I have a manual 1996 lx accord with the f22b2. The car has a rough idle and seems low on power (not sure what it should be). The biggest issue however, is that it will never go above 4800 rpm. Irrelevant of gear, throttle use or speed. It hits 4800 rpm and loses all power till 3k rpm. No cel or stored codes. Timing belt is aligned correct. It also stutters at about 3k rpm.ive tried unplugging the o2 sensors and map, and issue persisted. Despite all this, it instantly starts on first crank pretty much. Any ideas? Thanks

 

turns out, it's running extremely rich on ALL cylinders to the point the sparkplugs turn black after a couple miles.

 

Just speculating, could the cat be plugged up…
that’s just a suggestion

 

Sorry, I forgot to mention this. I got the car as non running and the cat was clogged, so right now it is open header. The map, tps, ect and rpm are all reading properly on live readout so I don't believe it's any of those sensors either. A new cat is in the mail so I will be able to test soon with both o2 sensors plugged in.

 

Here's a video of the issue

 

Sparkplugs? Basic maintenance?

 

I've done spark plugs and the ignition coil. The airfilter is clean. All the spark plugs turned black within a few miles. The compression is fine, and the map, tps, ect and rpm are all reading correctly. Ive checked fuel pressure which was about 38 at idle. Since the rpm is reading and the car starts fine, am I safe to say the crank sensor is fine?

 

It could be a week crank sensor but the car does run, so unlikely.
running with no cat, I’d put one on, with a good o2 sensor. See what that does.
The other thing I’d check would be the idle circuit.
Honda uses a little piece of wax that expands and contracts to control the cold start idle.
maybe?

 

Is it odd that no matter what I do, and no matter how many times the engine cuts out, mis fires, or shuts off, that it has never thrown a cel or stored a code for anything other than an 02 sensor i unplugged. With it plugged in, it runs awfully but still never throws a code. Is there a chance the ecu is bad? I feel like I've confirmed almost everything is working as it should. Ive not seen a single sensor with a weird reading.

 

Anyone?

 

Make your own data logs. Drive around with a multimeter on various circuits, see if you can tell what loses function. Does the fuel pump circuit have issues? Maybe start there.

 

So the car finally threw a code (p0172) for running rich with long term fuel trim around -22%. If all the sensors are reading properly, airflow is not obstructed, and there is proper spark at idle, what causes the fuel injectors to have a pulse width of about 5 ms at idle? The short term fuel trim is almost always -27.3% at idle too. The car knows it's running rich af. From my understanding the pulse width at idle on a warm engine should be around 1.5-2 ms.

 

Do you have a full exhaust back on this thing or what?

 

Yeah. Both 02 sensors are plugged in as well. The only code its thrown is p0172 (too rich). The long term fuel trim is about -22% rn

 

So you have OEM style header, OEM style cat, a pre-cat o2, a secondary o2 inside or after the cat, and no exhaust leaks between the head and the second o2 sensor?

If the ECU is commanding that much fuel, then the MAP sensor might have it in the wrong area of the fuel maps. A faulty map sensor might not throw a code. You (the map sensor) should read 2-300mbar vac at idle. If you are above 320-> toward 1000mbar, def bad map. Map should read 1000mbar or close to with KOEO.

The only other sensors that affect fuel compensation are IAT and ECT. Invalid readings on any of those sensors will cause undue compensation. But prob not -22%. Sounds like a map sensor issue to me.

Do a basic sanity check on all sensors

 

I've checked tps, ect, vss, map. Map was about 5 psi at idle and 15 psi at wot. Ive also checked the vacuum at idle using guage on pcv port and got about 18 inches. Ive checked all the sensors i can think of, and tried unplugging them as well. None of the sensors are reading differently than what i expected. Is there a chance the ecu is literally bad or the injector resistor box is bad?

If the ecu knows its rich, wouldn't it shorten the pulses irrelevant of why it's rich? It shows its pulling fuel actively but still has long pulses

 

A sanity check means you double check the sensor is reading correctly. Your engine is not turbo, you shouldn't be seeing psi. That means your not sane. Sanity means you measure X amount of vacuum with a mechanical gauge and the map sensor reports the same value of X to the ECU.

For example. If the car hasn't ran in a day, and you go key on engine off, and the ect sensor says its -20deg outside but its actually 70deg out, that fails the sanity check. The sensor is not right.

You need to see 250-320 millibar of vacuum at idle. 1000mbar is atmospheric pressure. With key on engine off the map sensor should read atmosphere. Engine not running means no vacuum made. So it should read 1000mbar. If it doesnt, fails sanity check.

The engine might be making 320mbar mechanical vac, but if map is telling ecu its 1200mbar, its going to assume its under heavy load and dump fuel, regardless of what the feedback loop thinks

 

Map sensors in these cars quite literally read in psi, and yes, it is reading 5 psi. Hell, you can even convert the 5 psi to inches of mercury pretty easy. Thank you for calling me insane though.

 

Lol I thought that was good.

Well I didnt know if you understood what I meant. So 5psi to mbar is 340, thats a little high. Hondata, crome, neptune, etc, all those tuning softwares got me using mbar for units. I didn't feel like converting things for you and 5psi sounds like positive pressure not vacuum to me. I always thought working in mbar under 1000mbar aka 14.7psi just makes more sense, and anything over 1000mbar turns into psi, starting at 0psi @ 1000mbar. Thats why I visualize it that way

Well, maybe you could just get hondata/crome/hts/neptune and tune it? Lol

 

So at wot throttle I'm seeing about 14-15 ms injector pulses, but when it cuts out, the pulse changes wildly, and the ecu never cuts the power. Instead of dropping to 0v it drops to 8v and then back to 14. Is this the ecu just failing to send a proper signal? I tested the injector resistor box and got 6.2 ohms for them

 

Hmm. Well not sure if you know but the injectors are ground switched. Aka they receive constant 12v and the ecu grounds them out to open them.

Also no exp with resistor boxes. I thought by 96 all the injectors were saturated, not peak and hold. I didnt expect that car to have a resistor box.

I can't relate to the pulse width as I haven't really tested in that way before.

 

So I've confirmed im losing spark at high rpm when it cuts out. I've changed the cap and rotor and ignition coil. No cel. What could cause spark loss at high rpm?

 

Does the TPS test out ok? Is the TPS at around .45v closed and gradually increase in voltage smoothly as you open the throttle? Around 4.5v full throttle

 

So i haven't checked voltage but the live readout using obd scanner shows its working. 9% at idle, 95% at wot, and the in between correlates with my foot

 

Check wiring between distributor and ecu. Those wires are normally shielded. If they've been messed with maybe there's crosstalk at higher rpm's causing the rpm signal to be bad. Idk just more ideas