mlloyd1 said:
Agreed! I like to go for the low hanging fruit (as "they" say) first. Curiously it's amazing how much of an audible difference this makes.
I keyed into this years ago after moving from a house that had been rewired to a house with ceramic stand-offs and cloth coverd AC wiring. Couldn't understand, at first, why the resolution and lowend of my system took such a hit. A 12ga run direct to the breaker box brought things back to normal. These days the system is sitting directly over the fuse panel with a dedicated line that runs approximately 18 feet. I'm very lucky.
Regards, Mike.
Hi John,john curl said:
10uF has an effective impedance of about 265ohms at 60Hz.
That is going to draw a lot of current. Besides voltage, is there any other specification parameter that must be met to allow this to work safely? Or does "motor start" make it robust enough.
If I were using 240Vac and 50Hz, what value of cap would you suggest? 6uF would pull the same current but maybe it should be a lower value for the 240Vac.
Andrew,
I think John meant to say 'motor run' capacitor (typically a metalized polyprop in oil with a built in failure disconnect, 'protected') rather than 'motor start' cap which is often a specialized electrolytic, not appropriate for this application.
While an appropriate 'accross the line' capacitor will shunt differential (normal) mode noise, a multi microfarad cap is going to have substantial ESL limiting its efficacy to below <1 MHz, and appropriate bypassing extends it. The impedance vs. f graphs are interesting -
http://www.wima.com/EN/mp3x2ue.htm
FWIW
I think John meant to say 'motor run' capacitor (typically a metalized polyprop in oil with a built in failure disconnect, 'protected') rather than 'motor start' cap which is often a specialized electrolytic, not appropriate for this application.
While an appropriate 'accross the line' capacitor will shunt differential (normal) mode noise, a multi microfarad cap is going to have substantial ESL limiting its efficacy to below <1 MHz, and appropriate bypassing extends it. The impedance vs. f graphs are interesting -
http://www.wima.com/EN/mp3x2ue.htm
FWIW
265 Ohms will draw nearly half an amp at 120Vac. This is not a trivial load--almost 60W. It's not common to have to worry about heat dissipation in a cap, but this is one of those times.
I used something more along the lines of a .1uF cap across the line, then an isolation transformer (which I still had sitting around after a tube project), followed by another .1uF cap, a humbucking inductor, then another cap. All this is from memory. Use whatever you have on hand.
I didn't "spend" anything on the experiment, as it was all stuff I had on hand. The isolation transformer was obviously the most expensive item, but I had picked it up cheap at MPJA or one of the other online surplus houses. All told, I think I could build something like it for about $20, perhaps less if I was patient enough to wait for an isolation transformer to pop up cheaply enough.
For simplicity, you can't beat John's cap-across-the-line approach, but that only gives you 6dB/oct attenuation. I prefer a sharper slope and less load on the line, so mine was set higher. I haven't really thought about it in a while, but I'd probably do things a little differently now...at least in part because I've got other things in my junk box that need using up.
For DIY purposes you can build this stuff into the box and it becomes a permanent part of the power supply. It need not be a separate box.
Grey
I used something more along the lines of a .1uF cap across the line, then an isolation transformer (which I still had sitting around after a tube project), followed by another .1uF cap, a humbucking inductor, then another cap. All this is from memory. Use whatever you have on hand.
I didn't "spend" anything on the experiment, as it was all stuff I had on hand. The isolation transformer was obviously the most expensive item, but I had picked it up cheap at MPJA or one of the other online surplus houses. All told, I think I could build something like it for about $20, perhaps less if I was patient enough to wait for an isolation transformer to pop up cheaply enough.
For simplicity, you can't beat John's cap-across-the-line approach, but that only gives you 6dB/oct attenuation. I prefer a sharper slope and less load on the line, so mine was set higher. I haven't really thought about it in a while, but I'd probably do things a little differently now...at least in part because I've got other things in my junk box that need using up.
For DIY purposes you can build this stuff into the box and it becomes a permanent part of the power supply. It need not be a separate box.
Grey
We have been adding caps across power supplies commercially for more that 12 years. Others did it even before we did. Perhaps, you are right about 220V supplies. Maybe a slightly smaller, higher voltage cap would be best.
This filter is really 12dB/octave when you realize that the series inductance of the power line is part of the equation. Do a computer simulation and see for yourself.
This filter is really 12dB/octave when you realize that the series inductance of the power line is part of the equation. Do a computer simulation and see for yourself.
With any cap across the line it makes sense to put a bleed across it - 100K will do.
Otherwise if there is a switch downstream in the off position and the mains plug is then pulled it can leave the cap charged to line peak voltage.
Not very likely but my fingers can vouch for the possibility
6uF can give quite a belt ....
Otherwise if there is a switch downstream in the off position and the mains plug is then pulled it can leave the cap charged to line peak voltage.
Not very likely but my fingers can vouch for the possibility
6uF can give quite a belt ....
Toshiba
Hi,
I found this on a Toshiba web site, they still make a lot of j-fet products.
http://www.semicon.toshiba.co.jp/openb2b/websearch/resultDo.jsp
Are they useful in audio design?
Kamskoma
Hi,
I found this on a Toshiba web site, they still make a lot of j-fet products.
http://www.semicon.toshiba.co.jp/openb2b/websearch/resultDo.jsp
Are they useful in audio design?
Kamskoma
john curl said:
John,
Yes, but the power line inductance is always there, regardless of what we put on our end. For that matter, the step-down transformers and series resistance of the wires will factor in, too, but I'm thinking more in terms of what we can bring to the table, be it 6dB/12dB/18dB/etc. The more the better as far as high frequency rolloff. At least we know the frequency and it remains (relatively) stable.
I have yet to go as far as to use an amplifier with a 60Hz sine generator on the front, but I get tempted every time I think about it. Being out in the hinterlands, I used to have good AC, but those days are gone.
dimitri said:
Dimitri,
It ain't easy trying to juggle a wiggly eleven month-old boy, a bottle, a telephone, a four year-old wild woman, and type...all at the same time. I got as far as calculating the .5 A and no further.
Grey
- Status
- Not open for further replies.