Desktop PC interfering with broadcast TV

[JetIgniter2k]

My TV reception gets choppy every time my computer's CPU utilization goes above idle.  I made a makeshift near field sensor, placed it on top of the case and took some measurements.  It's dead quite in hibernate, noisy at idle, and flat out ridiculous with some moderate CPU utilization.  Largest spike around 125 MHz.  Any good ideas on how to mitigate this problem?  There is continuity between ground prong on a shared power strip and the power supply mounting screws.  I will assume a decent earth ground despite the horrible wiring in this apartment (main breaker kept tripping due to loose connection!!!!!)

[Rasz]

1 search for 'Spread Spectrum' in bios, turn it on
2 try with another power supply

[Armadillo]

Either is the lost of Good Earth connection [can be continuity but not To earth] or a cheap non original power supply, example those without EMI/RFI screening copper plates.
Use the Original Power supply [not any another power supply] and check the earth connection.

[JetIgniter2k]

I could believe it was the power supply but it is an 850W Thermaltake with FCC on the sticker.

I changed both the CPU and PCIe spread spectrum settings to "Enabled" but that had no effect.

I also cleaned up some of the chassis screw connections for better grounding.  I'm sill getting 17 Ohms between the side cover and ground though.  I'll try to clean that up tomorrow.  This is a performance case with lots of venting, so I might not be able to get great shielding anyways.

[Bashstreet]

Thermaltake Is one of the shittiest makers.

[Rasz]

I could believe it was the power supply but it is an 850W Thermaltake with FCC on the sticker.

doesnt matter, try with another one

[jmelson]

Is the power cord shielded?  You can get shielded cables, which help.  Alos, all the I/O cables (keyboard, mouse, monitor, USB, etc.) can also radiate.  Disconnect one thing at a time and see if the radiated signal changes.  It may NOT be the CPU box itself, but a crummy cable or I/O board that has no suppression at the connectors.

Jon

[Rasz]

pcie uses 100mhz reference clock, cpu has onboard fsb generator(unless older one), ram is much higher
1Gbit ethernet is 125 MHz, so you can also(after swapping supplies) try disconnecting network

[Armadillo]

Those operate on the DC and filtered capacitors especially these are supposed to be blocked by the input lines chocks + X and Y caps using the Ground Cable which should be the prime suspect as faulty Ground can also act as "emitting Antenna" or emitting body.
This assume that the power supply is working properly otherwise it is also a sign of some failures example very excessive power supply ripples.
Talking about Antenna. Maybe he should describe his TV Antenna System and the proximity of the two.

[Whales]

I have an Antec Earthwatts EA-380D with severe noise issues at the moment.  Can't listen to AM radio whilst it's on and near any cables (power, speakers, etc) connected to my radio.  Being on the same powerboard does not seem to be an issue, so it atleast looks like the input filters are working.

It's an 80plus bronze rated unit too.   A freshly bought unit from another brand gives me none of these issues.

[Armadillo]

Additionally, since the power supply is above 150W and Class D equipment, therefore it required to have the PFC circuit inside. Those multitudes of harmonics can range to RFI interference.
Hence the Power Factor should be checked while the power supply is working.
Hence, the PFC circuit is also the prime suspect. The PFC circuit could be failing, this assume all units carry the CE mark.

[shakalnokturn]

Those operate on the DC and filtered capacitors especially these are supposed to be blocked by the input lines chocks + X and Y caps using the Ground Cable which should be the prime suspect as faulty Ground can also act as "emitting Antenna" or emitting body.
This assume that the power supply is working properly otherwise it is also a sign of some failures example very excessive power supply ripples.
Talking about Antenna. Maybe he should describe his TV Antenna System and the proximity of the two.

At 125MHz I wouldn't count on the power cord earth wire to stop any EMI at all.
Broken or missing shielding on an Ethernet link could be something to consider. (Unshielded twisted pairs in apartment walls?)
You may consider experimenting with clip-on ferrites on the various cables to the desktop case.

[Armadillo]

At 125MHz I wouldn't count on the power cord earth wire to stop any EMI at all.

By what you say, then all desktops would fail but obviously that is not true according to the above discussion.
 

Edit: I think you are mixed up with the purpose of common mode noise on the twisted pair and the purpose of the shield.

A freshly bought unit from another brand gives me none of these issues.

[shakalnokturn]

Maybe... I'm only saying I would not consider the earth conductor having much efficiency reducing anything coming through the PSU socket at 125MHz.
I 'm certainly wrong, as I'm not going to take time to make measurements to defend my suppositions, besides that I'm still far from perfect, I continue to learn things every day.
Still, could you explain why the presence/lack of earth conductor would make some/all desktops fail EMI standards, I just can't catch the relation between my statement and yours...

I'm possibly confused on the Ethernet and shielding, would you take time to explain?

[Armadillo]

Firstly, the TV contains many stages of band pass filter circuits, hence withing standards, it should not be susceptible to the airwave interference like from mobile phones, WIFI, transmitters transmission etc.. all around you.
To interfere with the TV reception, the signal must be chiefly broad and spread spectrum of harmonics waves of sufficiently strength and as it relate to the desktop switching power supply in question, the prime suspect can be at the PFC/Rectification section, as the primary suspect.

Secondly you may not be aware that frequency like 125 MHz meant nothing without the signal strength / watts being associated with it. If you fear frequency, you should also begin to fear handphones.
The WIFI transmitter wattage is very much much larger than the 125MHz ethernet and if this 125 MHz can't even interfere with the computer internally, why is the undue concern?

I did not specifically state loss of earth that it will fail the standard but I said it act as an antenna for the spurious harmonics to be conducted and radiated and it could be the prime suspect.
Please take time to read and don't misquote others message, as correction involve unnecessary time.

If you are interested please read the input EMI filter. I am sorry I would not be able to dig in any further, I trust others could help you in the subject.

https://uk.tdk-lambda.com/faq/question.aspx?id=99

[Armadillo]

I'm possibly confused on the Ethernet and shielding, would you take time to explain?

The Ethernet cable is a weak "feeble" signal line. It is susceptible to signal interference hence the shielding protection and the twisted pairs. Not the other way around.