Fwd: Re: [Amps] Computer Grade Electrolytic Capacitors vs. Radial

[William Fuqua]

>Date: Fri, 06 Sep 2002 10:09:14 -0400
>To: "Steve Thompson" <g8gsq@qsl.net>
>From: William Fuqua <wlfuqu00@uky.edu>
>Subject: Re: [Amps] Computer Grade Electrolytic Capacitors vs. Radial
>
>I struggled with the question for some years until I found some 
>information in my
>Industrial electronics handbook published by McGraw Hill.
>
>         I found a graph ( this book was published in the vacuum tube 
> days) that showed for a
>single phase power supply with a capacitive input filter the ratio of 
>capacitor ripple current to
>load current vs rectifier efficiency. The max value was 1.414 for 100% 
>efficient rectifiers and zero series resistance.
>         Semiconductor rectifiers in high voltage power supplies are 
> nearly 100% efficient due to
>low forward drop vs output voltage. It seems to me if that ratio is square 
>root of 2 then there
>should be a simple explanation. But I have not found one quite yet.  I 
>have not found article that
>has addressed the issue. And the ones that say anything about it suggest 2 
>or 3 to one (probably a
>safety factor of 1.5 or 2 included).
>         I am trying to locate book now but can't find it. I may have 
> loaned it out.
>
>73
>Bill wa4lav
>
>
>
>At 10:41 PM 9/5/02 +0100, you wrote:
>
>>-----Original Message-----
>>From: William Fuqua <wlfuqu00@uky.edu>
>>To: Ian White, G3SEK <g3sek@ifwtech.co.uk>; amps@contesting.com
>><amps@contesting.com>
>>Date: 05 September 2002 16:24
>>Subject: Re: [Amps] Computer Grade Electrolytic Capacitors vs. Radial
>>
>>
>> >The 1.414 factor is the maximum assuming zero source resistance. It will be
>> >less with series resistance.
>> >
>> >73
>> >Bill wa4lav
>> >At 01:07 PM 9/5/02 +0100, Ian White, G3SEK wrote:
>> >>Steve Thompson wrote:
>> >>>
>> >>>>For an inductive input filter the RMS ripple current is equal to the
>>output
>> >>>current if you have  a >single phase  supply. For a capacitive input
>> >>>filter the max ripple current will be equal to the >square root of 2
>> >>>(1.414) times the DC output current.
>> >>>
>> >>>
>> >>>Doesn't it vary with source resistance and C/R values? I use a rule of
>> >>>thumb of 3x for capacitive input - it was very instructive putting a .1
>> >>>ohm resistor in the bottom of the capacitor stack and hanging a 'scope
>> >>>across it. The example circuit in the Duncan software has a cap input PSU
>> >>>putting 450V across 5k. DC current is 89mA, rms current in the cap is
>> >>>226mA and 248mA in the transformer winding.
>> >>
>> >>I haven't checked that example, but one "gotcha" in the Duncan program is
>> >>that the default results in the tables include the power-on cycles where
>> >>current can be very high.
>> >>
>> >>The solution is to set the "Report from" time a few cycles in.
>>
>>Good point Ian, I hadn't spotted that. I changed the delay before analysing
>>from 100ms to 1000ms - should be well into steady state. Now, dc current is
>>89mA, capacitor rms current is 202mA and transformer rms current is 221mA.
>>Ok, it's a bit lower than the original values but either stack up reasonably
>>with what I measure in practice. When rms current tends to be dominated by
>>peak I^2 and the current pulse in the cap is a narrow spike varying with
>>source and load R and cap value, it seems to me that the dc/rms ratio could
>>vary widely, and readily exceed 1.414. Can someone put me straight here?
>>
>>Steve
>>
>>
>>_______________________________________________
>>Amps mailing list
>>Amps@contesting.com
>>http://lists.contesting.com/mailman/listinfo/amps