Back in the early Ď80s I was an electrical design engineer for an automobile manufacturer. One day the experimental engineering garage called on me to diagnose and remedy a strange, periodic snapping noise that surfaced during the development of a new vehicle.
Peering down into the dark, I confirmed that the noise was emanating from a large, arcing spark down low in the engine compartment. The arc was crossing between a small, metal in-line oil filter can and the vehicle's grounded chassis frame some distance away. The spark was big, fat and loud. It was at least two inches long and occurred about once a second whenever the engine was running.
In place of a conventional, vacuum-boosted brake system, this particular vehicle incorporated an experimental hydraulically boosted power brake system. It required an additional hydraulic plumbing from the engine-driven power steering pump to the brake booster system. The plumbing included a fluid filter inserted in-line to the pump's rubber, high-pressure feed hose. The metal filter housing was suspended by the insulated hose a distance away from the chassis frame.
I quickly concluded that a two-inch-long spark equated to at least a couple hundred thousand volts. My first thought was:† "How do you get that kind of voltage with a car's 12V system?" I found it particularly confounding because the vehicle, which used a diesel engine, didn't even have a high voltage ignition system!
I was familiar with generating very high voltages and their principles of operation, having built my own Van de Graaff generator and Tesla coil as a kid. I concluded that, somehow, we had inadvertently produced the "hydraulic" equivalent of a conventional mechanical Van de Graaff generator, which consists of three main parts:† an insulated motor-driven electron transport belt, a metal electron collector brush connected to the high voltage metal dome at one end of the belt and a source of electrons applied to the opposite end of the drive belt.†
In my conceptual "hydraulic" equivalent, the moving, non-conducting hydraulic fluid and insulated rubber hoses were the electron transport mechanism. The metal filter can and its internal metal filter element made up the electron collector. The engine-driven hydraulic pump and drive belt were the source of electrons to the fluid.
Since the engine-driven pump was of an all-metal construction, and was grounded to the engine and the chassis frame, it was difficult at first to envision how the pump could be a source of electrons. However, I substantiated my suspicion that the pump was the electron source when I attached a grounded test lead to the pump by rubbing it against the pump's belt-driven pulley and caused the arcing to stop. When the engine was stopped, a conductivity check confirmed that the pulley was grounded in this non-operational state.
I then concluded that the pump's spinning pulley and internal rotor assembly were electrically "floating" inside the grounded pump housing, due to the hydrodynamic action of the bearings and rotor and the insulated seals inside the insulated hydraulic fluid. The actual electron source probably resulted from the triboelectric friction of the rubber drive belt on the pulley.
Grounding the filter can housing to the vehicle frame eliminated the arcing symptoms. Upon further reflection, grounding the filter can only provided a good sink for the electrons, and there was still a large circulation of those electrons in the hydraulic fluid. I wonder what, if any, detrimental effects that electron flow would have had on the fluid and the system's other parts. As I recall, this configuration of power brake booster vehicle never made it into a production vehicle.
About the Author: Arthur Sundeen lives in Michigan, has a BSEE, holds 15 patents and runs his own electrical OEM companies producing aircraft instruments and radio antennas of his own design. You can reach him via our Sherlock Ohms blog comments.