Stock AP1 flywheel. Last time I did a compression test it was all solid....doesn't burn a drop of oil even with with 3-4 track days on the car. So I highly doubt leakdown would show anything. I Could give it a shot though.

 

Remove the map sensor and spray some cleaner in the hole. Let it soak a little. Then retry if engine stalls . If nothing happens, borrow a map from a friend (any 92-00 honda will do as long as it has the same shape). I have come across two cases where faulty map sensor had very slow response especially hot. Excess vacuum in overrun retracts the internal membrane a lot and when idle valve restores air, there is not much fuel injected due to map ‘stucking’ and reading a very high vacuum (aka very little fuel). This causes engine stall. Same effect with a semi closed map hole in the throttle body; it slows down map sensor response. But you should have some other symptoms as well (slight hesitation when you go suddenly full throttle at any rpm).

I really cannot think of much more considering your TPS and idle valve is fine.

 

key on engine off for TPS, close enough eh?

 

Now also getting very positive short term fuel trims which has maxed out my LTFT.....that points to a massive vacuum leak somewhere that I cannot find.

 

time to bust out the propane torch..

 

Are you suggesting what I think you're suggesting?

 

using an unlit torch in the engine bay spraying propane slowly to find the vacuum leak, yes

 

Welp. I got out the propane torch....and found nothing. I sprayed it on all the areas that I thought could have any kind of possibility of leaking at all. IAT sensor replaced without any changes. I performed an idle relearn and now it seems to never stall, but the RPM's do get very low.

I guess onto the "real" issue here is the fuel trims are kind of out of control at idle sometimes. I'm a little skeptical of this OBD reader I have, so I ordered a new bluetooth one off of Amazon; that will be here tomorrow and maybe I can confirm some of these things.

I took a data log, which is attached.

A few screenshots. After reviewing the data, I basically found at idle it is adding fuel (+ STFT) that eventually just maxes the LTFT out if I leave it there. When I go drive the car, it evens it all out and I usually have a LTFT in the low single digits by the end of the 20 minute cruise. So, still that is telling me vacuum leak. I also noticed this parameter "O2B1S1" in a "%" format. On all the other cars I've owned, they have had wideband O2's, so it was easy to tell AFR. I'm not sure if this % means Lambda or not. If it means Lambda though, the car is running extremely rich all the time. Which makes no sense either, so I'm going to vote a no on this.

Anyway, here are some screenshots for people that don't want to open the file:

Idle:



Clutch in coming to a stop #1:



Clutch in coming to a stop #2:




I'll edit the top section of things I have tried. Thanks for all the help from everyone so far. Hopefully this will help someone in the future.

Edit: I know this makes it hard with the data acq being so slow as well.....but here's also OBD O2 Voltage vs Time. Keep in mind I did go drive the car during this log. Is seeing 0V normal?

 

I did try this as well even though I am using a new MAP sensor. Figured why the hell not at this point.

 

@HawkeyeGeoff thanks for clarifying about the air pump's nipple.
If you think your car is running really rich based on that O2B1 % reading then wouldn't it be best to install an AEM wideband just to see what the AFR reading really is while idling, cruising and WOT?
How many vacuum leaks could there possibly be? Could an injector or two need sonic cleaning?