We're doing a physics project that involves a version of the Kelvin Water dropper experiment. Everything is working great, and we have the following system: system

In the image: two plastic containers (one in the bottom and one in the top). The bottom one is filled with water and has a water pump that pumps water to the top (using tubes for circulation). The water drip from two holes in the top box, go through two cans and into two (half) bottles. One side will become charged (say positively) and when we connect the cans and the bottles in an X-shape as seen in the image, the other side becomes negatively charged and thus we have a sort of "battery" (potential difference).

We would like it to work the same, only without connecting the cans and bottles in an X-shape immediately using crocodile clips (which requires fiddling with the cables inside the system). Instead we would like to connect it from "outside", meaning somehow extend the reach of the cables.

We've tried to connect two cables (a series of 2) on each side instead of one but that didn't work (also with thick cables that have less resistance), and we're stuck. We need to somehow figure how to make the same connection every time without putting hands inside the system, and would love to hear some ideas.

Thanks and advance! Also, we're sorry if this question seems too broad or not detailed enough - please ask if something is not clear.

  • $\begingroup$ It would probably help to link to a simple description of the experiment (the version that you are doing if possible) because while I know I've heard of this bit I can't bring the details to mind. Second, the photograph is a useful reference, but having a labeled schematic as well is generally helpful for people trying to understand the photo (and to establish a uniformity of jargon in the answers). $\endgroup$ Commented Feb 14, 2016 at 21:08
  • $\begingroup$ Unfortunately we can link only to this simple description en.wikipedia.org/wiki/Kelvin_water_dropper (only instead of spark gap we use it to create a potential difference elsewhere). We're actually in the process of making a labeled schematic but that might take a while. Thanks for the quick response! $\endgroup$
    – Trouble
    Commented Feb 14, 2016 at 21:14

2 Answers 2


You don't need thick cables because the currents that are flowing are minute. What you do need is excellent electric insulation. To make a "switch" with easily available materials, get yourself a piece of plexiglass sheet, drill holes for panel mount banana jacks and solder half of a crocodile clip cable on the jacks. As long as you keep the plexiglas sheet clean, you will be able to do your wiring changes with banana jacks from the outside and the experiment should work fine. Make sure that the distance between the jacks is sufficient, a couple if inches should work just fine.


To achieve the "excellent electric insulation" mentioned by @CuriousOne in his answer what you must not allow is the black and red wires in your diagram to touch each other or anything else which is not as good an insulator as you might think eg wood.
The insulation of those red and black leads in your diagram may not be good enough at the high voltages generated by your machine but also layers of surface moisture may compromise the insulation properties. So, as much as possible, keep those leads away from everything and dry them and the plexiglass sheet with a fan heater or a hair dryer.

Try and avoid points or sharp edges on the high voltage parts as charge can be lost by corona discharge.

Here is a picture of an electrostatic generator. Note that those cross connections which are made of metal are not even covered with insulation. Of course they are insulated from one another by the air.

enter image description here

The black and red leads are of such a length that they hang in mid air.

The two cylinders at the bottom right are Leyden jars(capacitors) to help produce a beefier spark Please note that with Leyden jars in the circuit because more charge is stored before a spark occurs the apparatus if=s much more dangerous.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.