Understanding electric fence energizer

[DW86]

Hi all,

In order to better understand how electric fence energisers roughly work, I've been trying to model one in Falstad. I tried a simple setup of taking a 5VDC input which is stepped up to around 30V with a buck-boost converter. The 30 VDC is charged in a capacitor, which is then discharged over a 1:100 transformer to create a voltage spike. I've done the discharging with a manual switch in the model, but normally I suppose this is done with a timer or microcontroller of some kind. I've attached the model as an image below, just after starting the discharge of the capacitor over the transformer.

Now I suppose the negative voltage at the transformer side is negative because the coil of the primary side is trying to resist the rise in current due to closing the switch, but I didn't really expect electric fences to work like that. Do they actually operate on a negative voltage at the fence or am I doing something wrong with this model?

I also noticed that the resistance between the two terminals on the secondary side has to be rather large, otherwise no voltage spike will occur. I don't think 1M Ohm is realistic when an animal touches the fence though. Curious to hear what I'm missing here.

 

[Easy peasy]

dot notation - and you could gnd one side of Vout

we have designed electric fences - and they are not like your design.

 

[BradtheRad]

Since fur is on animals I have a hunch a spark jumps through hairs or along hairs when close to the wire. Before that moment the distance is several inches through air.
Sort of like the Marx generator included in the circuit menu of Falstad's.

We don't know where arcs or sparks go inside the non-conducting secondary of the flyback.
We don't know how high voltage gets in a loadless boost converter.

 

[DW86]

@Easy peasy: Thanks for mentioning dot convention, you were right, it's fixed now. I am curious what the designs you work with look like then. I've based this schematic on the one I saw here (also see picture below):

Reverse Engineering an Electric Fence Charger

Reverse Engineering an Electric Fence Charger: Reverse engineering something can be a great form of both entertainment and education. I’ve often purchased items just for the satisfaction of disassembling them to determine how they function and how they are designed. Along the way you…

www.instructables.com

It also looks similar to the unit we have for the horses, which is powered by a 12V battery. The connection to the fence on that unit is also through a transformer, a fairly large capacitor, an inductor and then another capacitor. I'd assume the inductor and one of the capacitors is for creating a buck-boost converter to charge the second capacitor which is then dumped on the transformer.

 

[betwixt]

Not sure that schematic actually does anything useful at all. Looks like a disaster for the 4024 clock input, an uncontrolled Q1 and an op-amp that could be doing anything, powered between neutral and 15V with its input anywhere between ground and several KV.

Brian.

 

[DW86]

@betwixt, could you explain your concerns about the schematic? From what I'm reading about it, the clock pulses seem fine (see picture below), the transistor Q1 is controlled by the output pin Q7 on the counter and the op-amp input voltage is reduced through a capacitive voltage divider. From the pictures it seems like this is a store bought unit, so I doubt it doesn't do anything useful at all.

 

[betwixt]

I can see what it is supposed to do, my concerns are that the CLK input pin on the 4024 are clearly driven at higher voltages than VDD. It is fed from rectified AC at the input but VDD is dropped to a lower voltage by D2, L2 and R2 and the current through R1, D1 and the ICs. It is using the static protection junctions in the IC to prevent damage. I liken this to driving a car into a wall to stop it instead of applying the brakes, both are effective but one is far riskier than the other.

Q1 base emitter junction is driven by a spike of about 15V as pin 3 of the counter goes high, it has nothing except the junction itself to protect it and it has no leakage to the emitter to fully discharge C4. Again, cost cutting by putting components under stress.

Pin 3 of the 741 is floating to DC, while accepting this is a 50 year old IC and has relatively high input leakage current, it should still be tied to something and protected from overvoltage and to define a default output voltage. A capacitive potential divider is not a good solution to drop the voltage, especially if the PCB substrate is part of a capacitor and subject to environmental effects.

There is also a problem of the whole thing being connected to neutral, this is not a clean signal and connecting the transformer secondary to ground, as is necessary for it to work as a fencer, means the 741 sees not only the spikes from the output but mains borne interference too.

Brian.