ArchEnemyEJ1

Okay so about 3 months ago I started having issues with extremely rough cold starts, no warm up cycle, and after achieving normal operating temperature the car would stall sporadically after letting off throttle to coast to a stop or park or whatever. After stalling it would take a few seconds of cranking and WOT to get it started again and no CEL. It was not a sudden onset thing, it gradually got worse, and now when the vehicle is hot it almost always stalls after letting off the throttle. So what I am working with: EJ1, B20/b18c1 head, p28 ecu with crome tune. What I have done so far: Cleaned and tested original IACV, after no improvement, replaced IACV with functional used one, still no improvement but will go through warm up cycle. After replacing the IACV I attempted to reset the idle, and the car will not idle with IACV unplugged for me to set the idle, even with the idle screw completely removed. They only success I have had is removing the vacuum line to the IM, and I tried setting the idle that way and it did not work. For giggles I replaced, plugs, wires, cap, rotor, coil, and ignitor (control module), timing at warm up and cooling fan temperature is spot on. Replaced, Coolant temp sensor, map sensor, o2 sensor was replaced about a year ago so I have not touched that yet. Also had a valve adjustment and compression test done about a month ago and #1 showed 120psi and 2-4 tested at ~128, no leakage. Tested fuel pressure, 42 at idle with vacuum line in place, and 50psi unplugged. Pinched return line and pressure rested at 72psi, after key off, fuel pressure held at 42, so I think thats all good.

I tested the intake air temp sensor (air charge) and the input (reference) voltage is only reading 3.4v which I thought was supposed to be right around 5v. One of the videos I found on youtube said the ECU could be at fault for improper reference voltage so I reset the ECU to be sure and it still only read 3.4v. The IAT resistance was ~1970 ohms, and have not been able to find what a factory resistance for the IAT is. Also tested the TPS and at closed throttle it read .47v and WOT read 3.6v which I thought was also supposed to read ~5v, and no sporadic jumping between closed throttle and WOT.

In my experience, I have never seen an bad o2 sensor have a major affect like idling and stalling and cold start. But I will test the resistance on that turd as soon as I can track down the acceptable ranges.

So my questions are, first and foremost, WTF? Secondly does (or could) the TPS voltage reading affect the IAT reference voltage? I do not think that it would, but I have to ask to be sure because I am running out of ideas. Since this is an issue that gradually got worse to the point where it is at now, could it really be the ECU? Should I plan on replacing the TPS and even if so, the reference voltage for the IAT not being ~5v is concerning.

Update***

Picked up a TPS, and now I can not get the new one or old one to read over 2v now, not to mention I cant even set the closed throttle at .5v because I have to twist the TPS past the bolt holes in order to get there, and that won't allow the throttle to open all the way.

Also spent all day checking the engine harness and there was nothing that looked bad, but in another attempt to set idle, I noticed that turning on the defroster and blower on high drops the RPM to around 150-200. I read on another forum that Canadian model Civics (which mine is) do not use an ELD, so could the alternator be the issue? I have to ask because I feel like I am running out of options. All the reference voltage at the ECU are normal too.

***copied and pasted this from my post at AZHT since it seems to be a little slower than HT. Thanks all.

eghatch9295

iTrader: (1)

o2 sensor readings do nothing in open loop, so would not mess with startup

ArchEnemyEJ1

Copy, but the issue is not only rough starting, but also dying when throttling off after reaching driving temp.

eghatch9295

iTrader: (1)

any cel on? do you have a known good ecu you can test with? if not, crack it open and look for burn marks, leaking capacitors, etc.

ArchEnemyEJ1

I'll crack the cover open tomorrow and check it out.

Former User

Did you or someone else mess with the throttle stop screw?

The low IAT reference voltage suggests an ECU problem, as does the low WOT TPS voltage, which is about 1V lower than it should be. What are the reference voltages for the TPS, MAP, and ECT?

ArchEnemyEJ1

Great question Ron, but no the throttle stop screw has not been adjusted. MAP and TPS reference voltage are at 4.97. The IAT and ECT are only showing .8v but both show 12v when checking for continuity, I checked a couple other 2 wire sensors just for the sake of curiosity and the VTEC oil pressure sensor was only showing .4v and the coolant temp sensor for the gauge (on the tstat housing) was showing 8.8v.

I also opened the the ECU and it looks like new.

Former User

This suggests the TPS is bad, assuming you did the TPS output voltage measurements correctly. You should double check the readings.

On what wires? Post pictures or colors.

I'm confused. Continuity is an Ohm value.

This is the fan switch. The voltage should be closer to 11V or higher. What's the voltage reading across the two battery posts?

ArchEnemyEJ1

Right, that is why I purchased a new one, and it is not any better. That is about the 10th time I have checked reference voltages in the last 2 days.



Will get to that shortly.



Yep you are correct, meant to say ground.



12.2v

Former User

You may want to describe in detail how you are making the TPS and fan switch connector voltage measurements.

ArchEnemyEJ1

Voltage dropped. Red/Yellow Wire. Tested at the ECU and it is the same.

Former User

In the picture above, is your multimeter set to test a 12V battery? What is that setting?

You should be dialed to the 2V (2000 mV) or 20V DC voltage range, depending on the voltage tested.

Also, what are you using as ground for each voltage test?

ArchEnemyEJ1

Sorry, I thought I did. For the TPS I set my multimeter to 12v, connected the black probe to the negative battery terminal. Used the red probe to check the reference voltage on the yellow wire. It currently reads 4.9v. I then checked the original TPS at closed throttle by probing the middle (signal/red) wire and the voltage read .47v. I then opened the throttle to WOT and the voltage read 3.6v but have no sporadic voltage jumping from closed throttle to WOT and vice versa. I purchased a new TPS and retested the same way, and can only get the closed throttle to .18v and WOT reads right at 2v.

For the fan switch, the black probe is connected to the negative battery terminal, and probing the green wire with the red probe.

Former User

I'll first give you time to digest post #12.

ArchEnemyEJ1

Negative battery cable for ground.

Per the users manual It is for checking 12v systems/batteries under load, or checking voltage readings on 1.5v, 9v, and 12v circuits.

Former User

Also try the sensor grounds for comparison.

Seems unusual to use the battery testing setting when the appropriate DC voltage setting is available. Try an appropriate DC voltage range setting to see if any readings change.

ArchEnemyEJ1

Well ****, setting it to 20v everything is a little better now.... fan switch reads 12v, TPS can be set to .5 at closed throttle, and WOT is 4.5v. IAT still only reads 3.6v, ECT shows 3.2v.

ArchEnemyEJ1

Too many youtube videos show to keep it at the 12v setting, so I followed suit.

Former User

Yeah, don't use the battery testing setting to take voltage measurements.

Focus now on the ECT and IAT reading. Did a previous post already describe how you are making those measurements? Are the connectors unplugged? What are the wire colors in the sensor connectors?

ArchEnemyEJ1

Well Ron, I decided to take a break from electronics for a minute. and turns out the shop that did the compression test, didn't, or just did not want the hassle of rebuilding the bottom end and just told me it was good, I do not know. As stated originally, they told me compression was a little low in #1 but was acceptable. I got some loaner tools to test the compression and see if maybe I had a blown head gasket between cylinders or some crap like that because I had no overheating. So I tested the coolant for hydrocarbons and it passed. But #1 is only holding at 60 psi which is crap. added a little bit of oil to the cylinder and it shot up to 130 where it should be, so looks like this is a compression issue after all.

Looks like I was chasing my tail with the electrical crap because of some bad interweb tutorial and probably never had any electrical issues to begin with. Sorry to waste your time. At least I know the proper multimeter setting now, and that its time for a rebuild. Thanks Ron!

Former User

Are you sure injector #1 is not stuck open, flooding the cylinder and lowering compression?

What were the psi readings for the other cylinders?

How did you do the compression test?

ArchEnemyEJ1

125-128

warmed up the engine. removed plug from cylinder number one (lots of oil considering the plugs are only a couple months old), attached compression tester. cranked engine for about 5 seconds. took reading. removed tester, replaced spark plug, then repeat for 2-4.

how would you know if the injector was stuck open?

Former User

Constantly sprays fuel, rather than timed pulses.

Compression tests are done with warmed engine at WOT with fuel injectors disabled.

ArchEnemyEJ1

You really think an injector would leak 70 psi? Honest question.

Former User

I don't make suggestions that I feel are highly unlikely. Honest answer.