JetsterDajet

Sorry for this late-breaking but critical news, guys. I'm still learning about this car and Hondas in general. I assumed the exchanged engine was the same as the one that was pulled out of the LX but it seems that's not the case.

ezone

Awwwwshit.

How was this swap done:
Block and head dropped in using the existing harness and exhaust?
Or was the harness changed with the new engine?
And was the exhaust changed with the new engine?

And how does the cruise problem figure into this (possible clue), what did you unplug to get VSS to work again?



You need to figure out which engine harness was used and maaaaybe which computer was used.

Which manifold and cat was used? Is the cat out in front of the engine built into the manifold, or is the cat further back under the car?

Engine harness:
Where does the front O2 sensor connect to it?
Where does the rear O2 sensor connect to it?

Read the engine code on the front of the block.

AFAIK a D17 won't have an oil pressure switch on the VTEC solenoid, but a D17A2 will, but that's a swappable part.

Just wait on that. Figure out what you have first.

JetsterDajet

Alright, I will try to identify the engine block, exhaust and harness and get back to you.

JetsterDajet

The engine block is stamped "D17A", so we're dealing with a JDM VTEC.

I'm not sure how to tell if the harness is stock or pulled in, the exhaust either. The cat is right on the front of the engine, not under. The primary O2 sensor connects right above it and is easy to get to. The secondary is connected under it on the right side -- I would have to get under the car to remove it.

From the speed sensor thread, we determined that when terminal "O" is disconnected from the fuse box, the voltage is no longer pulled down on the speed sensor and it works again:

The ECM as far as I can tell is the stock LX one -- my understanding is that you need to cut some plastic to remove these and everything looks intact under the passenger side dash. The grommet in the firewall that the engine harness passes through looks quite loose and pulled about -- seems they may have actually pulled the old harness and fished a new one through.

Here's a picture of the whole engine compartment and closeup of my O2 connectors and cat:

ezone

Looks to me like the cat is part of the manifold so I'm guessing that's the LX exhaust, and both of the O2 sensor harness connectors are right there so it's probably the LX harness, ---even though they should be anchored to a bracket that's probably missing now.


To pull out the PCM one would need to cut out the plastic strip behind the glovebox in order to unbolt the metal brace immediately behind it




Did you ever unplug the OTHER sensor (the bottom sensor) and make sure that causes code P0141?
Just to make sure nobody got the sensors screwed into the wrong holes.



Another check might be wire colors and correct positions in each harness connector.... Even though 3 wires in each harness connector are the exact same colors, they may occupy different positions in the cavities.

3 common colors
black/yellow (heater, 12v+, fuse 4)
black/white (heater ground controlled by the PCM)
Green/yellow (sensor ground)

and the odd wire in each
Primary sensor = white
Secondary sensor = white/red


Kinda makes me wonder if someone happened to yank all the wires out of a connector, then stuck them back in the wrong holes..... Swapping sensor ground with heater ground might cause a hair pulling problem like what you have here LOL

JetsterDajet

Disconnecting the secondary O2 sensor is throwing a P0141, seems to be right. I checked the connectors and the positions of the colored wires match, with the primary having the WHT and the secondary having the WHT/RED.

I've seen some are able to pull the ECM without cutting the plastic by brute force, so no cut plastic doesn't seem to mean, for sure, that the ECM hasn't been swapped. Looks like my next task is to yank the ECM and verify if it's stock or not?

JetsterDajet

Can't yank it right now, I don't have a drive or extension long enough to reach the nuts in the back. I'll have to make a run to Harbor Freight tomorrow. Let me know if you think of anything else to check and as always thanks so much for your help.

ezone

Ok then check the two ground wires in the primary sensor connector, did they get swapped?

black/yellow and black/white should match up with the pair of like-colored wires on the sensor plug (most sensors use like-colored wires for the heaters)



I'm just making wild guesses here, trying to come up with something that's entirely possible and avoid ECM replacement.

Disconnect the ECM and sensor, and make sure the heater ground wire is not somehow grounded to the chassis?

If it had an ECM for a D17A2 engine, it would have been setting VTEC codes IMO, so I'd expect it to be original or at least correct for the stock LX non-VTEC engine..

The VTEC solenoid and valve on your engine, does it have an oil pressure switch sticking out of it?
Are there ANY wires connected to the VTEC solenoid?

You might be able to read ECM numbers with a mirror and flashlight (never tried it myself)



ECM = PCM to me LOL.

JetsterDajet

Pulled the ECM. It's a 37820-PLM-L04 and appears to be stock and never removed. Also pulled the PCB out of its case just to check for any obvious shorts/burns/loose solider joints. Looks fine.

So, someone just dropped the engine in and left the stock exhaust, engine harness and ECM.

The BLK/YEL and BLK/WHT wires on the harness match up with the like colored black wires on the sensor connector.

With the sensor and ECM disconnected this is no continuity to body ground on the heater ground wire.

Drove the car up and down the street, just for kicks. Same code. Even cleared out the codes while driving -- computer kicks it back in after about five seconds.

And I thought this was going to be easier than the speed sensor...

JetsterDajet

Are we 100% I have the right part? I found this review for a Denso 234-9005 (the more expensive A/F sensor) which seems to be a similar situation to what I'm in:

My daughter's 2004 Civic LX with a D17A1 engine was showing a P0135 OBD II code. That meant the upstream oxygen sensor had a bad heater circuit. I purchased an NGK 24425 unit that a local parts store listed as a compatible replacement part. It did have the proper connector and wire length but the sensor design was slightly different. Once installed, the NGK unit threw a P0134 OBD II code which indicates no activity detected from the unit.

A bit of research on the internet revealed that the '04 Civics use something called a wide band oxygen sensor (also called an air-fuel ratio sensor)- something not mentioned in my manual or on the parts house web pages. At any rate, I eventually identified this Denso 234-9005 as the most likely candidate for the correct part by comparing photos to the actual unit. All the positive reviews here reinforced its likelihood. Going to Denso's webpage also identified it as the correct part. Due to time constraints I had to pay a few dollars more and buy one locally but it did the trick. After about 500 miles there have been no OBD II error codes. This IS the correct part for a 2004 Civic LX with a D17A1 engine. Don't waste your time trying to save a few dollars with a cheaper unit that will probably turn out to be incompatible with the car.

These Amazon reviews were extremely helpful in my attempts to get the correct part. Thumbs up to all the reviewers and to Amazon for maintaining this exchange of information and experiences!


While I still have the Denso 234-4074 installed... Note that while he has a D17A1 and I don't... that's what my car originally had and seems the original harness and ECM are in, so if it gets that it should be happy, yes??

ezone

Funny....I ran the part# in my catalog and it comes back as for 2002 not 03.
I don't think that makes any difference though as far as running the engine, but it may be a clue.

EDIT: Part number was for the PCM

ezone

Look up the sensor part number on the website of the sensor manufacturer.

http://densoautoparts.com/find-my-pa...icle-selection

https://www.ngk.com/Automotive-Oxyge...-NGK-c904.aspx

JetsterDajet

Yeah, looks like everything I'm reading says the D17A1 (exhaust and ECM) wants 234-4074, which I have in there. Wonder why that dude has success.

Guess I have to bite the bullet and get a replacement ECM. Any idea what the dealer's going to want to flash a new ECM to my key?

JetsterDajet

Seems I may be able to test a foreign ECM by obtaining an immobilizer brake code from my dealer? This would be a good way to test an ECU without committing to the reflash. Do they charge for that info I wonder? With the engine running with the immobilizer code will the ECM still properly check OBD II codes and properly throw/not throw them?

JetsterDajet

I haven't performed these tests yet -- will get to them next just so I can say I did everything.

ezone

Prices will vary, I'm sure. We (my dealer) used to charge about half of what other nearby dealers charged, now we charge about the same.


I might be able to help get that code.
You may need to know the VIN of the car you got the ECM from as well as your current VIN.

AFAIK yes.

JetsterDajet

Alright, time to get back to basics. I fabricated a test light that lights up quite nicely across the battery terminals. It does not, however, light up across the BLK/WHT and BLK/YEL heater wires.

Again, in ACC mode or engine ON, I read battery voltage on the BLK/YEL wire and continuity to ground on the BLK/WHT wire. Just for kicks, I checked for continuity to ground on the BLK/YEL wire and... it has continuity. I'm no electrician, but this doesn't seem right. There should be no ground in my voltage wire... right? I didn't even realize I could read voltage on something if it's grounded.

ezone

Blk/yel comes from fuse #4 in the dash fuse box and is 'hot' only in ON and START key positions.
Should NOT be hot in accy position.

Blk/wht is a controlled ground that comes from the ECM and should only be grounded when the ECM turns on the heater. (AFAIK that's only when the engine is running)



If you check the positive wire when it's not powered up, it can show as grounded due to other loads in that circuit with less resistance to ground than your test light bulb. Don't worry about that, just make sure it's getting powered up when it's supposed to.


For a real simple test light I sometimes tell people to use a #194 license bulb, unfold the little wires and they will usually stick right into two female terminals of a socket and make contact nicely. Works good on a power door lock actuator connector.

JetsterDajet

With the test light connected across BLK/WHT and BLK/YEL, it never comes on, engine on, off or in between -- I don't think there's any cut in the wire, seems the ECM isn't turning on ground enough to turn on the heater. If I leave one side of the test light connected to BLK/YEL and connect the other side to body ground -- light comes on. I figured I'd hard wire the sensor's heater ground to body ground. Sure, the heather would be on all the time, but it should get it on and the code off, right? Then I can get it smogged and who cares after that. I spliced a wire into it and it doesn't care, P0135 still throws.

ezone

Stupid thought: Did you erase the code before you did the test?
The ECM might not even try to operate the heater if it's already stored a code for it.

Erase codes, retry this test.
IF it still can't supply the ground, then you have proven the ECM is at fault.
Sounds like the hot side is OK.

Nope, won't work that way. It will code no matter what, and as long as it codes you can't pass the test. Right?

The ECM has to control the sensor ground. You can't make it happy by wiring around it.

JetsterDajet

Damn bastard ECM won't take a hot heater even when I make it hot, eh...

I think the code was still there when I checked across the wires with a light. Let me give it another shot -- if it still doesn't light then it sounds like the ECM won't engage ground and it MUST be jacked up -- yes?

ezone

Next step: Continuity test to the ECM, then a Terminal tension test (pin drag test), at the plug on the ECM, on the female terminal of that wire in question.

B1S1 heater black/white goes to terminal A1 at the ECM.


If you ensure the connection to the ECM is good then the ECM must be at fault.

I keep saying check the terminal tension......this is an unusual failure IMO....and more than once I've had my *** handed to me by a broken wiring terminal. I'd hate for you to bite the bullet and stuff an ECM in only to find out the problem is still there, you know?

Dale Irish

I realize this is several months old now, but was this ever rectified? I am batting the exact same issue with codes P0131 p0135, and now p0141. I believe I had a bad ground wire between the Ecm and #1- O2 sensor connector (blk/wht wire). I spliced in a temporary wire from the ECM plug the first O2 sensor plug. That seems to have cleared the p0131 and P0135, however that somehow triggered p0141. So now I have to go and check power and ground to the lower oxygen sensor.

just wondering if this was solved