[Tubeburner]

I plan to install 1.0uF 400v 0r 600v capacitors across the receptacles of my power conditioner. Is there a big difference in capacitors, say a Auricap compared to a Solen for noise absorption?  The Auricap is $22.00 and the Soln $3.00. Any advice is appreciated!

Thank you!

[Occam]

Yes, there is difference, both objectively and subjectively. The actual construction of the caps define its ESL which limits a mains interference capacitor's efficacy at increasing frequencies.

I find that I can get needed results with appropriately purpose rated X2 and Y2  capacitors, and it addresses some liability (and reliability/safety ) concerns. That being said, I've never heard of Auricaps failing when applied to these functions. Audience, in their own power conditioners use their 600vdc caps in 120vac mains products, and 1200vdc caps on 240vac products.
Never use a foil type (as opposed to a metalized) capacitor on mains, as they are not self-healing.

[mgalusha]

+1 on Paul's response in all regards.

[Folsom]

X and Y rated caps are guaranteed to fail in a safe manner. Using other capacitors from the audio market is probably more than perfectly safe, just not guaranteed. (insurance companies won't blame an X2 capacitor, etc)

For one individual capacitor, ones from the audio market may tend to have a wider bandwidth so they will do a bit more filtration, and tend to have very good (at a cost) ESL/ESR properties. Good ESL/ESR you could say attracts the frequencies for filtration. (its actually more complicated than that). If you use two .47uf instead of one 1uf, you'd lower the ESL.

For safety reasons do not use under twice the mains voltage. In most cases you'll be over, or way over, if you live in a 120v country.

[S Clark]

Would you guys mind explaining to a non EE type?  What are X and Y rated caps?  What is ESL, ESR, and what is "good" esl/esr?

[Folsom]

PDF talking some about ESL and ESR with capacitors coupling.

Here is ESL and ESR wiki's. (lower is better for power conditioning)

Here is an explanation of the X and Y names.

[kinku]

This may be a stupid one.If you look at the topology of Felix the filtering cap before CMC are actually common to power supply line( they all are parallel to live and neutral ). If they are all parallel why there  is a need for a capacitor bank before each CMC. There can be one bank of 1,0.1 and 0.01 uf common to all CMC and still work the same.Have anyone tried using a huge Felix with all this pre CMC banks in one unit (which will all together form a huge bank of filter caps) and found better than one with just one single bank of caps before CMC?

[Speedskater]

Remember that for 120V 60Hz units the total capacitance to case is 78nF.  For 240V 50Hz units it's 41nF

[kinku]

Speedskater can elaborate on  what you said, for illiterates..please?

[Speedskater]

It's from:

Power and Grounding for Audio and Video Systems
A White Paper for the Real World – International Version
Jim Brown - Audio Systems Group, Inc

pages 17 & 18

LEAKAGE CURRENT AND APPLIANCES
Equipment connected to the power line will draw current through capacitors intentionally connected between the power line and its enclosure (noise filters), and through unintentional parasitic (stray) capacitance and resistance inherent in power supply components. The return path
for this current is the Equipment Ground (Green Wire, PE).
The resulting IR drop along the Green Wire (PE) raise the potential (voltage) on the enclosure.
IEC 61140 established four classes of electrical appliances that may be connected to the mains supply based on their leakage current.

*****************
Capacitance is the primary source of leakage current, so we can solve the equation for capacitive reactance and current for these two classes of appliances and standard power systems.
*******
A single Class I appliance is permitted to draw no more than 3.5mA, but the leakage current on one Green Wire (PE) can be significantly greater if several appliances are plugged into multiple outlets connected to the same Green Wire (for example, four outlets in a permanently installed "quad box" or a multi-outlet strip plugged into an outlet.


I skipped much of the chapter and it was difficult to copy what I posted.

see:
http://www.audiosystemsgroup.com/SurgeXPowerGround.pdf

[Occam]

Speedskater,

The Felix power filter has no Y caps at all (although its acceptable to use a Y cap for an X cap's across the line function). There are no caps from either hot or neutral to ground. From my experience, I've found that dumping noise onto the safety ground causes more problems than benefits. Given that, it would take a herculean effort to make a Felix that's ground leakage merited concern.

Kinku,

This may be a stupid one.If you look at the topology of Felix the filtering cap before CMC are actually common to power supply line( they all are parallel to live and neutral ). If they are all parallel why there  is a need for a capacitor bank before each CMC. There can be one bank of 1,0.1 and 0.01 uf common to all CMC and still work the same.Have anyone tried using a huge Felix with all this pre CMC banks in one unit (which will all together form a huge bank of filter caps) and found better than one with just one single bank of caps before CMC?

Excellent question! Yes, one could combine those front end caps which are all in parallel to each other and the incoming mains. But substituting one 5uf cap for 5 1uf caps, or 1 .47uf for 5 0.1uf, etc.... will compromise performance. All things being equal, multiple caps paralleled will offer lower ESR and ESL than an equivalent single cap of the same total value.

Regards,
Paul

PS - I will again reiterate my caution against using anything other than properly rated X or Y caps for the X cap function on the Felix. There are NO objective reasons or metrics for using anything else. [I can't speak to Audience's teflon film caps in power conditioning]  If, for subjective reasons, you choose to ignore the liability issues, I will point out that Audience uses their own caps dc rated at 5 times the mains ac voltage.

[Folsom]

Audience also has there stuff in a pretty decent metal box.

You are not technically worried about capacitance between mains and ground, you just use capacitance to measure leakage. Leakage over a certain amount is bad, causes problems, and isn't considered safe. The ratings on capacitors used in these applications have so little leakage they would have to be utterly failing to come close to the limits.

I prefer a neutral to ground capacitance, by a large margin. It has to be done correctly though (you can't do it for one device in a system). Audience uses isolated ground sockets, just to point out.

[Speedskater]

Well in any case my statement only applied capacitor's connected to the case or Safety Ground (EGC).

But by coincidence, in noticed a different cap value in a newer Jim Brown PowerPoint then in a second coincidence I saw in Henry W. Ott's new book.
 That various safety agencies have different leakage limits. The UL limit (0.5mA) so the max value is 0.01 uF.

Those that have access to a technical collage library, Henry W. Ott's new book 'Electromagnetic Compatibility Engineering'
has 30 pages in Chap. 13 about power line filters.

[Speedskater]

'DofS' the leakage is dictated be the cap value, the line frequency and the line voltage. It works as a coupling cap.

Yes, mild steel makes the best filter enclosures.

[Folsom]

Steel is fairly fire safe.

And you are wrong on leakage. You need to read more on how capacitors work. In theory no amount of capacitance would ever have leakage, but capacitors are not perfect so there is some. Manufactures give this rating. I'm not doing well at explaining this to you, but the point is that I know what you are talking about and you have it wrong. Capacitance measurement is used to measure leakage, but leakage is not capacitance. No one would ever rate leakage based on something that has zero current flow (perfect capacitor), because it would only have potential and no true leakage. No current flow = no leakage.

Again check the leakage rating on capacitors. It is so far below all of the safety ratings that it is not a concern in these applications. Depending on capacitor quality you might have to be dealing with hundreds of farads (not micro) to get a leakage that is notable.

[Speedskater]

That's a different leakage!  This is the intended current flowing through the capacitor in the normal way that capacitors work (they pass AC current). This is filter or circuit leakage.

On line to line caps, Mr. Ott writes:

Because this cap is not connected to ground, its value is not limited by leakage requirements.  Typical values for this cap range from 0.1 uF to 2 uF.  For safety reasons, a resistor, typically 1 Mohm is sometimes added in parallel with the cap. This resistor is used to discharge the caps when power is removed.

You should read his book.

[Tubeburner]

Wow, my many thanks to all who provided information and are still helping with this topic. This is very educational to me. It's great to be in an audio community that shares knowledge.

Thank you!
Ernie

[Folsom]

That's a different leakage!  This is the intended current flowing through the capacitor in the normal way that capacitors work (they pass AC current). This is filter or circuit leakage.

On line to line caps, Mr. Ott writes:

Because this cap is not connected to ground, its value is not limited by leakage requirements.  Typical values for this cap range from 0.1 uF to 2 uF.  For safety reasons, a resistor, typically 1 Mohm is sometimes added in parallel with the cap. This resistor is used to discharge the caps when power is removed.

You should read his book.

No current flows through a capacitor. By definition they stores electrical charge and release it on the same side it was stored on. Look at the schematic symbol for a capacitor, it shows that there is no connection via current flow. You linked to a PDF talking about using capacitance as a measure of leakage (with a discrete circuit that uses leakage to form capacitance). If you applied that method of measuring to a situation with a Y cap, your instrument probably wouldn't be able to pick up a reading of capacitance anymore than without the Y cap, if done like instructed in the PDF.

Your example doesn't argue or try to say otherwise from what I've said. He is only pointing out that the amount of leakage a capacitor has doesn't mean anything across line, but you do have to take it into consideration in a Y configuration (I point out the amount that any remotely respectable capacitor at a value we'd use has so little it isn't worth mentioning). It isn't saying that in a Y configuration it is leaking current equal to capacitance, because that wouldn't be true of a capacitor. That'd be true of something like a resistor from line to ground.

[Napalm]

No current flows through a capacitor. By definition they store current [...]

Electric charge, Mr. Destroyer, electric charge....

By definition current is the rate of flow of electric charge, it cannot be "stored".....

And while current does not flow "through" a capacitor, it will "flow" through its *terminals*....

[iluzun]

"Never use a foil type (as opposed to a metalized) capacitor on mains, as they are not self-healing."

Sorry, but what is 'self healing'.  I've several pr. of .22 caps lying about from rolling in my preamp.
On my Audio Magic Nexus conditioner,  can I replace the .01 Sonicaps from line to neutral on the input iec
& the outlets, w/say Clarity MRs, but not CuTf V-Caps?  How to determine an alternative optimal value when
moving from a .01?

I understand that different 'materials  can still impart
their sonic signatures even when not directly in the signal chain.  I've heard plastic, metal, & oil all thru various  conditioners on my 'mini' server located on a different level of our house, wirelessly to a router, &
wirelessly to my Modwright Transporter.

Perhaps the MRs vs the Auricaps would make a nice comparison? 

Thanx for any input....