How to select the correct AC line noise suppression caps?

[thorpej]

This isn't strictly a "tube audio" issue, but it's something I have been thinking about in that context...

I'm interested in learning the correct way to calculate the appropriate value for AC line noise suppression caps, for filtering RF noise and random transients from the AC line before they hit the power transformer (the kind that might happen when, for example, the mini-fridge in your office fires up its compressor). Perhaps my Google fu is not strong enough, but I have found plenty of information about the kind of safety and voltage rating the caps need to have, but nothing about how to select the capacitance value.

 

[mroboto]

I made a few of these power bars.

http://audiokarma.org/forums/index....aps-additions-questions.747210/#post-10131613

Check for Class Y2 ebay item (not affiliated) 191220268223 for line to ground and neutral to ground.

And 0.1uf 275V Class X2 for line to neutral.

 

[analog addict]

http://www.justradios.com/safetytips.html

http://www.justradios.com/X1Y2capacitors.html

 

[Tom Bavis]

Line-to-line after the power switch can prevent a POP when switching power off - .047 or 0.1 uF will do it. Line-to-chassis should be .01 or .015 but NO HIGHER to minimize leakage current (through YOU!). Can be omitted if a 3-wire cord is used.​

 

[thorpej]

Right, so I'm familiar with the justradios article about safety caps, and I understand where you use the X vs Y type. My question is how do you go about picking the specific value of [0.047uF to .1uF] for the across-the-line cap? If it's a rule-of-thumb, what's the math behind it? (I always use 3-wire cords, so no need for line-to-chassis caps in my usage..)

 

[drtool]

OP, I understand your question and wondering the same thing myself. Do tube amps like (clean power) as much as SS.

 

[thorpej]

OP, I understand your question and wondering the same thing myself. Do tube amps like (clean power) as much as SS.

Well, noise is going to make it's way in, and if it's high-frequency transients (like those annoying "POP!"s that happen when he mini-fridge fire up), you can hear them! Tube or solid state, doesn't matter.

(For my main system at home, the point is somewhat moot, because I run my entire AV system on conditioned power, but I figure I should start adding these noise suppression caps to my builds anyway as a best-practice.)

 

[BinaryMike]

Greater line-to-line capacitance yields greater differential mode noise suppression almost without limit, but values above 1~10uF begin to alter power factor. It's usually best to add L/C filter sections instead of just increasing capacitance when you get near this range.

 

[gadget73]

With two wire cords and talking specifically about line to chassis caps, its more about maximum potential current flow from chassis to a ground source. Mostly its intended to keep someone from zapping themselves. 0.01uf at 60hz is basically 260K impedance. With 120 vac, thats a maximum current of 4.6 microamps.

Its about 30 ohms at the bottom of the AM broadcast band though. It gets lower as the frequency goes up too. If the goal is RFI suppression, I suppose you could calculate impedance based on expected interference frequencies. At some point I doubt it really makes any difference though. Once the impedance is sufficiently low, making it a teensy bit lower still won't improve anything.

 

[thorpej]

Ok, got it. Thanks guys!