URGENT: Radiated emission for TOPHX series putting a only C snubber.
Dear Sir,
I have a big problem in my 36 W SMPS design. It was failing in radiated emission class B tests in 70 to 100Mhz..
In order to pass this test, First I put a RC snubber like DER187 and similar designs between Source and Drain, but it does not change anything, after that I put only 100pF 2kV high frequency Ceramic disc capacitor and I see good results.. The SMPS pass all the tests..
I have a question, putting only a capacitor between source and drain is a dangerous design? or it doesn't make any problem? Can anyone can help asap?
Best regards,
Comments
First of all like you I thought that, the reverse recovery of the output diode can cause this high frequency emission and I change the snubber and etc. but it doesn't change any thing.. Then I change the layout but it does not help anything.
I also add ferrite bead serial to auxiliary wire it helps alot but does not satisfy the class b limits.
Then by the help of spectrum analyzer and a air probe, I check all the possible points before adding capacitors.. I use magnetic shield to every point one by one in order to see which part causing this situation, the main problem was the turn off of the mosfet.. Secondly when I decrease the pirmary voltage the radiated emission is increases..
I change the snubber in the primary side but it doesn't help.. finaly I decide to add paralel RC snubber to Drain and Source.. Finally I remove R and only with C, it effect and suprress all the noise.. I tested in the laboratary and see that İt passes the emission class B.. Now below 120 V AC input full load it fails all the limits..
I can send you the results of the emission tests and schematic if you can give your personal mail..
Best regards,
I'm following up with a private message.
-The Traveler
kemred -
Our company was on break for most of the week of July 4th. I think we may have lost touch. Were you able to get your EMI issues resolved?
-The Traveler
As I said we connect 100pF parallel to DS of the TOP257YN. Then I measure the current passing through the drain of the mosfet by the help of shunt resistor, and observe that maximum current is nearly 3A peak at turn on off the mosfet.. Which seems acceptable, I did not make any improvements,
Best regards,
kemred -
Can you resend me whatever files and informaton I requested earlier? I can't seem to locate them in my inbox.
-The Traveler
As I said couldn't send them by private message since I could not find how ot attach files, there is not attachment part in this web site( http://www.powerint.com/en/user/64790/contact ), can you help how to send files by private message?
My email address should be shown on the private message. Just use your normal email software to send me the information.
-The Traveler
Before you start adding caps and snubbers to all the different components at random, you need to track down the component that is causing this high-frequency noise. This is a fairly easy task and can be accomplished with just your oscilloscope and a 100x scope probe (a high-voltage differential probe would be better if you have it).
Some of the more common sources of such high-frequency ringing are:
Output rectifers at turn-off
Primary clamp circuitry at turn-off
Improper PCB layout of clamp circuitry
Improper PCB layout of output rectifer traces and snubber components
Improper PCB layout of primary winding and drain trace current loops
Heastsinks that are attached to high-frequency switching nodes
Some of the more less common sources of high-frequency ringing are:
Improperly chosen output post filters (not very likely)
Ringing and/or harmonics in your EMI filter.
Again, these are pretty easy items to check. Simple hook up your scope across the different switching devices and looking for classic ringing waveforms. These ringing waveforms will often get hidden in what looks likes "noise" at the edges of turn-on and turn-off waveforms. Zoom into these areas and look for classic ringing waveforms. Measure the frequency of these waveforms...are they in the frequency range you're experiencing problems with?
Another good tool you can use that is simple and very easy to make is a sniffer probe. Take the tip off a 1x or 10x probe and use a probe tip socket on the end of your scope probe. Make a small bundle of coiled wire in a circular loop and attach one end of the coil to the probe tip and the other end to the probe ground connection. Being careful not to touch any electrically active components, move your probe around to different areas on your PCB and the PSU components. See if there is an area of the board that is radiating more noise than others. This type of measurement won't give you exact values of the radiated EMI but it will point you in the direction of the most likely culprit. My $20 would be on your output rectifier snubber circuitry and heatsink. My next best guess would be that your clamp circuitry isn't setup quite right.
If you can attach your schematic and PCB layout, I can give you some suggestions.
It would also be helpful if you could take some scope captures of your drain voltage waveform and zoom in on areas that have high-frequency ringing present. I would also do the same for your output rectifer as a good starting point.
-The Traveler