Electronics > Beginners
EMF Neutralisierer, real or tinfoil hat?
coppercone2:
look at military EMI filters for TEMPEST grade $$$
Brumby:
Another Idea is to set up your workbench so that you can "rig for silent running". By this I mean you identify every bit of RF noisy gear and be able to remove power from them when you need a low noise environment. It might be worth trying out Dave's $10 EMC probe idea to see what's noisy.
Desktop computers would need to be unplugged as even when they are "off", there is power flowing so that the soft on power button can work. If you want to be critical, view tablets with suspicion and take the batteries out of any laptops and phones.
Electric motors in vacuum cleaners, fans and kitchen appliances (just to name a few) can be really ugly if they are on the same circuit as your lab, but even if they're not, just being nearby could be enough to ruin your day. And if you've got someone doing arc welding in the area - go do something else until they're finished.
For lighting, use an incandescent globe or LEDs powered by a linear constant current source from a battery or simple old school transformer power supply. LEDs and steady DC are a brilliant combination.
If you need to run a scope which has a switching supply, then I suggest you try an Dave's EMC probe to find out just how much hash it is introducing into the environment - and what it looks like - so that you can take that into consideration.
There's a lot more that can be said - but you get the idea.
FriedMule:
Thanks for the great tips.
Yes it is about to kill all noise sources but some is properly hard to be without.
When Dave do tear down of more expensive gear, there are often some sort of metal shielding, to shield against RF and EMF, would it not be possible to make a box and place the SMPS inside that box?
dmills:
The trick with screening boxes is that any slits or openings must be thought about carefully (they form radiators at a high enough frequency), and every electrical connection penetrating that box must be filtered in a way appropriate to the problem you are trying to solve.
Also the metal must be thick enough and of sufficient conductivity to make the skin depth at whatever frequency you care about much less then the thickness, this is HARD at low frequency.
Bringing all the wiring out from the same side of the box is a good idea, as is keeping the internal doings only connected to the box at that edge (reduces circulating currents in the box walls).
Really switchers are fine, even for low noise work, you just have to do the sums to see what you need to do to make the power supply noise a non issue, and then follow thru.
Now filtering the output across a sufficiently broad band, that can be tricky, a post switcher LDO helps, but usually the PSRR goes to pot surprisingly quickly (But they will deal with most of the annoying DC - 1MHz or so region), then LCR filters for the quick stuff.
Regards, Dan.
radiolistener:
all depends on what is your requirement for low noise.
If you're need to work with signal levels about -40..-50 dBm, then switching power supply can be shielded with no problem.
If you're need to work with -80..-110 dBm, then shielding may be complicated or even problematic.
For signal levels -130..-150 dBm even good quality and good shielded switching power supply will be a problem.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version