Power ground

[Clarence L.Snyder]

Clarence Wood wrote:
> 
>   I have a 1982 280ZX turbo with the Bosch L-Jetronic fuel injection.  I have been told that after modifying timing, etc. the battery ground should be disconnected for a period of time in order to reset the ECM.  I have also been told that both, negative and positive, terminals should be disconnected.  The time interval has been anywhere from 5 min to overnight.  Does the power ground,  as described below, explain why both terminals should be disconnected to reset the ECM??  Or am I reading to m
>
Most (japanese) ECUs have several power supplies - at least one of which
is full time, non switched, to maintain non-volatile memory (it's
volatile with power disconnected) On a Toyota, for instance, there was
both an ECU 1 and an ECU 2 fuse. Removing the right one would reset the
codes on the ECU in about 15 seconds - unless you had the brake pedal
depressed, in which case enough voltage filtered through to maintain the
memory. Removing both terminals will definitely ensure the memory is
dumped, but will also, in all likelihood, loose all settings on your
radio etc.
Remember, CMOS memory is basically a whole slew of capacitors which hold
a voltage to indicate o logic one, and no voltage for logic zero (unless
you are running negative logic, where it is reversed) Virtually vo
"power" or current flow is required - so ANY voltage applied to the
circuit can prevent resetting.

 
> At 12:14 PM 1/2/99 -0500, you wrote:
> >Geoff Richards wrote:
> >>
> >> Anyone have time to explain the difference between 'ground' and
> >> 'powerground'?
> >> TIA
> >> Geoff
> >I'll give it a crack.
> >In a DC power system on a car, the negative(usually) terminal uses a
> >"chassis ground" for power "return".This is your "power ground". It can
> >have a voltage drop of up to .1 volt across it for accessory and
> >charging loads, and up to one volt for starter loads.
> >For signal circuits, such as sensors, sometimes the .1 volt drop across
> >the frame ground could cause calibration errors, so a "signal ground" is
> >provided. Sometimes the signal can even "float" above chassis ground, as
> >through a protection diode with a, say, .6 volt forward voltage drop.
> >Can't use chassis ground return for such a signal.
> >Hope this helps understand it.
> >
> >