[R-390] Need help fixing R390A not to trip GFI breaker

[Steve Byan]

On Aug 21, 2007, at 1:59 PM, Barry wrote:

> It is true that the caps may not be leaky; however, if you look at  
> the schematic for the line filter I used in my "conversion", it,  
> too, has caps between both legs and ground.  What puzzles me,  
> though, is why these caps do not set up a current path to ground  
> like the ones in the R390A filters do.  I assume it is the value of  
> the caps used create enough reactance to minimize the current path,  
> but not sure.

If you look at Richard Loken's post:

On Aug 21, 2007, at 1:32 PM, Richard Loken wrote:
>
> Once more with feeling!  Those capacitors ARE NOT LEAKY!  They are  
> 0.1mfd
> capacitors and Xc=1/2 * pi * f * C and that says:
>
> 	Xc = 1 / 2 * 3.14 * 60 * (1 x 10-7)
>            = 26K ohm
>
> 	I = E/R
> 	I = 120 / 26000
> 	  = 4.6mA
>
> and that will trip any properly working GFCI because they should  
> trip at
> between 4 and 5 mA.

I'm not sure which filter you used from this catalog page:

> http://www.mouser.com/catalog/631/958.pdf

but I'll assume it was a "GENERALPURPOSE IEC CONNECTOR FILTERS". The  
Mouser catalog page specifies a maximum leakage current each line to  
ground at 115VAC 60Hz of 0.2mA.

Turning Richard's math around,
	R = E/I
	R = 115 volts  / 0.2 mA
	R = 575 Kohms

Actually we're talking reactance, not resistance, so we're really  
talking about the magnitude of Xc, not R. So I'll switch to using Xc  
in place of R:

	Xc = 1/2 * pi * f * C

	We want to find C given Xc = 575 Kohms and f = 60 Hz:

	C = Xc / (1/2 * pi * f)
	C = 0.575 Megohms / (1/2 * 3.14 * 60)
	C = 0.575 Megohms / 94.2
	C = 0.0061 uF

So the caps in your filter are less than about 0.006 uF in value.

Regards,
-Steve

-- 
Steve Byan <stevebyan at mac.com>
Littleton, MA 01460