SR20DET Loss of power/hesitation/misfire

[2tone_s13]

So I'm having an issue with my s14. I'm not at the end of my ideas yet but I am reaching a point where some input might help. If anyone has had a similiar issue and resolved it or any ideas as far as tests, tricks or checks that I can do, id love to hear them. I figured involving the community might help some one at some point too.

Black Top S14 SR20DET

Symptom - Intermittent loss of power, hesitation, misifre, bad idle (rough and pulsates down to 250rpm/comes back up)
This issue will very frequently come and go. It'll come and the engine will "pop" back into the way it should feel. More evident when accelerating. Very easy to feel under light throttle/load/slower speeds.
At higher speeds/loads it is a lot harder to feel.

Actions/diagnoses so far..

First - checked fuel pump. Was missing sock. Replaced with ISIS 255 ltr/hr pump. Came with new sock. All good.

Second - tried another known to be good ignitor chip (friend has an s14)

Third - cleaned MAF. no change

Fourth - i already knew of a vacuum leak in the brake booster. I finally got smart and thought maybe that is causing my performance issue so I pulled the vacuum line and plugged it. Didn't change the issue.

Fifth - checked spark plugs. all looked the same. old and gray. none were wet or dark. i was pretty confident this was normal wear and tear and was a good sign. nonetheless they were old so they were replaced with NGK's gapped to .028" One thing we noticed was #1 and 4 pistons looked abnormally dirty. I only looked at #1 myself and thought it might be significant. It kinda looked wet. Dare i say, oily. It's just kind of weird that the plugs weren't wet at all. For now, I'm just contributing this to the age of the engine.

Sixth - checked TPS signal. seemed good. around .4V closed and 4.5V WOT

Seventh - checked MAF signal. seemed ok. 1.5V at idle.

Eighth - tried to pull codes on ecu. couldn't get the light to blink. Might try more eventually.

My next planned action is to try another known good ecu, checked for clogged fuel filter or leaky injector o-rings, look hard for any vacuum/air leaks, clean iacv (not expecting that to fix it)

[Kingtal0n]

A few things you might want to consider

-Coilpacks might be in need of replacement
-Do a boost leak check by using an air compressor
-check the ground at the back of the head for the coil packs
-check/change your ECCS relay (do you have an alarm? Or turbo timer? It may interfere with the ECCS )
-Where is your igniter mounted? is it even mounted? Is it supported in such a way that it does not vibrate?

What else can cause an intermittent misfire? and funky pistons?
-I would perform a leakdown and compression test for peace of mind ASAP.
-Are you on stock turbo/injectors/ecu? Which ecu? NA/WC? Which harness?

I have seen wiring specialties harnesses give this issue, where pins at the ECU intermittently make contact or not depending on how they feel that day.

So far my best guesses are coil packs and the harness/grounding followed by compression/leakdown (with respect to the valves, perhaps a valve is sticking open briefly) did you ever look under the valve cover? is it all black and gooey in there?

good luck!

[2tone_s13]

Well I got to the shop on base today.. first thing I did was a compression test. Low numbers across the board. All were around 100 psi and #3 was at 110. I believe I may have found my issue.. haha either way I'd definitely want to rebuild this now. Or throw it in the bin... one or the other. This would explain the "oily" look to my pistons.

Keep in mind my lack of resources.

I tried to do a leak down test just to put my mind more at ease by making sure it was leaking past the rings. The shop didn't have a proper leak down tester so I tried to hook the compression tester up to shop air but the fittings were different. Go figure.

Did a "power balance test" by unplugging my coils one by one. Each one made a difference even while it was idling like crap. So I don't believe it to be a coil.

I saw a ground directly behind the 4th coil. It was plugged into a metal bracket on the firewall which was also bolted to the same bracket on the other end of the wire. I didn't look into that any further. I figured it looked untampered and thus correct.

This is when I called it quits. I might do a compression test on a good running sr just to make sure the gauge I was using isn't crap, but I'm pretty sure it was fine. I'll definitely start keeping an eye out for another motor or start considering a rebuild.

Has anyone else had low compression numbers yield these symptoms? And a rebuild alone resolved it?

Just to answer your questions Kingtal0n.. thank you for the response btw. I don't have an alarm or turbo timer. Yes, my igniter is mounted (lol) and it is bolted down securely. Everything is stock except for the s15 turbo. ^_^ I'm not sure what model ecu or harness it's using. I could check the numbers if it's really that pertinent, but I think the compression numbers speak the loudest. Under the valve cover looks fine.

[Kingtal0n]

Actually if your compression numbers are the same across the board you are probably fine.

You should check the back of the head for the single bolt locating the coilpack harness. That little clip on the firewall is not nearly as important.

Compression can show low if you crank slowly and forget to hold the throttle open. Also if you have large cams, or if your camshafts are installed a tooth off, they will also show low across the board like that.

I definitely think you need to re-do the compression test. this time, make sure all plugs are out first, and use a REALLY strong battery, and hold the throttle open. jumper cables the car if you have to. If it shows around 120+ you are in the clear at that point. Many Sr20det engines with high mileage have sticky/oily/carbon'd up valves that will show low compression as well. One such engine was showing like 70psi on one cylinder, and low on the all the rest. I thought it was a blown ring at first, but we ran some toyota carbon cleaner over the valves (with the head off) and cleaned it up really nice like new, put the head back on and it had great compression again.

As to the coil packs. it is very common for them to work sometimes, and not others. I think you should swap all four out just to be safe. it is fairly affordable.
When I think of your wet piston, it makes wonder about your injector O-rings next. Which can also lead to a misfire and screw with your wideband readings. For instance if the cylinder is too wet to fire because liquid fuel is entering, your wideband will show lean because of all the unburnt O2, despite it being too rich.

[2tone_s13]

Thanks for all the input man. These are all great things to consider and look into. I was already second guessing my compression test so I will be pushing a little further with that like you suggested. I will post again when I get a chance to tinker with it.


Sent from my iPhone using Tapatalk

[2tone_s13]

Just a quick update. I'm going to be working on the car again tomorrow but I managed to take a look at my coil ground. I'll take some pics tomorrow as well.

When I first found it, all 4 ground wires from the coils came together into one wire. (I don't know if this is correct) The wire then came together and went to the back of the head. Then the wire on the firewall was just looped to itself.. kind of stupid. What i did was move one end of the ground on the firewall to the head as well.

So now it's.. 4 Coil grounds --> one wire to the head --> from the head to the one wire going to the firewall (as well as the big engine ground cable.. haven't looked at that yet)

Like I said, I can take pics and I will be doing more diagnoses tomorrow, but I wanted to update with this before that.

This didn't fix the issue.

[2tone_s13]

FIXED. Well, 95% sure. You know how cars are.. I followed the FSM diagnostic procedure for "Hunting" symptom. First thing you check is the oxygen sensor and that fixed it. Still don't know if its the sensor or a wiring issue, but my money is on the sensor.

So just for the future if anyone is having this issue.. Rough idle, inconsistent idle, misfire, hesitation, loss of power. Disconnect your O2 sensor and if that makes it go away then its your o2 sensor circuit.

I found my s14 zenki sr20det FSM here: http://www.240edge.com/manuals/s14-sr20det.html

You can find also find ka, ca, other sr, and rb26 manuals there. YAY!!

[Kingtal0n]

O2 sensor does not function at wot. So "loss of power" is not a typical symptom of a bad O2 sensor. When the sensor goes bad, then your closed loop operation is compromised, the engine will start "learning" the wrong fuel % values for each fuel map cell and the result is either too rich, or too lean, when driving during closed loop. yours sounds like "too lean", which would cause a consistent lean misfire, during idle/part throttle driving.

it would not cause, for instance, a misfire during wide open throttle.

[2tone_s13]

My issue wasn't happening at wot. I know what the o2 sensors job is and how it contributes to fuel control. What baffled me was how it could cause symptoms as serious as mine. If the signal or whatever was messing up the fuel delivery so bad then idk why it wouldn't just fail safe into an open loop mode and run off of pre-stored fuel values.


Sent from my iPhone using Tapatalk

[Kingtal0n]

The computer has no way of knowing if the O2 sensor is bad, and it has a maximum adjustment for fuel % removed or added just in case something like this happens.
The ECU has no way to know the engine is running rich or lean. Your job during diagnosis is to unplug the O2 sensor to remove it from the equation as a possible source of problem, since the engine does not need it to run properly. You could do the same thing with the TPS the car will still run 98% normal without the TPS connected for instance.

You should probably reset your computer now, since it has learned all the wrong values from the bad O2 sensor, just give it a clean start. (unplug the battery and hit the key).
I am assuming you do NOT have a wideband or you would have caught this issue far sooner. Highly recommend you add one.

[2tone_s13]

The upstream o2 sensor's job is to tell the ecu how rich or lean the engine is running.. The maximum limit to adjusting fuel trim rich or lean shouldn't get to a point where my engine was running the way it was. So I'm just curious what exactly was going on.

I tried unplugging my TPS and my engine did NOT run 98% normal.

Yes, resetting my ecu is probably a good idea.

[Hoffman5982]

Been battling this issue for well over a year now. I just changed o2 sensors the other day and there has been no change, but I'll unplug it to rule out wiring. My next step was new coil packs, injectors, and knock sensor. At that point every sensor and fuel component will have been replaced. When it's not in its "mood", it pulls extremely hard and runs perfect, so I'm confident that my motor isn't blown.

[2tone_s13]

Must not be too bad if it's been a year haha my motor would die sometimes it was so bad. The idle would drop down to 250 and bounce back up. I recommend using the fsm to help diagnose your issue.


Sent from my iPhone using Tapatalk

[Kingtal0n]

Must not be too bad if it's been a year haha my motor would die sometimes it was so bad. The idle would drop down to 250 and bounce back up. I recommend using the fsm to help diagnose your issue.


Sent from my iPhone using Tapatalk

ohhh wait. what kind of wiring harness is in the car. did you wire it yourself? I am guessing not.

Also, how did it run with no tps? If you had to put a % on it. Isn't the ecu stock? A stock ecu should run the engine fine with no tps. I know a few engines that went for years without one, and the owners never noticed.

And last, yes, there is enough adjustment in the hands of the O2 sensor to make your engine run in the 17:1 range, if you had a wideband you could have seen it.

[Hoffman5982]

Must not be too bad if it's been a year haha my motor would die sometimes it was so bad. The idle would drop down to 250 and bounce back up. I recommend using the fsm to help diagnose your issue.


Sent from my iPhone using Tapatalk

Mine dies every once in a while as well, mostly when it's cold though

[Vdubbin32]

Im having similar symptoms, did you figure it out?