Here is one i have been fighting for years. I wonder if anyone else has any ideas.

2000 Civic EX (d16y8) (250k ish) (MT)

Car will run completely fine for maybe 10 miles. At which time the idle will drop extremely low (but never die). car will shake, oil light comes on as the pump it barely spinning at that rpm. Temp is right in the middle of the gauge where i would expect it. If i pull over and shut it off for a few minutes it will be good again, but the time/distance it takes for it to start dropping again will be less.

No appreciable difference with the clutch held in or not, or if in gear or not.

There was a time i thought this was a head gasket issue, but it almost feels impossible at this point.

Things i have done:
Reman head install
OEM head gasket.
OEM gasket set everywhere else during head install
Torqued to 65 pounds with ARP studs (in correct order)
plugs, wires
distributor
fuel injectors
bled coolant system
checked coolant system pressure. dropped about 1 psi per minute with no external leak
plugs look normal
TPS is correct calibrated
replaced IACV
Compression is between 180-190 on al cylinders
set idle to the book
set timing
replaced oxygen sensor
no CELs
3" exhaust with new Cat
Smokes for vacuum leaks (nothing)
took IACV off and plugged holes with a piece of cardboard, idles as it does with IACV unplugged (fully closed)
Blocked throttle body opening, engine dies as one would expect.
Set throttle cable slack
IACV function appears fine. Idle with unplugged is low (450 or so) and raises when plugged in, so it is indeed opening and closing.
valve lash adjustment to the book

im sure there are few things im forgetting but ive been through everything can think of so far and this issue persists. The only thing even slightly anomalous is the coolant pressure test.
I cant for the life of me figure out why it gets worse the longer its running. I also dont understand why the IACV is letting it ever idle as low as is does when it starts to get bad. I can raise the RPM with the pedal, and the IACV should be doing this on it's own.

im considering something is mechanically dragging and pulling the RPM down (maybe power steering pump or something else on belt), but unless i completly misunderstand the IACV it should let in a little more air and keep the RPM in spec

 

I had the same exact problem in my 2000 DX hatch D16A.

How's the stock cat/manifold look? any cracks where the upper 02 sensor screws in?

It was my lower 02 sensor not reading right because of the common crack that's located on the upper o2 sensor bung, Eventually it threw the P0420 code and that's how i narrowed it down, Very easy fix. Just added a spark plug fouler to the lower O2 sensor, I used the spark plug fouler or the cheap way but they sell these too,

PLM O2 Fouler Cel Fix Spacer — Speed Science

Watch this,

 

Nothing obvious as far as cracks go, but i will certainly look closer.

I took it for a decently long drive (highway and city) and captured data. (excel file attached)

To me it looks like it is almost always lean and adding a fair amount of fuel to compensate.

Ive only replaced the post-cat o2 at this point. manifold o2 could be the original as far as i know.
Map sensor is almost certainly original, as well is the TPS.

Maybe the data log will mean something to someone

 

take out the idle screw in the throtle body and shoot some carb cleaner through the hol, then adjust the idle with the screw so it hits 750rpm when warm and see what happens.