For the price, buy the thing and install it inside the chassis of your radio and problems are solved.

My modern all valve hifi amp has one of those filters and it is totally noise free.

I built a couple of valve audio preamplifiers and fitted those filters to them as well.

Noise free operation and as far as I am concerned they are the best thing I have ever seen.

Highly recommended.

The caps as I suggested are easily mounted within most valve equipment. The valves themselves are pretty robust when it comes to spikes & surges. Albeit it is cheap & easy to put an MOV across the mains at the primary.

A few years back I had an Astor RK in that I reckon had been hit by lightning. The valves & transformers were OK, only one resistor survived & even the mica caps were stuffed.

It may be what is being used to produce the DC (similar to CFL,s) not the LED's that makes the noise.

I recently had a FAX generating signal around 175kHz getting into an IF train.

Marc

In the shop I was contracting at the local electric utility offered to install the "New and improved" energy saving flourescent lighting.

The building owner did that and it totally killed all AM and shortwave reception in the entire shop as well as taking out all the TV channels Channel 6 and below. We were on cable too!
The electronic ballasts generated so much RF noise it swamped out the cable system in the building. It was in excellent condition too, as the provider had just put in new service drops and we had just finished an entire rewiring of the shop to bring up the signal levels.

They refused to believe it was causing a problem, until the head electrician came in & I shut off the entire flourescent lighting circuit breaker, and all the signals came back to normal.

They tried RFC's, multiple RFC's on the electronic ballasts, etc,over a period of 3 weeks with no success. They finally had to come back and replace the "New and improved" energy saving electronic ballasts with the old style transformer ones.

It was so nice to be able to receive signals on what I was working on.

My new workshop space has the very thin (4ft?) type tubes installed. Not sure of their type number as I haven't handled one yet. I hope they are not in ballast-free fittings. I guess I should take a valve radio there sooner than later to see whether or not I have an RFI-infested cave.

I have a new house & all 4ft tube fittings are electronic. The worst RFI comes from the new Fridge, merely adding fuel to the regulators being asleep at the wheel, as there is a lot of crap stuff getting in.

Telstra actually had a fridge light wiping out the internet here.

The workshop is fortunately a metal shed. It is 1980's and all of the fittings are old type florescent. The one above the bench is a twin 85 watt with 8ft tubes.

I have managed to keep the interference generators out. But the power supply still has MOV's & caps to get rid of lightning spikes & RF from the miles of 24KV aerials.

Marc

My experience with electronic ballasts goes back a few years and whilst I am not up to speed with RFI output I know there is vast differences in build quality between brands.

I've found Tridonic Atco and Osram ballasts to be the best quality. Cheap Chinese brands have a high failure rate and probably spew out RFI like a volcano.

Then again, Australian laws relating to interference are also very stiff. It is against the law to use any item that causes interference to anyone else's reception of an authorised radio signal. This means that if you install a light fitting and this causes the bloke next door a disruption to radio or TV programming then you have to turn the light out until it can be modified to not cause interference or simply remove it altogether.

Though in saying that, who is policing communications laws in Australia these days? The police aren't because it falls outside their jurisdiction. Are there any DOC inspectors like there were 20 years ago? Not that I know of. It is therefore reasonable to suggest that interference will become a bigger problem down the track because it is all too easy for a manufacturer to stamp "This product complies with AS/NZSxxxx" on the things they make without performing adequate tests and producing results required by the standards.

Then again, with all the high frequency electronics and all those triacs and SCRs operating at resonance, is it even possible to produce things like an electronic ballast without them emitting interference?

I note from the American Radio forum that they are having similar problems to us from CFL's and what they like to call "wall warts" ( cheap & nasty emitting switch modes often in plug packs).

I have already reported on a FAX putting out RFI into the Phone line and getting into a 175kHz IF.

Going back to a previous comment on CFL's it is the DC to AC inverter in these that is the root cause. The oscillator seems to vary the "load pulse" at a rate that reaches RF, with such a small load that presents as a short discharge is rapid, where a ballast is more like a resistor albeit that it is a dirty load.

I have one Heat generating Halogen bulb, it is on a transformer so has no RFI generating capacity.
I wonder about LEDS.

Essentially a DC device which should not generate RFI. But if you have a switch mode PSU to avoid heat generation, therein lies an issue.

It seems apathy is becoming a major issue: but they don't seem to care.

Marc

My new workshop space has the very thin (4ft?) type tubes installed. Not sure of their type number as I haven't handled one yet.

It sounds like the 28 watt T5 lamps. Lighting companies like to brag that they put out more light than the regular T8 lamps but it's bull. T5s also do not last anywhere near as long and the ends go black much quicker too.

On an unrelated note, you can buy a T5 converter module that will fit into most T8 fittings however these are very poor quality and you have to remember to remove the starter or you'll blow the bejesus out of it. They really aren't worth the bother.

Warm white T8 LED tubes are the way to go if one is a purist and wants to save the planet and a few bob. The starter still has to be removed though I would be playing it safe and get an electrician to remove the iron core ballasts and starter holders as they are not needed with LED lamps. Not that replacement of these lamps should be on the cards but if that does have to happen at some stage turn the lights out first as, unlike a fluoro lamp, the pins on a LED lamp are live if one end is still in the tombstone connected to the active wire.

It sure sounds like I have the T5 tubes. An interesting comparison here:

http://www.lightingsolutions.ca/news/t8-vs-t5-fluorescent.html.

There is something on that website that my browser doesn't like - probably a badly-written JavaScript.

Neither Firefox nor Chrome has a problem, so I'm guessing you're using IE.

I have all five major browsers here (mostly for testing) though only use IE for browsing. I'll try on one of the others later tonight.

I'm installing Jaycar EMI filter as recommended by Simplex further up this thread, but I have a question about earthing.

The filter module only has an earth terminal on one side - I'm assuming the mains connection side.

Should I bridge the chassis earth to that terminal, or leave it floating? I patched in the module with only the active and neutral connected on the radio side, earthed on the mains side. Some noise is suppressed but some 50Hz hum remains.

Maven

I tried one out as welI and bridged it to the chassis. It didn't make any difference to my noise. I'm been driven mad trying to track it down. After reading Marcc's post again I'm going to try turning the fridges off (and hopefully remember to turn them on!)

With a valve radio you have to establish how the crap is getting in, or where it is coming from.

Source 1; Is the antenna picking it up. that can be proven by shorting out the antenna. Where that involves an antenna in a power circuit, best to use a 0.01 HV cap.

If the crap is getting in with the antenna shorted, then, unless the set has some way of generating crap (as opposed to receiving it from a large building on a hill in...) then it is likely from the mains.

This is more apt to get through if the set has no decoupling cap on B+ ( 0.22mfd - 0.047 mfd) parallel with the second filter electrolytic: Electrolytics do not filter out RF well.

As I have written frequently my normal attack to reduce it, is to have a 0.1 -0.22mfd (mains type 275VAC) Active to Neutral then a cap 0.0047mfd active to earth and same neutral to earth.

To kill the common lightning episodes here, most filters have an MOV A to E and another known A to earth (all of my dozens of plugs scattered over 5 sub boards & one of the meter boxes are right) Some filters I have seen are a bit like the old car radios, with a winding around a ferrite core, on both N & A to get RF.

Whist the A to E may sound odd it is deliberately aimed at tripping the RCD, or CB to help take the surge off devices.

Marc