Noise suppression capacitor

Check out www.electrocube.com

First you need to know what frequencies are going to cause problems, so get hold of a spectrum analyser (local universities are often very helpful if you need to borrow high tech equipment). The best solution will depend on the frequencies of the noise and how it's getting to the outside. You may need to redesign the electronics to have a lower clock freq, have on board filters to limit rise times of critical signals, change components, add screening or just have a simple low pass filter on the power leads. It depends greatly on where the noise is coming from and what standards you must comply with. There's a lot of info on designing filters available on the internet, but it's not so simple in the real world since the components (inductors or capacitors) are not ideal (their properties change with frequency, temperature, age, production run) and there are price /performance tradeoffs. You'll need to design and test a few alternatives. Jeff

Hello,

9 kW is a lot of wood - I would think its no such good idea to suppress the noise only by capacitors because they will be very large and expensive when you look at the currents they must pass. Additionally you would need large inductors.
Any kind of noise in that application will be generated by harmonics of your 50 or 60Hz system. The more harmonics you can avoid the cleaner your EM-radiation will be, as Jeff posted in (2).
The way is to make a PFC, a Power Factor Correction, which is mostly done by an electronic switching device inserted in the line before your load and which makes the load current consumption follow the line voltage with the same shape. If, per example, your Voltage has a sinus shape and your Current has not then harmonics will be generated. The internal frequencies of PFC are very high relative to 50 Hz so the required capacitors can be very small compared with those you need without PFC.

Regards Uwe

It seems to me that 9 KW is fairly significant. You want to ananlyse your output before you disign your filters. Its been a long time since I did any electronic design work but as memory serves me, there are a couple of fairly basic answers. Perhaps i am being too presumptious, but if your device does not generate a lot of harmonics, a tuned circuit, where inductive reactance and capacitive reactance are equal at the resonant frequency might just be the ticket, or perhaps a pi filter. On the other hand if harmonics are a problem, a more elaborate system might be rquired. The arrangement of inductors and capacitors will have a significant impact on your device.

Inductive Reactance (Xl) = 2pi*fL

Capacitive Reactance (Xc) = 1/(2pi*fC)

Resonant frequency (Fr)=1/{2pi*(sqrt LC)}

where

X = Reactance

F = Frequency

L = Inductance (henrys)

C = Capacitance (farads)

I haven't messed with electronics in 10 or 15 years but if memory serves me, these formulas are correct. If I am misleading this person, would someone please correct my mistakes for him.

1, It needs a large Filter-Network not a simple capacitor

2. Is the appliance a linear-device, ie drawing sinusoidal current or else making lot of noise ?

I agree with at least a couple of the originators responses: 9KW is not a real biggie, and likely with an off the shelf L/C filter and/or PFC device, you may be compliant with the requirement specification itself. Good EMC and I/O line testing (by a certified lab) would be the proof of the pudding.