Model of EMC Conducted Emission Test Reciever - Quasi peak measurements

[Klaus Vestergaard Kragelund]

Hi

I'm playing with using pseudo random PWM on a motor drive to minimize
the conducted emission level of the inverter

For that purpose I'm testing a system using a standard Rhode&Swartz test
reciever with the following frequency band 150kHz to 30MHz (9kHz
bandwidth, 5kHz step and both average and quasi peak measurements)

The PWM converter clock is running at 18kHz and modulated +/-20% at
100Hz rate (in the beginning triagular shaped mdulation)

But the measurements take a long time - do anyone know how to model the
input of the quasi peak detector to mimic the result seen on the test
reciever? If I get a model I can run some simulations instead.

As far as I know the quasi-peak detector consists of a rectifying diode
in series with a RC filter and a load resistor on the output of that
filer (shorter rise time than fall time to transients)

Thanks

Klaus

 

[Mark]

Hi

I'm playing with using pseudo random PWM on a motor drive to minimize
the conducted emission level of the inverter

For that purpose I'm testing a system using a standard Rhode&Swartz test
reciever with the following frequency band 150kHz to 30MHz (9kHz
bandwidth, 5kHz step and both average and quasi peak measurements)

The PWM converter clock is running at 18kHz and modulated +/-20% at
100Hz rate (in the beginning triagular shaped mdulation)

But the measurements take a long time - do anyone know how to model the
input of the quasi peak detector to mimic the result seen on the test
reciever? If I get a model I can run some simulations instead.

As far as I know the quasi-peak detector consists of a rectifying diode
in series with a RC filter and a load resistor on the output of that
filer (shorter rise time than fall time to transients)

Thanks

Klaus

Randomizing the frequency does not reduce the emissions, it just
spreads them out so that you can pass the test. But it still creates
nearly the same level of interference and to more frequencies. I know
this has become a popular technique to pass EMI tests but I think it is
a bad idea and just HIDES the problem.

Mark

 

[Paul Burke]

Randomizing the frequency does not reduce the emissions, it just
spreads them out so that you can pass the test. But it still creates
nearly the same level of interference and to more frequencies. I know
this has become a popular technique to pass EMI tests but I think it is
a bad idea and just HIDES the problem.

The EMC standards are a hoop to jump through, not an objective quality
standard. They say meet this curve: you meet this curve, you've passed.
Don't start inventing extra obstacles, there are plenty of stupid,
unnecessary ones already. RoHS/WEE anyone?

Paul Burke

 

[Terry Given]

Randomizing the frequency does not reduce the emissions, it just
spreads them out so that you can pass the test. But it still creates
nearly the same level of interference and to more frequencies. I know
this has become a popular technique to pass EMI tests but I think it is
a bad idea and just HIDES the problem.

Mark

This is true, *but* passing the test is a necessary requirement. That
the test is less than useful is a whole 'nother can of worms.

Cheers
Terry