![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Computer PSU RFI?
- Thread starter DWD1961
- Start date
Nobu
[H]F Junkie
- Joined
- Jun 7, 2007
- Messages
- 10,050
In the air, it shouldn't be too bad–they are almost entirely enclosed in a grounded metal box. A bit of rf might escape through the fan hole if the noisy bits are exposed there and it has a simple wire fan guard, but if it's punched holes then that should be shielded pretty well, too. That'd just leave the wires, but anything going through there should be pretty weak, I think.
The biggest source of EMI/RFI is going to be the mains plug. If the PSU doesn't have a good transient filtering stage, it's going to backfeed the mosfet switching frequencies into the mains. If it doesn't have one at all, which is common on shitty power supplies, it's going to be an RF blaster. The DC side can also be a source of RFI if the output filtering stage isn't good, or has a problem. Again, shitty supplies, but bad capacitors can also create an RF blaster. I've had ATX PSUs with bad capacitors blast out so much RF noise that nearby radios would pick them up and CFL lights would start to flicker.
Other RF blasters are "wall warts" of the SMPS variety. They almost never have a transient filtering stage, or if they do, it's usually not very good. The smaller they are, the less likely they are to have a transient filtering stage because of the size and space requirements of transient filtering components. But even the classic linear wall warts create EMI in the 50-60 Hz range, as seen by what they do to nearby CRT monitors making the picture go wonky and shake/make psychedelic colors. They can also be picked up sometimes as a 50/60 Hz buzz that sounds like a ground loop.
As for charts on EMI/RFI, you're basically on your own there. There are guidelines and regulations for manufacturers, but they won't really cover real world scenarios because EMI/RFI varies based on numerous conditions that can't all be quantified in a lab setting. Then there's also the fact that some manufacturers flagrantly ignore such rules/regulations, or have product design issues that prevent their EMI/RFI circuitry from working properly.
If you have some super sensitive RF application, I'd look into building a well grounded Faraday cage around said device and use a linear power source.
Other RF blasters are "wall warts" of the SMPS variety. They almost never have a transient filtering stage, or if they do, it's usually not very good. The smaller they are, the less likely they are to have a transient filtering stage because of the size and space requirements of transient filtering components. But even the classic linear wall warts create EMI in the 50-60 Hz range, as seen by what they do to nearby CRT monitors making the picture go wonky and shake/make psychedelic colors. They can also be picked up sometimes as a 50/60 Hz buzz that sounds like a ground loop.
As for charts on EMI/RFI, you're basically on your own there. There are guidelines and regulations for manufacturers, but they won't really cover real world scenarios because EMI/RFI varies based on numerous conditions that can't all be quantified in a lab setting. Then there's also the fact that some manufacturers flagrantly ignore such rules/regulations, or have product design issues that prevent their EMI/RFI circuitry from working properly.
If you have some super sensitive RF application, I'd look into building a well grounded Faraday cage around said device and use a linear power source.