Fuel pressure regulator and STFT

Car
I have a 97 Honda Prelude, it has about 96k miles on it.
It has stock H22 engine, 10.8 compression, uses 94 octane.It has a return based fueling system, using an old school vacuum based fuel pressure regulator. It is a speed/density based system using the MAP sensor.
Has always been kept in excellent condition, oil changes, valve timing, belts etc.
It's driven very gently barely crossing 3500 rpm , never used VTEC which activates at 5500 rpm.
Plan to keep it forever

Observations:
I noticed I have very high negative fuel trims on deceleration and idle.
I can see jumps ranging from -10 to even -22% in some areas.
As the rpm increases the STFT becomes a decent 2 to 3% and even sometime remains at 0 at higher rpms.

Guessing- the FPR struggles to keep the pressure in check , maybe faulty FPR or a weak vacuum. So as the speed increases the trims look ok either because the engine needs that much fuel or the effect of the faulty FPR simply works better with low vacuum , the low vacuum compliments the weak vacuum and happily provides the higher pressure as diaphragm is not applying that much pressure.

Current state of other sensors:
-AFR primary O2 sensor works perfect , ranging from 0.1 to 0.8Volt, waveform is symmetrical on both sides of .45
-Secondary O2 sensor works perfect , holds a steady 1 volt or 0.8v , but when the rich condition happens it jumps up and down a bit (upto 1.27 volt)
-The MAP sensor is near perfect , ranging from 0.9 V to 2.8V [as per factory shop manual it should be 0.9 to 3V]
-LTFT is around -3% , never seen it go higher than -5% or -6%

IMP - once in a while I get a P0138 from the B1S2 secondary sensor that it's sending out very high voltages for a long period of time( greater than 5 seconds as per shop manual).
This has always coincided with state when the engine was running rich and the STFT was doing -18% or -22% trying to cut the fuel.
I have an aftermarket exhaust but it has been there for 3 years and no issues, its just recently I started getting these P0138 so I'm guessing the FPR is causing richer conditions

I use both generic and Honda factory proprietary scanner modules.
- DTC and corresponding freeze frames, Mode6 from the generic module
- VTEC, Intake bypass solenoid, MAP voltages etc from the Honda module

Help needed
-This weekend I'm going to try and check the fuel pressure with/without vacuum line , pinch the return line etc and see if the pressures are within spec.
-Clean the MAP with MAF cleaner spray (maybe help the voltage signal come back to 3V instead of 2.8V)

Please advise what I need to look at ? any inputs greatly appreciated.
I need this intermittent P0138 to go away ..what else can I check ?
(I also have purchased the book, I'm working my way through it )

Symptoms when driving from cold to warm?

I didn't record that but I can go and observe today.
Generally I heat up the car before plugging in the Obd2 , it's usually -15 below zero here .

What would you like me to log ? is the following good enough ?
I'll just switch it on and let it warm up till it reaches closed loop
-STFT
-LTFT
-Rpm
-B1S1 and B1S2 voltage
-Engine load
..anything else ?

In terms of driveability there is no loss of power , hesitation etc while going from cold to warm or extremely hot after a hour long drive.

With a gas analyzer you can see if O2 sensors are lying! O2 lazy or slow!? Need to test them loking with a scope!

On deceleration normally looking at shoter term fuel trim has no real value as the injectors are shutoff completely until the engine approaches idle speed. As the longterm fuel trim is well within acceptable limits unless you have excessive fuel or other running symptoms you may have nothing to major worry about ,

But I am puzzled by the occasional P0138 1.25 volts from the downstream O2 sensor --- I would be be wondering about the downstream sensor itself or its wiring. I would want force rich then lean conditions with added propane then introduce & vacuum leak to do oxygen storage test on O2 sensors and cat

ISTR on Hondas of that era it was quite easy to contaminate the upstream O2 sensor by flooding the cam/rocker cover with oil when refilling with oil after an oil change.

On deceleration normally looking at shorter term fuel trim has no real value as the injectors are shutoff completely until the engine approaches idle speed.

ok ..that make sense.No point watching fuels trim in deceleration.

As the long term fuel trim is well within acceptable limits unless you have excessive fuel or other running symptoms you may have nothing to major worry about ,

ok ..I'm at -3% LTFT now even with the spikes of -15% to -22% STFT trims. Long terms has always been between -3% to -6% , currently its settled at -3%

But I am puzzled by the occasional P0138 1.25 volts from the downstream O2 sensor --- I would be be wondering about the downstream sensor itself or its wiring. I would want force rich then lean conditions with added propane then introduce & vacuum leak to do oxygen storage test on O2 sensors and cat

Based on your doubts I just went and checked the pending codes and I see a P0136 also (O2 sensor circuit malfunction B1S2)
So currently I have a stored code of P0138 and a pending of P0136

ISTR on Hondas of that era it was quite easy to contaminate the upstream O2 sensor by flooding the cam/rocker cover with oil when refilling with oil after an oil change.

I think you have very good point here.
The H22 engine uses FRM cylinder liners which burn oil and require the engine oil to be refilled quite regularly. So I'm quite frequently topping up the engine with oil and then driving off immediately. Maybe a few ml every second week.

I may be contaminating the upstream and downstream 02 sensors without realising
All this freshly poured dripping oil could be easily sucked into the exhaust ?
I think I'll try and clean both the sensors also and be more careful to not drive off immediately after filling.


Another question - does a 2.81V from the MAP sensor be of any concern ?
At KOER the MAP should output 3.0 V but I just read 2.81V
I can just spray MAP some MAF sensor spray into the tube if its dirty causing it to not to generate the full 3V ?

Thanks !!

With a gas analyzer you can see if O2 sensors are lying! O2 lazy or slow!? Need to test them looking with a scope!

ok ..I can go to the local emissions guy here and see if I can get some data

Tried to measure the pressure from the fuel rail ..but the prelude has a weird size banjo bolt without a schrader valve. So the gauge did not fit.
Will be getting a new gauge - thinking of ATD fuel pressure tester with the special adaptor

P0136 and P0138 is really pointing to a sensor or wiring fault.
For pressure checks very often on the cars I work on it is easier to connect at the rear. I have two right angle push lock connectors that allow me to connect from the top in the rear seat area.

P0136 and P0138 is really pointing to a sensor or wiring fault.

got it ..both are pointing to second sensor
-It's like the heater circuit is out of spec on KOEO and I get a P0136
and then when the car is being driven I get a P0138.
The high spike in the STFT of -16 , -22 etc coincided with the P0138 codes so I was focussing on the fuel pressure.
ok ,I think it makes more sense to check the O2 sensor circuit instead of the fuel pressure then.

For pressure checks very often on the cars I work on it is easier to connect at the rear. I have two right angle push lock connectors that allow me to connect from the top in the rear seat area.

Thanks, problem with me is that the Prelude me to remove the rear seats to get access to the fuel pump.
It's located right in the middle of the chassis making it really inaccessible.
Prelude has a M12x1.5 adaptor built into the fuel rail for pressure checks so its very accessible ..I just need to get the convertor so that it can provide a schrader valve for the pressure gauges to attach to.

But I think I will now focus on the o2 sensor 1st.