Sorry to keep beating what might be a dead horse by now, but I've been
thinking about the issue with my Altima again where the revs suddenly jump
from around 1100 to 2000 as I am easing into the throttle... and I might
have an explanation for it.

It's a 99 Altima, 5 speed, with 117,000 km.

When I got the car it was idling around 1000 RPM even when warmed up
completely.  I was able to adjust the idle, but only down to about 750rpm
with the TPS disconnected, and the service manual says it should be 650 rpm,
+/- 50.

However, with the TPS connected normally, the revs settle around 750, which
is within spec (700 +/- 50 IIRC).  What I realized today, is that what is
probably happening is the ECU is closing the idle control valve further than
normal in order to maintain the target idle speed.

With that in mind, I suspect the ECU is programmed to more or less match the
opening of the throttle by opening the idle control valve as much as it can
so that when I take my foot off the gas, the engine doesn't immediately
spool down, which can be annoying when coasting.

Now, given that the IACV is closed further than normal at idle, and assuming
that the ECU is programmed to open the IACV with the throttle (I suspect
it's probably a linear map of TPS angle to IACV opening?), then it makes
sense that as I reach the beginning of where the IACV starts to open, it
jumps to what would *normally* be only slightly more opening than at idle,
but in my case because it's more closed at idle, there's an abrupt change
instead of a smooth increase in speed.

Does that make sense to anyone else?

If that's what's going on, then where is the extra air coming from that's
causing my idle to be too fast?  Is there anywhere else metered air might
get into the engine?  I suspect maybe I have a PCV issue because the revs
will drop to 400 rpm when I am in idle adjust mode with the TPS unplugged if
I pull and plug the PCV breather tube between the valve cover and the
intake.  There's a fair bit of vacuum at the valve cover fitting at idle
which is adding to the airflow through the MAF at that point.

I'm hoping I can find where the line from the PCV valve to the intake
manifold is connected, and disconnect it to see what effect that has on
everything...  too bad the PCV valve is so well buried on these cars!