Hello, if I put a 4,7 microfarad cap (which I own) across mains (in europe 220V), can this have real benefit in cleaning noise, power factor correction, clean harmonics.... etc...??
or 100 nF is enough?
Thanks
or 100 nF is enough?
Thanks
A large capacitor will pass a significant current at mains frequency.
Like anything to do with the mains, leave well alone unless you know exactly what you're doing. And what problem you are trying to solve.
Capacitors can do all those things, but as a designed response to a measured problem, not just adding a value at random, which will just as likely make things worse and possibly do damage.
If you have noise issues, try a filtered mains lead or inlet, if you have the other problems, hire a design engineer.
Like anything to do with the mains, leave well alone unless you know exactly what you're doing. And what problem you are trying to solve.
Capacitors can do all those things, but as a designed response to a measured problem, not just adding a value at random, which will just as likely make things worse and possibly do damage.
If you have noise issues, try a filtered mains lead or inlet, if you have the other problems, hire a design engineer.
Hello,
First, the cap should be rated for going across line voltages. X rated caps are made to fail safe if they should fail, not like most of non X rated caps. They do cost more than most, but are required by UL and those concerned with safety and the lack of fires.
Second, if you were to put this across the line voltage, need to add a bleeder resistor such that the cap gets discharged when the power is off. Otherwise it could be setting there with voltage across it and shock you badly.
Unless you have really bad power, much smaller cap should take care of any RF and hash coming into the line.
First, the cap should be rated for going across line voltages. X rated caps are made to fail safe if they should fail, not like most of non X rated caps. They do cost more than most, but are required by UL and those concerned with safety and the lack of fires.
Second, if you were to put this across the line voltage, need to add a bleeder resistor such that the cap gets discharged when the power is off. Otherwise it could be setting there with voltage across it and shock you badly.
Unless you have really bad power, much smaller cap should take care of any RF and hash coming into the line.
4.7uF charged to around 300V would be enough to make you jump. Hence it would need a bleeder resistor.
Cleaning noise? No, a filter would be better.
Power factor correction? Yes, if you have an inductive load like a motor or an old-fashioned fluorescent tube light fitting.
Clean harmonics? 4.7uF is far too small to make any difference.
100nF is enough for what?
Cleaning noise? No, a filter would be better.
Power factor correction? Yes, if you have an inductive load like a motor or an old-fashioned fluorescent tube light fitting.
Clean harmonics? 4.7uF is far too small to make any difference.
100nF is enough for what?
Hello,
Small caps are normally used to keep RF filtered out. Not unusual to have a filter on the incoming power at all. A IEC connector input with a filter and switch is the easiest way I know of to do this.
Small caps are normally used to keep RF filtered out. Not unusual to have a filter on the incoming power at all. A IEC connector input with a filter and switch is the easiest way I know of to do this.
At 50H, 5uFd is 637 Ohms reactance.
My power line series impedance is 0.4 Ohms at an outlet; yours may be 0.1 Ohms.
637r hanging on 0.1r (even 0.4r) does almost nothing.
It may do some at higher frequency. At 500Hz the cap is 64r and the line may be 0.2r. Still hardly anything. At 5KHz the cap is 6 Ohms, the line may be 1 Ohms, and this just may attenuate 5KHz one dB.
Back at 50Hz, the 5uFd cap will pull 0.37 Amps of steady current. This is small, and may or may not register on your Electric Bill (depending how your supplier handles reactive energy metering). If fully billed, about $11/month.
The resistor is perhaps needed and ALWAYS a good idea. If the cap gets disconnected from the wall near a peak of the AC wave, as DF96 says it will hold 300V for some time. If you put your fingers on the "dead" plug, you get a 300V shock. Through skin, this could be 300mA on a 5mS decay. This is more than is usually allowed for human contact. While a shunt transformer would bleed the cap fast, transformers or fuses may fail "open". Big power capacitors always have a permanent bleeder.
My power line series impedance is 0.4 Ohms at an outlet; yours may be 0.1 Ohms.
637r hanging on 0.1r (even 0.4r) does almost nothing.
It may do some at higher frequency. At 500Hz the cap is 64r and the line may be 0.2r. Still hardly anything. At 5KHz the cap is 6 Ohms, the line may be 1 Ohms, and this just may attenuate 5KHz one dB.
Back at 50Hz, the 5uFd cap will pull 0.37 Amps of steady current. This is small, and may or may not register on your Electric Bill (depending how your supplier handles reactive energy metering). If fully billed, about $11/month.
The resistor is perhaps needed and ALWAYS a good idea. If the cap gets disconnected from the wall near a peak of the AC wave, as DF96 says it will hold 300V for some time. If you put your fingers on the "dead" plug, you get a 300V shock. Through skin, this could be 300mA on a 5mS decay. This is more than is usually allowed for human contact. While a shunt transformer would bleed the cap fast, transformers or fuses may fail "open". Big power capacitors always have a permanent bleeder.
The 370mA is about where I figured too, but thought of it more like a 75 watt light bulb hanging on the line. 🙂
Wouldn't the impedance of the device(s) connected down stream determine how well the cap worked an HF drain?
Wouldn't the impedance of the device(s) connected down stream determine how well the cap worked an HF drain?
What is an "HF drain"?
The effectiveness of a CR low pass filter depends on R and C, where R is the net source impedance i.e. it includes the effect of the load too. However, in any mains supply situation the impedance of the load is always negligible when compared to the source so the load can be ignored.
The effectiveness of a CR low pass filter depends on R and C, where R is the net source impedance i.e. it includes the effect of the load too. However, in any mains supply situation the impedance of the load is always negligible when compared to the source so the load can be ignored.
I have a large film motor run across the 120v mains here, isn’t going to consume much power anytime soon I don’t believe. Not about to use my meter to measure it either, but it’s always been ice cold...
Even a tenth that size makes you jump. A cap across the mains was commonplace in older TV's with 0.1uF being typical and even that can bite if it captured the mains at its peak.
Lol, we used to have an engineer that always seemed to get caught by these... imagine engineer standing at bench, mains lead in hand bashing the plug top on the work surface to smash it to pieces. He always felt better after that, even though the plug had to be replaced.
Our new Hotpoint washer bites severely if you unplug it and happen to touch the L and N on the plug without waiting a few seconds.
0.47uF is the largest I have actually seen on domestic equipment and I would class that as 'dangerous', or at least severely unpleasant to the uninitiated.
4.7uF isn't going to do any cleanup of 50 Hz - 20 kHz noise on the AC mains. It's only effective in the 100's of kHz to low MHz.
A Torus Power isolation transformer would reduce noise above 2 kHz. It's -12 dB per decade above 2 kHz. But they aren't cheap!
Firstly, do you have a hum noise issue? Secondly, look at cleaning it up with changes to your signal cabling and its runs, signal cable isolators, etc. That'll be a lot cheaper and more productive than monkeying with anything on the AC side.
In general I'm amazed at the focus on trying to fix power issues on the AC side of things where very little actual improvement is possible without designing/buying a product with a better power supply.
A Torus Power isolation transformer would reduce noise above 2 kHz. It's -12 dB per decade above 2 kHz. But they aren't cheap!
Firstly, do you have a hum noise issue? Secondly, look at cleaning it up with changes to your signal cabling and its runs, signal cable isolators, etc. That'll be a lot cheaper and more productive than monkeying with anything on the AC side.
In general I'm amazed at the focus on trying to fix power issues on the AC side of things where very little actual improvement is possible without designing/buying a product with a better power supply.
The performance of any filter will depend on the target noise and how well they align to cancel or damp any noise.
"Downstream" is a purely human concept. The electrons (well, electric charge) roam wherever they want.
And for a tube driving a capacitive load, you mostly work with the source impedance. Technically source||load. But nearly always source < load. 12AX7 with 100K RP and 220K next-grid, source is 39K and source||load is 33K- little difference. 15%?
In Utility Power circuits, 15% sag under load is dim lights that flicker when load changes. We more often work to 5% sag, or load is 20X higher than source.
Any house-size power supply will be a few Ohms at most.
That's at 60Hz. Power lines are typically very inductive. On my 500' line, not so it really matters. (0.05r inductance, 0.35r simple metal resistance.) But there's miles to the dam generator and many miles to the gas/electric station. OTOH these are increasingly fat wires to big generators of good regulation. You can't fight a power company with a puny 5uFd.
Interestingly, if you put hundreds of uFd, in an industrial area, someone might PAY you. Because inductance dominates power network action, it is frequently economic to add capacitance across the lines. Aside from lines, large motors are inductive. They pull a current waveform out-of-phase with the voltage wave. The difference is not Real Power. But it adds to real heat in the wires and transformers, which must be oversize. Or the inductance approximately cancelled by shunt capacitance. BIG capacitance: 40KVAR at 240V is 1,000uFd.
Not counting ESR, which may be small; no, it does not consume POWER.
But current sloshes back and forth from the generator through the cap and back. This heats wires and transformers.
*Do Not Try This At Home!!* Measure the temperature of your feeder wires, with and without the cap. Control all other loads, and supply voltage. Give time to stabilize. They WILL run warmer. As mentioned up-thread, perhaps 0.3Amps warmer. For a fist-size cap, probably not so you can read with common thermocouples. Anyway measuring "live" metal can be deadly.
Traditionally, industrial customers have large reactive current for large motors. Homes with (incandescent) lamps and heaters are/were nearly pure resistance. And industrial customers have experts; residential customers are clueless.
Industrial power is usually billed in several "chunks". There is the Real Power, heat (or work done). There is reactive energy, current sloshing back and forth heating the wires even though no useful work is done. This ratio is Power Factor. And there is a Peak demand charge to encourage steady all-day loading rather than short surges.
Your cross-line cap "should" incur Power Factor charge. But PF metering is costly, PF problems are very rare in residential loads, and you can not explain a PF charge to Joe Homeowner. (Also a cap is usually "good PF", though I have never seen that distinction made.)
The cap is just a frequency dependent voltage divider in that circuit, like most filters. In this case the higher the frequency, the lower the impedance.
I suppose it really won't make much difference at audio frequencies if the mains impedance is low.
And it wouldn't do much for common mode noise, anyway.
The cap is not a frequency dependent voltage divider. It is half a frequency dependent voltage divider. The other half is the mains impedance.
Which is what I said, or so I thought. Maybe I just assumed that folks here would know that a voltage divider isn't made from one component. 😛
BTW - the cap is half of a frequency dependent voltage divider. Better?
BTW - the cap is half of a frequency dependent voltage divider. Better?
So the thermal part is spread out over the inductance/cabling, makes sense now.
Please forgive my ignorance as to thinking that would have been isolated to the cap in question...
Speaking of sound quality and line conditioning, I have found that a 2.5k R across line and neutral to lower the noise floor, when located near the device, is a ham radio trick I picked up.
That could be a 7w incandescent bulb. Am guessing that there is a damping effect, haven’t got out the scope to try and make a visual.
Most of the line conditioning i have tried is like the cleaner-wax you get for your car, polishes a little bit but won’t take out any deep scratches.
Please forgive my ignorance as to thinking that would have been isolated to the cap in question...
Speaking of sound quality and line conditioning, I have found that a 2.5k R across line and neutral to lower the noise floor, when located near the device, is a ham radio trick I picked up.
That could be a 7w incandescent bulb. Am guessing that there is a damping effect, haven’t got out the scope to try and make a visual.
Most of the line conditioning i have tried is like the cleaner-wax you get for your car, polishes a little bit but won’t take out any deep scratches.
- Status
- Not open for further replies.