EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: madhu.wesly01 on January 10, 2019, 01:46:20 pm
-
Hi,
Good Day!
I have an application developed, for that I want to get certification on Conducted & Radiated Emission(FCC & CE certification). This application requires an external AC-DC adapter for getting powered-up. When they were tested individually, both got passed. But, when the application powered-up with the AC-DC adapter and test, then it is getting failed. It would be great if anyone give pointers in this regard.
How to come out of this issue with-out much effort??
Thanks in advance.
Best regards,
MadhuWesly.
-
This is a very common issue.
Power supply manufacturers test their products using simple resistive loads, which is perfectly valid. The product passes emissions - just! - and they place it on the market.
Unfortunately this leaves little or no margin for any noise generated by the product being powered, and the extra noise from the product pushes the total over the limit.
Also, some supplies are noisier at part load than they are at full load, which means a supply which is able to meet the peak power requirement of a product is actually producing too much noise on its own under normal conditions.
Power supply manufacturers are very reluctant to accept the fact that their products need to pass with margin under all conditions in order to be usable commercially.
As a product designer, you have a few options:
- improve the filtering and shielding of your own product, so it produces less noise of its own
- use a different power supply, which may be quieter when used with your product
- include EMC filters with your product as part of the package, and inform your customers that they must always be used
None of these options are quick or free, and that's just too bad. We've all been there.
Sometimes you can convince a power supply manufacturer to improve their product, but a better option (unless you're a big company with a lot of influence over your suppliers) is usually to buy a few different options from different manufacturers, and test them all to find which is the quietest when used together with your product.
-
How was your device tested without a required adapter..??
Tim
-
How was your device tested without a required adapter..??
Tim
Bench PSU?
Anyway, do you have full specifications of your PSU? Many PSU vendors offer PSUs with and without Y-cap, with Y-cap version obviously a bit more expensive. If the one you have tested is without, simply using the Y-cap version may be enough to pass...
-
My first suggestion would be to try and isolate where the source(s) of emissions are. Assuming you have a spectrum analyzer, you might be able to narrow down the source with a simple loop probe made from some coax. I would make the loop pretty small to help with localization... maybe 8mm-10mm diameter. If you find that the emissions are coming out the power connector, then you might try adding a series ferrite / pi filter before your internal regulator/supply. You might also consider a ferrite loop being added to the external power supply near the connector if it doesn't already have one. Kind of hard to suggest much more without seeing the circuit / layout / report, etc.
-
I've found that a 2-4 mm loop of copper wire on the end of a scope probe is an astonishingly effective H field sensor for locating noise sources. Just don't use the probe tip as a form. Measure it and use something slightly smaller. The larger the loop the more sensitive it is. So having large loops to get close and small ones to pinpoint the source has become SOP.
But just a bare probe without any ground lead so that only a few mm of the center conductor sticks out past the shield is often better. I made up a condom out of heatshrink to prevent shorts and have been quite amazed at how well it works with the DSO at 1 mV/div and the tip dragged across suspected sources.