Here goes. I've got a 98 honda civic with the D16y7 (stock engine for this car). Just paid a boat load of money to get a CEL that came on moments after it left the shop. If you've followed my life at all, you know that I got fucked on the deal and refuse to let those clowns touch the car regardless if they created the issues with the CEL.

My old man sent his scan tool down for me to dig a little deeper. Aside from knowing it's P0133 (Bank 1 Sensor 1, slow response) I now have some tools to figure out what is going on. I already put a new o2 sensor in, knowing full well this code is rarely o2 related. It obviously didn't fix the problem, thus I continue to dig. Checked for vacuum leaks a few times and never really came up with anything.

Tonight I pulled the live data and could see that the computer is seeing a voltage and swing from both oxygen sensors. Everything else looked good- although if this requires additional info, I'd be happy to support with additional findings. What I did find was that at idle in neutral, closed loop, the computer sat at about 16% load. This pointed me to the TPS which is reading about 9% at idle, and only up to 64% when floored. Mind you, I've not done a damn thing to this engine in 120k aside from routine upkeep and replacing parts when they puke. The TPS shouldn't have a variable to show these signs, so I've got to assume it's toast. The car has felt underpowered and less responsive, which I equated to to the CEL all along.

Before I go blindly throwing money at this thing on a hunch that the TPS is fucked, anyone got any tips or tricks to try to better diagnose this thing? Looks like Honda builds it into their throttle bodies, so finding an after market sensor is almost null and I'd rather not go with used.

 

I would imagine there is some way to bench test the TPS resistance. Don't you get some smoking deal on auto parts where you work?

 

I play by the rules and try not to abuse my connections. I could potentially say "get me this and that" but I don't like owing people favors. Our primary electrical supplier doesn't re-pop the TPS for this car, but it's available from Dorman. Last time I asked the Dorman for some help, he agreed- then I never heard from the rep again.

I know how to do a resistance bench test on a TPS, but don't have an analogue resistance meter to watch for a steady increase (god, I've got to just buy one already). It was hitting 64% with every WOT, so it might not even jump around in testing but could still be bad. I'll have to pull the duct work and make sure I'm getting full throttle plate motion and check it against the scanner- just too fucking cold and dark out to do it now.

Looks like I've found the part also, so I won't need to get a new TB entirely.

Could a potentially faulty TPS create enough bullshit in the ECM to throw a o2 CEL?

 

If it has a cast exaust manifold pop the heat shield off and look for cracks around the o2 sensor and at the "merge" of all the cylinders. They crack a lot and can cause o2 sensor codes and cat codes. I have replaced a lot of them.

 

From what I can recall, this is not the cast piece like on other models. I've got the CX which has the cat as part of the manifold, so make your conclusions as you will. Haven't heard any exhaust leaks either.

 

I agree with checking for cracks. Usually runs right through the o2 bung. Plays hell on the switch rate

 

No cracks of any kind that I can see or hear. Still fighting with this thing Any advice is appreciated.

Pulled the carpet out from under the throttle pedal. The TPS is still a little "off", but it is showing throttle percentage to be a bit more accurate than what it was before.

 

Have you closely checked/tested the o2 wiring/connector pins. Are there any other codes or stored codes you may not see with the scan tool your using ? Have you done a reset and if so how long till the mil came back on. There is always the possibility of the new sensor being ng as well.

 

Just go back from intentionally throwing the CEL. I'd cleared it out, then data logged with the scan tool. As I ran around, I'd periodically pull over, shut the engine off and scan for pending codes. There was no cool down between scanning and logging again, so it stayed in closed loop. It wasn't until about the second or third shut down that I found a code pending, so I flipped the scanner into "record on MIL indication" knowing full well it the CEL would present itself in the next few miles. Sure enough, the MIL kicked on and the logger went live, however it recorded nothing out of the ordinary, and there was nothing unusual leading up to the MIL coming on. This isn't a graphing logger, it only takes a snap shot of the desired sensors. I monitored all of them and didn't see anything irregular aside from a single frame when the ECM went into open loop. This was well after the CEL had kicked on. With each frame, I was seeing a fairly dramatic voltage change in both o2 sensors, so they are reading- eliminating the potential for a broken or frayed wiring.

No other codes are present.

I won't discount the potential of the new sensor being bad, however I am putting the likelyhood of it being bad pretty low. Again, not reject that possibility, but the car threw the code immediately after coming out of the shop, and had no previous problems up until then. The OE sensor that was in their initially may have been good, and I put more weight on another issue potentially causing this problem.

 

A few more bits. Before going out to data log, I checked under the hood for broken or disconnected vac lines. I still haven't found any. I disconnected the vac line to the FPR and saw an idle change when creating the leak. It was a large enough change that I wouldn't think raising the fuel pressure was the only factor. Additionally, this tells me my FPR diaphragm is not torn.

Also poped the spark plugs off one at a time and did not a change in idle. No dead cyls.

 

Missing any engine ground straps ? Can you link your thread from what you had done at the "hack" shop.

 

I'll have to look into the ground strap, that's a good one. When I first bought the car, I had some extra cable laying around from a subwoofer install. I replaced all of the ground straps with it out of boredom, so it should be grounded better than the factory, and a disconnected strap should stick out like a sore thumb since it's bright red. While I was under the hood this weekend, I did note that the ground from the chassis to the head was still intact along with from the negative to the chassis. I don't recall if I had one on the transmission.

Here's the thread:

https://www.theturboforums.com/smf/index.php?topic=144147.0

If the weather is bearable this weekend, I am going to throw a new cap, rotor, plugs, wires and a fuel filter in since it's been a while. This should eliminate a fair amount of un-monitored variables. Given that the timing belt had 120k on it, I also intend to double check the ignition timing- a new belt may have thrown some additional advance in, and also I have no idea what it's set at base... so it really can't hurt.

 

I'll wait till you do the tune-up before going any further.

 

Ah, well don't let a cap and rotor scare you away. I did the fuel filter, C/R/P/W not too terribly long ago. Drivability before taking the car to the shop was perfect! Now, I'm down on power, feeling a little miss every now and then, and on occasions my idle will drop very low, but no stalling. Aside from it feeling underpowered (which might be in my head) the problems are very inconsistent. If it were a bad C/R/P/W I'd have these symptoms much more often.

 

Didn't want to send you on a goose chase, so I figured wait and see after you do the tune-up stuff then move on from there. Any shavings or metal dust around the dist shaft and dist housing ?

 

After thinking a few things through this has more then likely got to be related to the trans being out. Something got f-ed up. Pinched or disconnected wiring/vacum hose or something along those lines. I'd look real close between bellhousing and block area. Check closely and I'm betting you'll find something.

 

Good advice. It's not something I would have considered. I will hopefully be able to work on it tomorrow barring it's not too cold or windy out. About the only thing I can think that may oppose something getting pinched is that I'm seeing a o2 voltage at the ECM (read through the ObdII), so it's not completely broken. Pinched or snagged is going to be harder to trace, but I'll definitely look into it.

The honda Dizzy doesn't really look like a ford or GM at the base. The instead of having a separate retainer that keeps the distributor pressed into the block, there is a 3 bolt flange integrated into the distributor base that bolts to the head. I didn't notice any shavings before but will look closely when doing the cap/rotor. It is driven off the cam, so the timing belt may have had some effect- this is why I want to check base timing.

 

I mentioned the dist because I have seen them fail and you usually see a tell tale of metal dust or shavings when you pull the cap and look inside. It may not be the 02 wiring that is pinched. Could be something else. I just think it was not coincidence that your getting this code after a re-read of the repairs especially knowing the trans was out.

 

Yeah, I'd really like it to be non-coincidental but easily fixed. When I told the shop owner that the light came on 5 minutes after leaving, he was pretty defensive since I'd already put an o2 sensor in. If I figure out that it was on them, I won't hesitate to get the coals ready for dragging.

 

Working on it a little bit today. It is pretty brisk out, so I don't know how far I'll get.

Pulled the cap and it looks surprisingly good, same can be said for the rotor. I may re-gap the plugs and put them back in since they appear to have opened up a little, but all of this stuff looks like it is in VERY good shape. Of course I snapped on of the bolts holding the cap on, so I need to drill it out now, but otherwise there's no signs of abnormality or even wear for that matter.