Electric Automation Forum
Forum » General Discussion » Three-level boost converter in PFC vs. conducted EMI
Topics: Three-level boost converter in PFC vs. conducted EMI on General Discussion
Start by
09-07-2013 10:15 PM

Three-level boost converter in PFC vs. conducted EMI

What do you folks think about application of three-level boost converter in PFC?
It seems very compelling for me, rather for multivoltage input then for real PFC in three-phase power supply.

But ...
Both (+) and (-) of the DC bus are HF switching potentials in relation to input lines.
That is because there are diodes on both (+) and (-) lines.
It means that the number of elements and conductors with fast voltage transitions is increased.
Obviously there will be isolated DC/DC converter/converters after this non-insulated boost.
I predict increased problems with CM noise. Another problem is that I want to have entire control of DC-DC converter on protective earth potential. Of course drivers of switches in the converter are insulated, but then they're exposed to additional voltage transitions from PFC. On the other hand, since there are only n-channel devices, in DC-DC converters only those switches with emitter/source connected to (-) are free of this disadvantage. I have never noticed any problems with any false operation of isolated floating switch driver.

Am I exaggerating with my afraids?
09-07-2013 10:16 PM
Top #2
09-07-2013 10:16 PM
I don't think it is Three phase PFC circuit, i,e, The three phase input current waveform is not same as input voltage. Definitely it is some kind of boost topology
09-07-2013 10:17 PM
Top #3
09-07-2013 10:17 PM
You aren't exaggerating your concerns.
This circuit is occasionally proposed for high efficiency "universal input" converters.
At Low line, the neutral is connected to the centre tap of the bulk HT capacitors.
Apart from considerations of keeping the even harmonics down so that the centre tap doesn't wander about in voltage and the problem of accurately sensing the line current, the couple of attempts at this that I've witnessed failed due to common-mode EMI problems; so bad that control circuits would not work properly on one design I saw.
09-07-2013 10:17 PM
Top #4
09-07-2013 10:17 PM
It is very essential that any output Bus terminal (+ or -) of the Boost converter must be connected at any moment to any of the AC line to avoid the switching frequency common mode noise.
When you use bridge rectifier, that connection is rotating between AC lines to - bus terminal, which is OK.
This is the big challenge for Bridge-less PFC, because the -Bus of the PFC boost is bouncing w.r.t. to AC line with high voltage at switching frequency.
09-07-2013 10:18 PM
Top #5
09-07-2013 10:18 PM
For low common mode noise it helps if the centre of the capacitors is not switching. adding an inductor is added into the negative leg, to make the circuit symmetrical, can help prevent this. Otherwise you have the major problems with the "bouncing" output creating common mode noise problems. With the inductor these can be controlled with suitable y caps creating a quiet virtual ground
Reply to Thread