FWIW: I see the following DC filters being used in preamp circuits of a local manufacturer, not exactly cheap (though I got BNX002 as surplus) but attenuation is said to be 40dB from 1MHz to 1 GHz. It is rated at 10A so I am not sure if it would show the same attenuation in lower current/higher impedance circuits: FILTR EMI .1GOHM 10A 1M-1GHZ - BNX002-01
As far as the Blowtorch is concerned, this is the wrong direction to go, in my opinion. We experimented with these sort of input filters, for decades. I put a fairly sophisticated one in the AC input of a master tape recorder that I made for Dave Wilson in 1983. He removed it, after comparing the sound quality with and without the filter. This got me to be very careful about these line filters. Then, a couple of employees at Corcom, made a transformer based line pre-conditioner, that I still have one of. It used a low leakage transformer in the signal path, deliberately frequency limited to 20KHz, in order to filter out RFI. This unit was very popular about 20 years ago.
In any case, the sort of commercial line filter similar to Corcom, has been rejected in many cases by hi end listeners. It is most likely the 'quality' and the secondary characteristics of the parts that make them, rather than the schematic, that is the problem.
In order to reduce some potential confusion, I do use a lot of 'regulators' in my Blowtorch preamp, but they may not make any sense, taken in isolation from their use in the Blowtorch, itself.
In any case, the sort of commercial line filter similar to Corcom, has been rejected in many cases by hi end listeners. It is most likely the 'quality' and the secondary characteristics of the parts that make them, rather than the schematic, that is the problem.
In order to reduce some potential confusion, I do use a lot of 'regulators' in my Blowtorch preamp, but they may not make any sense, taken in isolation from their use in the Blowtorch, itself.
Apparently, i tried one by RP with 100dB insertion loss but rejected it.
www.radiuspower.com/pdfs/RPT5500.pdf
(if 10 bucks each for the BNX002 is not exactly cheap, €300 for a 20A Schaffner is what ?)
Last edited:
Line filters vs DC filters
Apples and pears IMHO: The links you show are AC line filters, what I am referencing are DC filters to be used in the path of the DC supplies, behind rectification and initial passive filtering. Positions may vary but I see them in some high end gear from a manufacturer from Berlin (selling very expensive shiny boxes). They seem to be using them at the entry point of the individual preamp pcbs (on board), after all previous DC filtering. As 1audio mentioned regulaters do not seem to be happy with higher frequencies and stop regulating at some point, so IF there is still anything coupling through the filtered DC supplies this could stop it. The filters are actually made up of ferrite beads and caps.
I am fully aware that there are mixed opinions regarding the use of AC line filters as there are claims that they introduce additional resonances - and the DC filters should not.
Apples and pears IMHO: The links you show are AC line filters, what I am referencing are DC filters to be used in the path of the DC supplies, behind rectification and initial passive filtering. Positions may vary but I see them in some high end gear from a manufacturer from Berlin (selling very expensive shiny boxes). They seem to be using them at the entry point of the individual preamp pcbs (on board), after all previous DC filtering. As 1audio mentioned regulaters do not seem to be happy with higher frequencies and stop regulating at some point, so IF there is still anything coupling through the filtered DC supplies this could stop it. The filters are actually made up of ferrite beads and caps.
I am fully aware that there are mixed opinions regarding the use of AC line filters as there are claims that they introduce additional resonances - and the DC filters should not.
Last edited:
True, but you can do both AC and DC filtering not extremely cheap.
http://www.radiuspower.com/pdfs/RPT5000.pdf
(088, 4 total, 2 for line rails, 2 for phono rails, plus shiny Swiss for AC )
http://www.radiuspower.com/pdfs/RPT5000.pdf
(088, 4 total, 2 for line rails, 2 for phono rails, plus shiny Swiss for AC )
The Schaffner line filters don't have any filtering to the ground wire. Filtering the ground wire may be well be a benefit, as there is RFI on it also.
My simple logic says that the best solution would be preventing any RFI entering any audio gear. That is, a CLC, or CLCLC filter, encased in a shielded metal case.
Possibly, the difficulty is to design and build such a filter that will not degrade the mains supply in any way, other than the RFI filtering. Possibly, many such existing line filters have too high impedance at 50/60 Hz, thus causing fluctuations of the mains voltage, with the changing power consumption of the audio gear.
I cannot see how a well designed line RFI filter that will not cause fluctuations to the mains voltage can possibly degrade the SQ of audio gear.
Possibly, when such an RFI line filter isn't shielded, right at the AC entrance, some RFI is simply transmitted into the audio gear, thus, reducing its' effectiveness.
My simple logic says that the best solution would be preventing any RFI entering any audio gear. That is, a CLC, or CLCLC filter, encased in a shielded metal case.
Possibly, the difficulty is to design and build such a filter that will not degrade the mains supply in any way, other than the RFI filtering. Possibly, many such existing line filters have too high impedance at 50/60 Hz, thus causing fluctuations of the mains voltage, with the changing power consumption of the audio gear.
I cannot see how a well designed line RFI filter that will not cause fluctuations to the mains voltage can possibly degrade the SQ of audio gear.
Possibly, when such an RFI line filter isn't shielded, right at the AC entrance, some RFI is simply transmitted into the audio gear, thus, reducing its' effectiveness.
"Tempest" line filters are for military applications and are specifically spec'ed for reverse conduction. In other words, they are supposed to block signals going from the equipment to the outside world where they could be picked up by sensitive receivers. They may or may not be good at blocking RFI coming into the equipment, I don't know.
/OT: I have done testing on data terminals that were not tempest certified, with receiving equipment and terminals outside the fence of a facility and we could easily get a duplicates of the terminal screens at 200 feet away. After that, we got lots of money to buy filters
jan didden
I saw between 6 and 10 of them in the aforementioned gear, 2 for each bipolar supply, so in case of separate PCBs per channel and for phono and line each.
Those tempest filters look indeed interesting. I have heard comments that line filters kill the the dynamics, especially of power amps, but whether that is because they interfer with the amp (not convincing, at least if properly dimensioned) or whether a quieter background "sounds" less dynamic is not entirely clear to me.
BTW, Ayre Acoustics have their own "non-magnetic" conditioners. If I understand them correctly the claim is that they are more long-term stable than ferrite based filters (ferrite supposedly degrade as they become magnetised).
Last edited:
"Tempest" line filters are for military applications and are specifically spec'ed for reverse conduction. In other words, they are supposed to block signals going from the equipment to the outside world where they could be picked up by sensitive receivers. They may or may not be good at blocking RFI coming into the equipment, I don't know.
/OT: I have done testing on data terminals that were not tempest certified, with receiving equipment and terminals outside the fence of a facility and we could easily get a duplicates of the terminal screens at 200 feet away. After that, we got lots of money to buy filters
jan didden
I have actually used the BNX002 for the TEMPEST applications in Handheld Computer design work, and can vouch that they are WONDERFULLY effective.
This has to do with carefully chosen elements inside:
- Ferritebead/ ceramic cap combination [effective into tens of MHz]
- Feedthrough capacitor [tens to hundreds MHz]
You might find them useful in audio, for snuffing out rectifier noise - but only if you wire them very close to the rectifier.
VHF and higher, entering through the power line will not be suppressed using a filter like this though - unless you shield your mains supply all the way back to the street. If you don't totally shield, any mains-borne VHF/UHF will electromagnetically couple into your circuit again, and bypass your filter. Remember, a quarter wavelength antenna for 250MHz is less than 300mm [12"] and 75mm [3"] is good at around 900MHz - very effective coupling mechanisms!
RFI is present everywhere.
The equipment case should block air born RF from entering the equipment.
Should there be RFI filter in a shielded metal (iron) case, right at the AC cord entrance, as far as I can see, it should block all RFI residing on the mains supply.
Joshua, You can try that, but remember that your iron case must have no gaps or slots for that to be effective, and your audio input cables, and your loudspeakers cables must have the same filtering as the power line, or the suppression effect at high frequencies will be poor.
Iron cases bring many problems of their own.
Iron cases bring many problems of their own.
Thanks,
No gaps or slots for such an application is obvious, it is what I had in mind.
Of course, tough where I live now, in the country side, RFI carried by mains supply may be much worse than RF radiation inside my house.
Is there any way around it? Is there any better solution?
What's stopping you from adding a simple blocker between the ground return and your floating isolation transformer.
(Martin, no criticism implied with the links, merely illustrating that the Murata filters are actually quite affordable. Afair, Dietie Baby has been using the BNX for quite some time, in previous models too.)
The Schaffner line filters are encased in a metal case, with feed-through capacitors. Any external choke will serve as an antenna for RF reception – not good. The ground line filtering must be inside the case of the entire filter.
There are 2 different RFI filtering solutions:
One is outside the equipment, the other is right at the mains entrance of the equipment.
In both those cases, possible ground wire filtering should be encased with the rest of the filter. Either one builds oneself such a filter, than that is taken care of. In case of purchasing a commercial filter, ground wire filtering should be included in the filter, not added externally to the filter.
I think this is "much ado about nothing". Of course, RFI is bad. RFI can come into the power supply through the line cord and the power line. Power supplies are often NOT optimized for RF, especially with toroid power transformers, etc.
The VERY FIRST idea to fix things is to put a commercial, usually reasonably priced, RFI filter right at the input. They usually give specs, and show the circuit even. Who could ask for anything more? Well, the first problem is WHAT they use for component parts. An RF ceramic is not necessarily a good audio capacitor. Ferrites saturate, and even magnetize. Not so good.
I hate to say it, once again, but 30 years ago, we thought that we 'discovered' commercial RFI filters as being a good thing for audio design. We bought all kinds, and now they sit in storage. They just didn't work out that well. I can only presume it is parts quality, not design quality. Please keep this in mind.
The VERY FIRST idea to fix things is to put a commercial, usually reasonably priced, RFI filter right at the input. They usually give specs, and show the circuit even. Who could ask for anything more? Well, the first problem is WHAT they use for component parts. An RF ceramic is not necessarily a good audio capacitor. Ferrites saturate, and even magnetize. Not so good.
I hate to say it, once again, but 30 years ago, we thought that we 'discovered' commercial RFI filters as being a good thing for audio design. We bought all kinds, and now they sit in storage. They just didn't work out that well. I can only presume it is parts quality, not design quality. Please keep this in mind.
- Status
- Not open for further replies.