Wankenstein

During drive to work o2 voltage didn't go below .02 until I let it idle in the parking lot for a few minutes. Voltage dropped and stayed at 0 for over 20 seconds yet engine seemed to idle smooth. I will post the video of it later tonight. With all the testing performed does this narrow it down to an ECU issue? I iwill recheck compression. I need to purchase an HEI spark tester to check the plugs.

ezone

I just looked at the 34 second video and can't tell anything from it.

IDK, let's see what the O2 and fuel trims are doing now.
All you changed was the MAP sensor?
No you don't. If your spark is strong enough to jump a 10mm air gap that's 30kV available from the coil (or at whatever point in the system you are testing from). Here's a couple quotes:

Quote:

"You can calculate the voltage needed for electricity to jump across a spark gap with a simple formula: voltage equals the air gap length in centimeters times 30,000."

Quote:

"But there are many factors such as the shape of the conductor at the arc, the air density, temperature, moisture, etc but I looked up a 6.25 cm sphere-gap at 25�C and 760mm and here are some of the values for different gaps (positive/negative impulse kV crest values given):

0.5 cm: 17.0/17.0
1.0 cm: 31.3/31.3
1.5 cm: 44.5/44.8
2.0 cm: 57.0/57.4
2.5 cm: 68.8/69.3
3.0 cm: 78.8/79.4
3.5 cm: 86.6/88.0
4.0 cm: 93.6/....
4.5 cm: 99.8/....
5.0 cm: 105.5/....

From A.I.E.E Standards June 1940 as tabled in "Gaseous Conductors", J.D. Cobine, 1958. These are close to the values given in IEEE Std 4-1995 "IEEE Standard Techniques for High-Voltage Testing"."

And if you have spark that can jump a one meter gap, that's gonna be about 3 million volts.

Wankenstein

I posted the wrong one ..sorry. Will post the one I made on my drive to work today.

Will be on vid.

From three weeks ago..several test (or so it seems)..You mean you don't want to re-read through 122 replies?..LOL. Here's a recap..check or performed:

*Cylinder drop test by removing spark plugs wires on at a time: all dropped same amount and can hear good saprk to plugs.
*Cylinder drop test by removing fuel injectors one at a time: 2-4 dropped the same, 1 (seemed) dropped but not quit as much as others (but very near)
*removed, tested and cleaned fuel injectors.. after installation repeated
drop test and all cylinders dropped the same.
*mechanical timing checked: good
*igniiton timing checked: good
*fuel pressure is correct according to Hondash and gauge on filter
*o2 sensor replaced
*smoke and air tested exhaust
*smoke tested intake
*sprayed TB fluid in intake at idle and at 2500 rpms
*FPR vacuum removed
*Hondash data-logging
* IAB adjusted

People on HT claim without HEI tester during a spark test the ignition coil could possibly be damaged. I trust your word over theirs.

Wankenstein

Here's the video of this morning. O volts at idle is mostly near the end of it.

https://www.youtube.com/watch?v=ptUg...ature=youtu.be

ezone

This can be true and it's always a possibility....but finding this out is one reason why you would do the test in the first place.

The original coil insulation (on many 6th gens for example because I haven't had to chase problems on a 5th gen in maybe a decade) can break down (get weak) over time and miles, and the spark checks can make a weak one turn to crap during the test. I've had it happen to me in the shop and it's kinda hard to tell a customer it needs a coil now too on top of whatever they brought it in for.

(What actually usually happens on the 6th gens is the spark eventually finds an easier path to ground internally, some leak spark through the side of the coil body to the metal shield up against it.)

Same could have happened when you lifted the plug wires and heard the spark snapping, and same during a compression test where the wires are just pulled out and left hanging. If yours was weak enough to fail during a maximum kV stress test, it would have already happened.

It doesn't even have to be a coil failure either....all it takes is the spark finding an easier path to ground.....like from the rotor to the distributor shaft.




tl;dr
The good ones will take it. The bad ones won't.

ezone

Video: still looks crazy lean most of the time

Wankenstein

Any suggestions?

When I pulled and tested the injectors I found it odd that the spray pattern/pressure (when pulsing with a 9v battery + forced air pressure pushing the brake cleaner) was weak and not fanned out at all. Maybe it's different when hooked up to harness voltage (12v ?) and pressure? Researching..they should be (obd1/obd2) 240cc saturated style injectors. Not sure what style/size a '99 CRV uses though.

I couldn't find a Honda related video. The injectors pattern and strength is similar though weaker (1/2 the pressure) to the injector on the left at 2:20 mark:

The MAP sensor I ordered and installed: Denso part# 079800-3280 is stock for '95 Civic and '94-'01 Integra LS head and intake manifold (my engine). 1999 CRV uses Denso MAF part# 37830-PAA-S00. As I understand it MAP readings help determine injector pulse time and rate.

This reads talks more about B20 displacement + ECU type + lean fuel delivery: http://www.carolinahondas.com/tech-t...lean-help.html
Another read that points to using bigger injectors post#12 or going to a B20 obd2 (p28) ECU with obd2 to obd1 harness connector.
Best solution is most likely a socket/chipped and tuned obd1 ECU.

Wankenstein

Here's a reply I recieved from HAMS today:
"Thanks for your email. Yes, your tuning shop is correct in that the P75 ECU doesn't provide quite enough fuel for the 2.0L engine. Chipping the P75 ECU and tuning with Crome would be a great solution if your tuner offers tuning with Crome. Hondata S300 is another, more advanced option to accomplish the same thing. However, if the engine is running slightly lean this will cause a higher than normal idle. A low idle is typically due to running overly rich or the ignition timing is retarded. Obviously there are many other possible causes, but these areas are in regards to the tuning/ECU."

ezone

Well....it's crazy lean and apparently it won't be resolved with the stock parts you have now, so it's now up to a tuner or who/what/however you want to mess with the ECMs.

Wankenstein

Really appreciate the help. At least it can be said everything else was tried first and not much spent in parts to this point.

I found a guy who sells p28 ecu socketed/chipped with Crome Pro and no wiring harness conversion needed for $150. He has a B20 with similar bolt-ons as mine and said he can base-map it off his setup. Told me it may or may not correct the lean issue without dyno tune. Good thing about Crome Pro is it comes with data-logging and additional tuning features.

Wankenstein

Just a thought: This engine doesn't have a knock sensor. If the P75 ECU is normally looking for a knock sensor and can't find one can it affect o2 sensor readings in any way?

I called the local Honda dealer and spoke with a lead mechanic who has performed several swaps that involved ECU chipping. He has reaffirmed that a stock P75 obd1 ecu is not designed to run a 2.0L properly and that without a tuned chip or dyno tune it will always run lean. He referred me to this site: http://www.phearable.net/information...ner-guide.html and said the owner has been tuning for a long time and sells great tuned chips and chipped ECU's. The chip/tuned obd1 ECU is tuned with a wideband o2 sensor and when installed the present o2 sensor will be disabled and MAP senor + new, chipped ECU will take over A/F ratios. So, dyno tuning may not be necessary at all.

I asked him about the fuel injector testing I performed and he said without at least 32 psi of pressure and a 12v source the injectors won't spray as they should when hooked up to the rail. I also told him my concern of possible pre-detonation damage and he chuckled and said I have nothing to worry about with this engine. He said he's ran setups on track racing that he pushed lean fuel limits and engines still survived. All this info puts my mind at ease for now. I will update once I get the new ECU in. Thanks again.

ezone

Don't you think it would code for a missing sensor?

I'm NOT crazy! LOL

Wankenstein

I'm not sure but I will ask when I talk to Phearable. I would assume it will unless they program it not to (if that is possible?). I can always repair a CEL as my father before me did: put a piece of electrical tape over the the CEL light..lol.

Confirmed
It's probably been running lean since it was swapped. Hopefully, the new modded ECU will solve this lean issue and maybe see some hp and tq gains.

Wankenstein

My questions to sales:
*Would optioning for a disabled o2 sensor be a solution to my engine's fuel trims/low o2 voltage issue?
*Would disabling the o2 sensor cause a CEL?
*Can disabling or enabling the o2 sensor be reprogrammed at any time by *Crome Pro/tuner?
*Does a B20 basemap tune account for the fuel injection differences compared to a 1.8L?

HAMS answers:
Disabling the O2 sensor will prevent the ECU from providing a trouble code for the O2 sensor (if faulty) as well as cause the ECU to ignore 02 sensor signals. This is a great option if the 02 sensor is faulty or you'd like to force the ECU to run off of the chip's fuel tables rather than make changes based on 02 feedback.

The 02 sensor can be disabled or enabled by your tuner using the Crome software.
Yes, the B20 basemap adds the necessary fueling to compensate for the additional displacement.

Phearable answers:
Its fine to run in open loop, in fact 90% of our customers run in open loop all of the time as their factory o2 sensor doesn't function properly. In theory the stock ecu & stock o2 sensor should reach stoichiometric with the stock p75 program and a stock o2 sensor.

On my drive to work tonight I noticed o2 sensor at 0 volts on deceleration and idle so it seems that is doesn't go into closed loop mode.

Wankenstein

Pic of unused connector ('94-'01 Integra LS harness).