sensor grounds

Grounding to a common point keeps ground loops from forming. What happens is that a small voltage drop can be measured from one ground spot to another eventhough in theory, all ground is the same potential. What then happens is a reference voltage of 5 volts may be 5.01 volts for the TPS but the computer's reference voltage is 4.99. Though 2 hundreths of a volt seems insignificant, the TPS on your old GM vehicle is looking for .56 volts with the throttle fully shut.

I would also think from a production standpoint running the wires back to the computer is better than having to ground to the chassis because a common grounding point allows you to have to put in only one bolt to the chassis instead of one screw per sensor, a royal pain in the butt.

-Bruce

I see your point about ground loops. The diagram for my computor shows several grounds circuits going from the computor to engine or chassis ground. I can't tell if the tps ground circuit grounds at the computor case or simply runs through the computor to another pinout where a ground wire runs to engine. That sure seems like a long way around, do you think it is because of ground loops? Now I wonder if each sensor ground has it's own circuit by way of the computor and then back to engine, humm? I have seen some computor diagrams where all engine sensor grounds go to one pinout at the ecm, that diagram wasn't GM, by the way. Just fooling around, I unplugged the tps, used an LED with a resistor for testlite, went from B+ post to ground terminal at the tps connector, the lite came on. Someone told me that circuit wasn't a constant ground, I believe it is. I don't know if a 12 volt testlite would have been safe to use. The wire on my 12 volt lite was larger than the wires at the tps connector so I used an LED. Anyway thanks, interesting.

The reason why sensor grounds go to the computor instead of the chassis is becasue it's more reliable. Imagine the rust and corrusions that can affect a sensor. They don't want to be cheap. Wires are cheap.

I always check it the same as a ground for those that runs to chassis. In almost all circuits, if it's ground, chassis is equally ground, except in a few cases.

Tibur

Thanks for the reply. In my first post, I probably should have used the term engine ground, instead of chassis. I don't know if the engine ground has a different value than a chassis ground, I guess I could check and see? My engine has alot more grime on it than the chassis. I don't know if the tps ground that goes to the computor, grounds through the computor case and then through the body or if the computor runs another ground to the engine block? If said ground circuit had a problem wouldn't it be more reliable, to clip the ground wire at the tps connector, leaving a long enough tail to work with, splice in new wire, then run to negative bat cable, wouldn't that be quicker and more reliable than going through the computor ground circuit? I realize if you did all grounds like that, you would have a fist full of grounds at the neg bat terminal.

Are you saying it is ok to use 12 volt testlite connected to B+ on a sensor ground circuit that goes to the computor? thanks