily "short out" magnetic flux from a power supply with a reasonable gauge sheet steel box – look at lambda’s old power supply packaging, perforated sheet steel box allows air flow but is quite effective in shorting out the magnetic leakage flux of the transformer

low frequency magnetic field control is about using the high reluctance of air in a flux path and conducting (with iron, steel, mu metal) the small flux that crosses the air path(s) away from your sensitive circuitry

look at commercial power amps and you will see that toroids can be packed pretty tightly since they have little leakage flux, I think the greater line isolation of E-I transformers is more worthwhile (do you know which manufacturers put drawn metal cylinders over their E-I power transformers because consumers have been sold on the false idea that toriods are “better” – they are better for the manufacturers bottom line)

[Shoog]

Don't worry about it. A toroidal would have to be sitting on top of your chip to make it an issue. Grounding is really where noise comes from and this is where you should devote your energy.

Shoog