On Sat, 3 Aug 2002, Morgan Olsson wrote:

>Me again trying to repair this electric fence unit...
>Seem no problem redesigning the charge cap charger, and fire cirquit.
>
>But...
>
>I have earlier repaired an electric fence unit in which a microcontroller
>adjusted the charge according to the load.
>(Trial-measure-increase/decrease style)  That one was a whopping 9 Joule
>per pulse unit.  (If anyone wants to know, it was only the FET switch

Ahh. You do direct shawarma from the sheep, right ?

>Most units i have seen is 1.5 to 4 Joule.

I don't know about sheep but people (like me) tend to react very badly
from 20mJ up.

>The one i repair i calculated to 6 Joule capacitor stored energy.
>Scaring, it has room for another equally large capacitor, yeilding 12J!
>
>It is possible this IC incorporate some clever function for load
>adaption, but it seem that would be an add-on ckt to connect to 3 unused
>pin connectors.  Maybe only used with the larger capacitor bank.
>
>Now, is it recommended to just shoot 6J in every shot...?  I mean
>personal safety. I know for sure i will keep my hand at safe distance, as
>the 9J machine blew 2W carbon film resistors like New year bombs when i
>tried to test it, and wire wound types arched...

Imho you need to know the capacitance of the wire to know this. Is there
some kind of voltage or current feedback from the output ? (look carefully
it may be a current transformer implemented as PCB traces).

I'm no expert on fence zappers but usually when you have an
inductive/capacitive load like several hundred feet of fence wire then you
treat it as a RLC resonant circuit and either tune it into the generator
or excite it by pulse. For example by placing a spark gap in series with
the generator secondary feeding it. The goal is to up the voltage on the
fence RLC to something that will go through wool. This may or may not take
6J but you will never know until you know the fence impedance. Do you have
a RLC bridge to measure the fence ?

One method I've seen used to test a fence (for cows) was to make a string
of 12 neon bulbs and connect it to the fence through a 100R resistor and
to GND. If they flash there is enough energy (12 neons fire at about 1kV).
Then you would turn down the converter until they stop flashing, then turn
it up until it starts flashing again. An active feedback would be good
because in bad weather there will be more leakage.

The only fence generator I've ever seen used two valves (!), one as a slow
oscillator the second as self oscillator exciting the fence through a
spark gap (no capacitor, no rectifier). It worked forever, you just had to
change valves and and electrolytic filter capacitor every ten years or so
(or so I was told - I only changed them once for the owner). I heard that
pure neon bulb based designs existed once upon a time. They may have been
thyratrons but who knows.

Peter

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu