[Prev][Next][Index][Thread]
Re: Get it over with
At 11:25 PM 2/7/97 -0700, you wrote:
>Subscriber: Kilroy-at-bscn-dot-com Fri Feb 7 22:51:17 1997
>Date: Fri, 07 Feb 1997 00:15:03 -0600
>From: Kerry Ludwig <Kilroy-at-bscn-dot-com>
>To: tesla-at-pupman-dot-com
>Subject: Re: Get it over with
>
>
>>>I am looking to build a circuit to filter out the 120Hz ripple of the
>>>rectified output from the trans. At least down to tolerable levels. Maybe
>>>less than 5%? I planned on using seperate filters for protection.
>>>
>>>Kerry "Kilroy" Ludwig
>>>Kilroy-at-BSCN-dot-com
>>>
>>>
>>
>>Kilroy,
>>
>>Ripple (60 hz) is normally handled by the filter capacitor and not the
>>inductor. The inductor is sort of a current spike filter.
>>
>>Richard Hull, TCBOR
>>
>
>How did I calculate the size of the capacitor needed to properly filter the
>circuit?
>
>Kerry "Kilroy" Ludwig
>Kilroy-at-BSCN-dot-com
>
>Kilroy,
If your DC rectification is full wave and 5% ripple is ok then I would
figure on a cap that would hold a charged voltage to 5% of peak under a
specified load for 8 milliseconds. (120 hz ripple). Let's say you have a
constant 50 milliampere load and your peak charging voltage is 5,000 volts.
This means your load impedance is on the order of 100,000 ohms. It is
assumed that your source impedance(transformer) is much stiffer than this.
(Z source <5,000 ohms) If this is the case, then you have a simple RC
network with a time constant of 80 milliseconds. (I like to give the system
filter time constant about ten times longer than the ripple period)
So we need a capacitor which can yield a Time constant of 80 milliseconds
with a 100,000 ohm resistor. Now the TC = RC or C= TC/R or .0000008 farads
= 0.8 ufd.
As a praticing engineer I always double the calculated value to allow for
dorks at the output end having lied about the load they were going to have.
So I would use a 1.6ufd capacitor of about 10,000 volt rating provided I was
using full wave rectification from a bridge circuit. Caps like this have a
list price of $200.00 or more. I can pick 'em up all the time at ham fests
for $5.00 each. Mylar or paper units are the best.
All the above mathematical gyrations assumes a steady unvarying load of 50ma
-at- 5000volts. (never ever the case in a telsa coil.) It assumes a very
stiff transformer (never ever the case with neon sign transformers.)
Good luck.
Richard Hull, TCBOR