the lesson here is that simply cleaning a ground connection may not fix
the problem. a number of posts mentioned cleaning off ground contacts to
improve such things as "ratty" idle - some with success, others not - but
without actually using test equipment it is difficult or impossible to
tell if electrical grounding/bonding is really the culprit, or if indeed
anything was really fixed by simply cleaning off connections.

On Sun, 13 Jul 2003 03:42:27 -0400, Daniel L. Katz <katzd54@YAHOO.COM>
wrote:

>recent adventures in '84 vanagon electrical bonding from first time
listee:
>
>1. as per list suggestions, i measured dc voltage between body and engine
>at high idle with accessories on and, sure enough, had a reading of
>roughly 250 mV. it turned out that indeed the port side bonding strap
>between body and engine was a high resistance. the solutuion, however, had
>nothing to do with cleaning contacts or snugging down bolts. instead,
>there was an 8 ohm contact resistance where the "engine side" connector
>was crimped onto the copper braid. obviously, the bonding strap was
>useless in this condition, and carried very little current; the crimp on
>the body side was good. new, slighly longer connectors, crimped and
>soldered onto the original bonding strap (for good measure, along with the
>many individual grounding wires from ecu, etc), solved the problem.
>
>as a side note, the much heavier bonding wire underneath van between frame
>and transaxle case was in excellent condition.
>
>2. the temperature guage on this van was unusual in that it read low, and
>had a very small range - way down there on the left end of the scale.
>serendipidously, i discovered a roughly 800 ohm resistance between the
>upper thermostat housing, where the temperature sender is grounded via its
>threads, and the rest of the engine. evidently, there is corrosion or goop
>or something keeping the mounting bolts from making good electrical
>contact. now, this 800 ohm is itself considerably greater than the
>resistance of the sending unit at operating temperature, and could easily
>explain the behavior of the guage. i simply provided a shunt using a test
>lead with alegator clips, and now have reasonable readings on the guage,
>with the needle below the led at operating temperature.
>
>note: this van resides in mcminnville, oregon, and will be for sale in the
>near future.