Filters above 100MHz need shielding?

treez said:

A…..Adding an RC snubber from switching node to ground at the Buck SMPS may well have a good effect. However, the sudden discharge of the snubber capacitor simply means there will be another radiated emissions problem, but above 1GHz, where formal testing does not take place?

Click to expand...

I don't see why this would make emissions worse, not better. It really depends on layout. If a snubber is properly designed, then it provides a low impedance path for high frequency currents, but in doing so it should divert those high frequency currents from other components which would radiate more.

B…….If we use common mode chokes with Y capacitors to the chassis, then any failure peaks that we may get are just as likely to be coming from resonance in the common mode LC filter, as from genuine noise emissions? As such, tweaking the Y capacitor value to try and bring peaks down is likely to be futile, as it will just mean getting a resonant peak at a different frequency?

Click to expand...

Then damp your filter?

G.......Why on the entire web are their No forums, or sub-forums, for people to ask questions about EMC testing of products?....is it because the actions taken to get kit through EMC testing is the deepest secret of any company?

Click to expand...

Because EMC stuff is highly anecdotal, and solutions can end up being very specific to a product/system. You can't just copy paste someone else's input filter design and expect it to work in another application. In the end the empirical results are what matter.

C……

D…….

E……

F….

Click to expand...

Again, if you have a specific problem you're trying to solve, then you're better off just trying all these options in the lab rather than asking people on the internet.

Again, if you have a specific problem you're trying to solve, then you're better off just trying all these options in the lab rather than asking people on the internet.

Click to expand...

Thanks, thats hit the nail on the head.. as you know, these radiated emc test labs are phenominally expensive places to book.
Also, as you know, the radiated EMC equipment and lab is utterly unaffordable.

Then damp your filter?

Click to expand...

Yes but how do you know the failure peak is from common mode filter resonance, or from genuine noise? The common mode filter concerns the stray caps to earth, so its impossible to calculate where the resonances will be.

The blue cable in the attached picture comes out of enclosure A. The blue cable is the output of a common mode choke PCB, which filters the output of a flyback SMPS, which is also in A.

..For the purposes of reducing radiated emissions from 30MHz to 1GHz…..

1…Do you think its worth using screened cable for the wired interconnection between the Flyback SMPS and the filter PCB in A? (instead of just twisted pair).

2….Supposing the assembly person strips back the screened cable too far, and before the blue cable goes into A, it is stripped into its separate go, return, and screen wires, which are then out in the open…..would it then be necessary to cover these in copper tape, so as to reduce emissions? (the copper tape being connected to the cable screen).

How far into enclosure A should the blue cable go before it is allowed to be stripped into its three internal wires?
***********************************************************************************************************************

Again concerning radiated EMC testing in a chamber (30MHz to 1GHz)….

The attached system is usually mains connected, and the metalwork of enclosures A and B are usually earthed via the mains lead earth wire. Now supposing I disconnect the mains plug, so that the enclosures are no longer earthed. I then put a battery in enclosure A and use it to supply enclosure B via the blue cable. Will the floating metal enclosures now act as antennas and cause lots of emissions problems?.....would it now actually be better if the metal enclosures were actually removed from the EMC chamber? Do you agree that the now floating metal enclosures of A and B are now causing more radiated EMC problems than they are solving?

****************************************************************************************************
Hi,

The following is an 0603 size ferrite bead with 2500 Ohms of resistance at 100MHz…..

https://4donline.ihs.com/images/Vip...1-1.pdf?hkey=6D3A4C79FDBF58556ACFDE234799DDF0

.......If such a small, 1.6mm long component can offer 2500 Ohms of resistance to 100MHz, then this surely implies that frequencies like 100MHz do not generally “propagate around” components after all? Since obviously, the 100MHz is not able to propagate around this tiny ferrite bead?...rather it suffers a path resistance of 2500 Ohms. So all this talk of high frequency noise jumping around, and bypassing filters, has to be taken with a pinch of salt?

Below is a 0201 size ferrite bead which offers 1800 Ohms of resistance to 100MHz.....its just 0.6mm long!!!...this surely puts to rest the notion that 100MHz frequencys can "propagate around" components and bypass filters etc?

0201 size ferrite bead (just 0.6mm long!!!!)

Bottom of page 19 of the following….

https://interferencetechnology.com/wp-content/uploads/2016/10/2016_IT_EMC_Filters_Guide.pdf

…..states that input and output traces to ferrite beads should be routed away from each other to stop the stray capacitive coupling bypassing the bead to the 100MHz….are they serious?...a 0201 ferrite bead has pads which are less than 0.2mm apart…..how far are they going to get the traces?

treez said:

Thanks, thats hit the nail on the head.. as you know, these radiated emc test labs are phenominally expensive places to book.
Also, as you know, the radiated EMC equipment and lab is utterly unaffordable.

Click to expand...

Then develop your own precompliance testing methods, like everybody else. You've already been recommended some approaches, but of course since EMC is highly anecdotal you'll have to tailor them to your needs.

Yes but how do you know the failure peak is from common mode filter resonance, or from genuine noise?

Click to expand...

Not sure what you mean by "genuine" noise. The antenna used for the measurement doesn't care.

The common mode filter concerns the stray caps to earth, so its impossible to calculate where the resonances will be.

Click to expand...

If your device has a wired connection to earth (which I believe it does, in the AC mains cable) then stray capacitance should not affect the CM filter's behavior significantly, because both ends of the CM choke should see fairly low impedance at high frequencies. On the source side it will be looking towards a low impedance earth connection (or a LISN), and on the load side it should have low impedance X/Y caps which will be far larger than a few pF of stray capacitance.

Not sure what you mean by "genuine" noise. The antenna used for the measurement doesn't care.

Click to expand...

By "genuine noise", i mean a peak on the radiated scan that doesnt come from a resonance in the common mode filter.

If your device has a wired connection to earth (which I believe it does, in the AC mains cable) then stray capacitance should not affect the CM filter's behavior significantly, because both ends of the CM choke should see fairly low impedance at high frequencies. On the source side it will be looking towards a low impedance earth connection (or a LISN), and on the load side it should have low impedance X/Y caps which will be far larger than a few pF of stray capacitance.

Click to expand...

Thanks, and you are speaking of the mains AC common mode choke.....my apologies, as i am only speaking here of the common mode chokes used in A or B but that come downstream of the flyback SMPS output.

Also, what if only a battery in A supplies power....?....in this case, should A's enclosure be earthed or not? (for best radiated EMC performance)

Also, in a radiated emissions lab, we need to use screwdrivers and other tools to disassemble the unit under test in order to tweak its component for the purpose of radiated emissions, we do this between every scan.
What happens if we just leave the tools in the EMC chamber whilst radiated emissions scans are being done?

Hi,

treez said:

Thanks, thats hit the nail on the head.. as you know, these radiated emc test labs are phenominally expensive places to book.
Also, as you know, the radiated EMC equipment and lab is utterly unaffordable.

Click to expand...

how do you perform your tests in the lab? There are several methods to investigate EMC issues with DIY equipment. E.g. using a foldable ferrite which can be placed arround your cable, make one ore more turns with a wire, and connect both ends of this wire to a connector e.g. BNC. This allows you to see what is going on at your suspicious cable. And now starts the clue, save the result on your spectum analyzer or scope (FFT) as reference and start to tweak your system. Now you can compare the results and its obvious which changes are beneficial. If you know in advance you have problems at certain frequencies this approach will help you to minimize disturbances at this frequencies with respect to your reference spectrum.

BR