Mains filter for small linear power supply

EMI control is seldom as simple as retro-fitting a filter, but I don't suppose it will hurt.

It's often something other than the power supply interacting with the grounding...

It would make it better, the B type is better on the data sheet.

Best is to just try it, with and without, see if there is interference.

( I left a lot of stuff out here, but you may not need any of it)

It wouldn't make it worse, but it may not protect against RF. Its not always safe to say that the filter will work well in either direction. Filtering the low voltage side from RF may be easier than the mains.

As for transient, if you need protection from lightning then GDT's should be used. But for most 3rd branch circuits MOV 's could be used if a lightning strike taking down the equipment is acceptable.

And a fuse between the filter and mains is a good thing.

Cheers

These filters are really common surplus items. I wouldn't expect miracles. I think the reply that suggested filtering the DC was spot on.

For yucks, I took one of these off the shelf filters and put it in series with the DC/DC for my notebook computer. I was trying to reduce the hash radiated by the car wiring due to the DC/DC pulses. It made the noise much worse. I guess it was resonant and the pings from the DC/Dc excited the resonance.

"Alan Slough"

** So there is an iron transformer, rectifiers and filter caps - right ?

How do you imagine RF signals and voltage spikes will get through them and your regulator ??

** Such filters are intended to reduce EMI going in the opposite direction.

.... Phil

Phil is right on. If there really is a problem with mains interference you usually take a transformer with an additional shield between primary and secondary. The filter probably won't make much difference tho. ciao Ban

? d

n.

Thanks to everyone for your replies, which are very much appreciated.

You are correct with regards to the components. On the primary side there is a fuse followed by a MOV which is in parallel with the primary winding. The secondary output is bridge rectified and I am using 1uF tants and 1000uF electrolytics to smooth the input to the voltage regulator. There is another tant across the regulator output.

Having looked at EN 61000, I'm not so concerned about electrostatic immunity and radiated EM energy immunity. It is the fast transient burst testing that I'm worried about and what steps I can take to mitigate this.

Presumably, all linear power supply manufacturers have a standard solution to this. Love to know what it is!

"Alan Slough" "Phil Allison"

You are correct with regards to the components. On the primary side there is a fuse followed by a MOV which is in parallel with the primary winding. The secondary output is bridge rectified and I am using 1uF tants and 1000uF electrolytics to smooth the input to the voltage regulator. There is another tant across the regulator output.

Having looked at EN 61000, I'm not so concerned about electrostatic immunity and radiated EM energy immunity. It is the fast transient burst testing that I'm worried about and what steps I can take to mitigate this.

** So your MOV, transformer and filter caps will have no effect ??

What universe do you inhabit ??

.... Phil

Those Tantals are calling for trouble.

What exactly do you want to protect? The transformer is not so vulnerable to fast transients, it can momentarily digest quite an overload. A continuous input voltage on the upper limit OTOH can be bad, if there isn't enough margin. It will heat up and eventually self destruct. A PTC in series with the primary might be a good idea. ciao Ban

to

s h

Isn't the transformer subject to insulation breakdown if the transient is high enough voltage? It's been my experience that when something breaks down, it is instantaneous. It's not like getting some device overheat a bit, i.e. if the event doesn't last long, the part can take the stress. With a break down, it's like flipping a switch.

But he has put a MOV across it. Wouldn't that protect from this fault mechanism.?

Please may I ask why? Is it because they are not so good at high frequency? I have decoupling ceramic disc capacitors between the supply rails local to the ICs.

I guess the AC mains immunity listed in section 4 of this webpage:

Thanks for your suggestions.

If the big electrolytic is not far away from the regulator you don't need anything there. Tantals can't stand the inrush current very well, when you switch on and sometimes explode, Jorg knows how it sounds. The one behind the regulator is needed for stability. Here it is best to take what recommends the data sheet, usually some ceramics, sometimes in series with a small resistor even (for LDOs)

With a transformer it shouldn't be too difficult to meet the specs.

ciao Ban