Wankenstein

Okay.

ST o2 correction (%): between 28.9 and 33.6
LT o2 correction (%): between 6.2 and 4.7

Okay

There's a a rubber cap on a vacuum inlet on the TB where evap purge looks to connect to.

Spritzed in TB cleaner and rpm's raised and idle smoothed for a few seconds. Tried it a few times and each time the same result.

Not sure. Are you thinking that if that's the case than less fuel might be injected than needed for a 240cc injector?

I pulled the TB to inspect idle screw passage and it looked clear. I do regret putting RTV silicone on the TB gasket though, as it took longer to scrape off. If any of the RTV got into the intake and cylinder would it possibly cause damage? I put a very light coating on it but when compressed from tightning it may have squeezed inward towards the bore.
Car ran fine for weeks afterwards until this current issue. Currently running well except at idle. I'm purchasing a new TB gasket tomorrow.

ezone

Seems to confirm the engine is running too lean.

Same check at 2500 RPM, does it still make it run better?
These numbers at 2500 RPM support the above.


I'm also curious about a wide open throttle test run up to redline through first and second gears (Maximum fuel demand), watching the O2 sensor voltage the entire time.....does it peg at .9v or .1v?

I was thinking the injectors for a 1.5 engine are too small to run a 2.0 correctly, fuel trims would always end up very high as the computer continually tries to add fuel..
It was a thought.

At operating temperature, neutral, no electric loads....What is MAP value at idle, and again at 2500?
Is ECT looking accurate at operating temp?

Have you tried a known good ECM?

Wankenstein

Trying to get those things done but now a new problem. I removed the TB yesterday evening when the engine was at operating temp to check and clean passage if needed. The RTV silicone I stupidly applied a couple months ago got on TB bolt threads..I got it off with wd-40 and brass brush. I have to wait for a day for an OEM gasket so I figured I'll try a felpro for $3. It matched up with the old OEM as far passages and bolt/stud holes but will not seal and I have at at proper torque spec.

Maybe the TB or intake mating surface warped because I removed the TB at operating temp or the because of the RTV I used. Maybe Felpro's gasket isn't working properly. I called felpro and they said check both surfaces with a straight edge and that a thicker gasket won't help. If the TB is warped it's not a real loss because it apparently has a sagging butterfly. The intake would need to milled if it's warped, correct?

I don't have a straight edge at my place (at mom's) but I used the smooth side edge (opposite of teeth side) of a 10" hacksaw blade to get a rough estimate and both looked very flat. I laid it crosswise, diagonal etc.. on several areas of both intake and TB mating surfaces. I plan to get my straight edge and feeler gauges to check it again. Just ordered an OEM TB gasket from local Honda dealer for however this plays out. I'm starting to believe that I am what's mostly wrong with my car

I attached some pics of TB and intake ..the only passage hole in the previous OEM and new felpro gasket is the upper left hand corner just below the stud. Not attaching for flatness inspection...lol

Wankenstein

I forgot to mention that the idle scew had RTV on it too and I cleaned it up as much as possible. May have prevented air from passing as I mentioned previously that adjusting the screw did not effect idle.

ezone

I can't picture a gorilla warping a manifold there.

Might have lumps of old gasket still stuck on a surface somewhere causing a high spot?

Stick TB and manifold together WITHOUT a gasket and see if they mate up nice and flat with eachother.

I remember thinking the IAC counts didn't look out of line on the scanner so throttle body and IAB was probably set ok and flowing a sufficient amount.

If the engine was actually running correctly (without fuel trim and delivery problems) you could watch the scanner while adjusting the IAB screw, you might hear airflow change and you might be able to see the IAC value change as the computer compensates for what you are screwing.

ezone

Oh hey I see it looks like you've got a manual trans IAC on the manifold....Have you taken it off the manifold and scraped the carbon buildup off of the screen?

Wankenstein

I scraped both mating surfaces carefully with a razor blade and throttle body spray. I'll try without a gasket to see how they mate up.

Wankenstein

About four months ago I cleaned the IAC and screen with TB cleaner. I didin't scrape the screen but it looked clean. I added a bit of RTV on the IAC rubber gasket at that time out of laziness because I didn't want to wait for new gasket/screen. The rubber gasket is very flat and hard and I felt might be a source of leakage. No more RTV for me other than where service manual says to use it. I'll order a new IACV gasket.

There's several 60mm TB's in Florida listed on Car-Part.com some are shown in local salvage yards...if needed.

Wankenstein

When I get the TB straightened out (literally) I intend to do a DIY smoke test as shown below in vid. Reading up on lean idle the following possibilities are what we are trying to rule out?:

Vacuum leak
Exhaust leak
MAP sensor
o2 sensor
Cylinder misfire
Inadequate fuel delivery

Anything I'm missisng?

ezone

Alcohol content in the fuel?

See the scotty kilmer video where hes smoking a big stogie for a smoke test.

Wankenstein

I did a cigar smoke test last week and I used a ziploc bag with an intake clamp over it to cover the TB opening. The only smoke I saw was coming from near the ziploc bag which is very near the TB gasket. At that point I had the old OEM + RTV combo TB gasket. Could have been the bag seal was leaking or could have been TB gasket..looked more like the ziploc seal though.

Thankfully, I used an Arturo Fuentes Meduro nubby cigar or i would have gagged myself to death with the backpressure. I didn't check for exhaust leak though ..only intake. The setup in the video will allow a steadier, thicker flow to check intake and exhaust.

Is there a procedure to test alchohol contenet in fuel?

Wankenstein

I removed the felpro Tb gasket and the TB and intake looked vey flat without noticeable space. I used a precision level and .002" feeler gauge measuring in every direction and both the TB and intake wouldn't allow the feeler gauge to slip under the level. On my way soon to pick up an OEM gasket from dealer.

If it seals I will get the other test finished you asked for. Smoke test will probably be another day.

Just got back from Honda and they ordered in an intake manifold gasket instead of a TB gasket..so, TB gasket will be here tomorrow. Parts guygave me: Genuine Honda Parts 36531-P1H-E01 Oxygen Sensor price $300. I'm thinking the Walmart $30 Denso# 234-4099 sensor may not be the fix. Unless, other things contributing to poor readings..in which case the old o2 sensor I removed may still be good or was never the right one in the first place..hence the intermittent, erratic, dropping idle this car has had since before I purchased it.

HondaPartsNow (the bestest) sells the same genuine sensor for $200: https://www.hondapartsnow.com/genuin...1-p1h-e01.html

Another site selling that sensor states this: "The 36531-P1H-E01 Heated Oxygen Sensor is interchangeable with part numbers 36531-P28-A02 & 36531-P28-A01 Heated Oxygen Sensor". Also states the part is manufactured in Japan and is direct fit for 92-95 Civic EX as well as Del Sol.

Updated: Doing more research it shows in the cross reference table (pdf) http://wiki.24subaru.ru/images/7/74/...rence_Rev1.pdf that the genuine OEM 36531-P1H-E01 is the same as Denso# 234-4099 and
NTK# 24542.
Now my head hurts.

ezone

I think the O2 sensor is reporting as it should, or as it sees it.
The engine appears to be too lean on the scanner.....and you said adding carb spray made it run better.

Wankenstein

Okay. I may pull the injectors and spray them into a bucket to see what they are doing. I've read it can be detrimental to an engine to continue running lean so I will try to drive it as least as possible until issue is resolved.

ezone

Look up some DIY injector cleaning?

I wonder what a stock 2.0 CRV would show for IPW-ms at hot idle, and compare to yours.

Wankenstein

I've seen a few youtubes where people pull the injectors, use a 9v battery, inject seafoam or other cleaner.

Maybe you'll get one in the shop tomorrow.I am requesting one from guy in the video below and hopefully he'll reply:

Wankenstein

OEM TB gasket solved the leak.
I found a tiny intake leak on PCV valve hose but still has lean issue.
At 2500 rpm TB spray drops rpms for a second and then seems to smooth engine a bit but, harder to tell compared to spraying it in at idle.

.9v

Idle: 4 psi
2500 rpms: 3 to 4 psi

Yes

Not yet

Fuel pressure on scanner/app reads 38-42 psi at idle points. Fuel pressure dial gauge on top of fuel filter reads around 32 at idle then jumps up to 38 during initial acceleration (in neutral) to about 2000 to 3500 rpms and then shows lower psi (around 30-32) when rpms are between 2600 and 5500 rpms. At what point in the fuel delivery system does the scanner read psi?

Me wife is getting tired of me using her car yet, I'm worried about running lean.. Is it okay if I drive mine back and forth to work?: 4 miles each way, 4 times a week

The parts consultant at local dealer said if we can't get it resolved there's a 20 year dealer mechanic that has a lot of experience with Frankensharts like mine. So, if need be that's an option. I plan to pull injectors tomorrow to test spray. May do smoke test too if I can rig up a bucket like shown in the video I posted. If it passes those two test then it seems like it may be an ECU issue.

One thing I've noticed about this engine even while I was test driving it before purchase. It always has a smell in the engine bay like mild burnt or heated rubber (not belts related)...it's tough to describe other than that.

Doing some research it could be that the p75 ECU may need a tuning or may not be the optimal ECU for this setup..maybe it needs to be dyno'd, chipped and tuned?: http://hondaswap.com/threads/ecu-for-b20-vtec.514606/

ezone

Did sealing up the TB change how the O2 sensor acts, or does it still read 0v@idle?

MAP is displayed in a value I'm not used to seeing. Can you switch it to display volts or inHg (vacuum)?
Does BARO display the same way?

Fuel pressure is supposed to change in relation to changes in manifold vacuum, so at idle there's 20" of vacuum applied to the pressure regulator, fuel psi is lower, then when TB is wide open the manifold vacuum goes to almost nothing, so fuel psi rises accordingly.
You can just pop the vacuum line off the regulator (at idle) and see the pressure change.
The factory provided no way for fuel pressure to be directly monitored though, so IDK where your scanner program is getting its value from, if it actually represents fuel pressure, or if it is calculated or inferred or based on a model.

Good, sufficient fuel delivery is available, so that means the combination of fuel pressure/flow and injector sizing is adequate for maximum demand....now need to figure out why it looks so lean at idle/low ranges.
Great, you're finding stuff that seems to agree with where my thinking is headed.

Wankenstein

Unfortunately, no..still stalling and zero volts when idle rpms drop to 300 and lower.

Checked but, PSI only.
This chart might help: http://injector-rehab.com/shop/mapsensor.html

This read shows me why (graph near bottom of page) you would like to see voltage: http://www.aa1car.com/library/map_sensors.htm

This read descibes a bit about MAP + PSI: https://www.yourmechanic.com/article...sor-map-sensor

Same..PSI only

This linked forum below has a lot of Honda swap/tuner enthusiast so I will post there concerning which ECU/Engine combination is best. Many dyno shops in the Tampa bay area too. Link:https://www.tamparacing.com/forums/index.php

ezone

I've become so accustomed to reading this in volts, it's difficult to remember 6 or more other ways other manufacturers or scanners may interpret it. (I can change how it's displayed on my scanner, but usually stick to volts) 15+ years ago I was far more familiar with various others because I dealt with them regularly, and could convert in my head without a lot of stress.

MAP @idle engine at operating temp, park or neutral, with no additional loads: 0.80v-0.95v, is expected.

Hondas used speed density, and MAP is the major player in determining engine load, thus determining the base fuel quantity calculation.
(If the engine uses a MAF sensor it's not speed density)

Wankenstein

Want me to test it with a voltmeter? I could pick one up at a junkyard for $5 .. not into throwing parts at it but, I can swing five beans without a setback if it doesn't help.

ezone

Backprobe it with a voltmeter is good enough...
Also check the other two wires on that sensor for a good ground and 5 volts. If either is off that would skew the output signal.

Got a vacuum gauge handy? Confirm you have 18-20" of manifold vacuum at idle. Just for giggles. I think you will unless the valves are too tight.

Wankenstein

Okay

I do have one but may have left in my son's car (my old '97 EX) which is in St. Louis until mid August.

Wankenstein

I should have asked this sooner: is this just a lean condition at idle or under throttle as well?

ezone

You said the O2 sensor voltage hits .9v during acceleration, so it appears it is not lean during that condition.

From what I saw it looked lean at steady throttle and no load, but had some changes as you were raising rpm in the videos. I can't tell what it would be doing while you drive though, you can watch that.

If you watch the O2 sensor response as you drive, you can see when AFR goes richer than 14.7:1 (stoichiometric) when the sensor shows higher than 0.45v.


Whenever the sensor reads greater than 0.45v, the AFR mixture sensed is richer than 14.7:1.

Whenever the sensor reads less than 0.45v, the AFR mixture sensed is leaner than 14.7:1.

The O2 sensor "switches" its output voltage high or low as it senses the air/fuel ratio crossing the 14.7:1 threshold....the ECM continually adds and subtracts fuel injection quantity (displayed as the fuel trim values) as it attempts to keep the sensor voltage switching back and forth rapidly, this keeps the average air/fuel ratio (near) 14.7:1. (closed loop feedback)

In your case (at idle and no load?) the ecm can't add enough fuel (fuel trim) to get the sensor to switch away from low voltage state (near 0v). I have been thinking this is because the base fuel table isn't suited for the engine it's hooked to, and have been trying to rule out other possible causes.

The ECM takes MAP, ECT, RPM, TPS, IAT and calculates how much fuel to inject based on a predetermined table (fuel map) then the O2 sensors allow the ECM to make minor fuel corrections (closed loop feedback) to keep mixture as close to 14.7:1 as possible during most low load and driving conditions.

Wide open throttle (max power) demands a fuel mixture much richer, more like 12.8:1 (actual might be 12.5-13.3, :1), and that can't be monitored by a plain O2 sensor, but the sensor shows high voltage when mixture is rich.
The ECM does not attempt to make minor fuel corrections during high loads, and ignores the O2 sensor (open loop, relying on the rpm/load/temp sensors and the fuel table to calculate fuel injection quantity)


/still no coffee

Wankenstein

Okay..will monitor. So far no stalling at idle with a/c on. Other than poor idle the engine seems smoother than ever while driving.

I will call a few dyno shops tomorrow and see what they recommend for which ECU and need for tuning. ECU is presently a P75..here's related info:

Check out reply# 8: https://honda-tech.com/forums/hybrid...-swap-1897725/

https://honda-tech.com/forums/hybrid...u-b20-2534866/

Google hits: https://www.google.com/search?num=50...k1.ky1r14wnU8I

Wankenstein

Reply #2: http://forums.g2ic.com/forum/assista...ng-and-the-lot

"You really want to use an ecu mapped for the B20. The LS ECU will run the engine fine in closed loop but in open loop it will run it lean. I was seeing AFRs in the 16-17:1 in open loop while using a P75 on my B20. It just doesn't deliver enough fuel for the B20 until it's compensating using the 02 sensor."

Reply #4: "The issue with running a B20 on a PR4 or P75 ECU isn't its performance while cruising around in closed loop. The integra ECU will do a good job keeping the engine running optimally in that condition. But in open loop the map is just too conservative for the extra displacement of the B20 and the injector duration is too short causing a lean condition. It is 100% an ECU map issue though because the B18 and B20 injectors have the same flow capabilities. So to answer your question it isn't a "fuel hardware issue" but an "ECU mapping issue."

Even though most info points to the P75 as the correct ECU option for Civic/B20
setup if it's not chipped and tuned it will run lean.

Chipped P75 ECU description: http://www.ebay.com/itm/Chipped-P75-...38.m2548.l4275

Does an ECU have a backup (CMOS) battery? if so, what's it's life span when car battery is disconnected?

ezone

I've not NOTICED any, but I've only ever checked inside them for evidence of leaking capacitors and burned spots.

I doubt they have any backup because codes and adaptives can be erased by removing the power sources from the ECM.

Wankenstein

Okay. Thanks.

Contacted seller and he said he used P75 ECU for automatic transmission (this car was originally and automatic coverted to manaul) and removed the chip essentially making it a manual ECU.

Contacted a local tuner who has experience with this ECU/engine combination and he knew exactly what I was explaining to him. He said P75 ECU is meant for 1.8L and runs about 10% lean with a 2.0L. Options are to Hondata chip and tune it for best setup ($1000 ) or purchase an after market adjustable fuel pressure regulator and manually set it as close as possible.

Tuning it with chip and dyno only adds 10 hp and 5 torque of which I don't care about anyway..just want it to run properly. Still, though if it's stalling now, when is it was just hunting without stalling before..will an aftermarket adjustable FPR solve that issue too or is there still an underlying issue?

I was informed to remove the o2 sensor and block off tail pipe to perform an exhaust smoke test ..blow smoke through o2 sensor bunghole (hehe..lol) Also, as I mentioned before I intend to pull injectors and watch spray pattern, pulse pattern and volume. I asked the man at the tuning shop if it's okay to run lean and his opinion it's okay to do so and that, as we discussed, it's most likely not lean when throttle is opened more.

As far as my question about a CMOS in ECU..my thought process is/was: Since I had the car up on jacks for nearly 10 days with the battery disconnected during steering rack removal. Maybe, ECU lost it's mapping to an extent further than just doing an ECU relearn?

Wankenstein

Called another dyno shop, gave him the backstory and current stalling situation and I was aked if it's chipped and socketted. He told me to open the ECU and and send him a pic by email or drop by his shop and he will take a look. I told him that a 20 year Honda mechanic (you) is helping me narrow down possibilties and it's seemingly a lean issue. He said his Hondata chip, socketted and tuning (using a 5-wire wideband o2 sensor..doesn't stay on vehicle) is $300.

If further part(s) or fixes are needed to resolve stalling then will be extra cost. He also stated that an adjustable FPR is not the way to go and most likely won't resolve stalling or lean issue. He also said the 4-wire, narrow band o2 sensor currently on the car is not going to provide enough information to narrow down correct a/f ratios. He said he's been tuning Honda's and other cars for the past 10 years in the Tampa Bay area and should have it solved within a day. Shop's website: http://www.provenpowertampa.com/

Before I purchased this car I had a strong feeling it was going to end up at a tuning shop at some point.