help about smps emi problem

[shanren]

howto reduce noise due to smps

hi,can anyone can tell me about how to reduce the rf noise of SMPS?
and a strange problem,a build a smps of about 50w ,24v2a.when it works,sometimes is good,but sometimes it will sent out 100Hz sound noise,i find when this hanpens,the voltage of mosfet drain is not the same as it works well.but sometimes it will be ok after a few seconds,and maybe so after a few minutes.this can cause a lot of rf noise when i measure the noise.
any information is wellcome!
thanks!

 

[mtawab74]

smps emi

Hi

I think your loop is unstable
check the frequency response of your feedback loop
the 0dB should be at frequency less than 1/5 switching frequency

also check your phase and gain margine

salam

 

[shanren]

about smps

no,i think the loop is ok.i use a oscillograph,find the pluse is stable.the problem is the noise is 100hz(i use 50Hz AC input ).
but i do not know how to measure phase and margine,i only have a oscillograph.can i measure phase and margine with it?and how?
thanks,mtawab74.
best regards!

 

[MADURAN]

switching frequency emi problem

Try to connect a capacitor (4.7nF - 10nF at min. 1000V; it depends of your isolation voltage) between power ground and output ground.

 

[mtawab74]

y2 capacitor in smps

Hi

I may missunderstand your problem

if you find 100 hz on your o/p

i think this is due to high coupling capacitance between primary and secondary winding.

the solution is

1- connect 1nf 250v/Y1 between primary and secondary grounds ( close to transformer pins)

2- connect 1nf 250v/Y2 between secondary ground and earthed case
3- connect 1nf 250v/Y2 between secondary output and earthed case

all connection should be so close to transformer pins.


Y1 = a capacitor that fails open ( not short) at voltage in the range of 3kv
Y2 = a capacitor that fails open ( not short) at voltage in the range of 1.6kv

salam