I've tried to jump blue two-wire service connector and yes, timing is constant 12 degrees at idle. But when it's shorted - idle is still rough and jerking/bucking when driving is still present.

My scantool is capable to show IAC and ECT temperatures, timing, both O2 sensors voltages, MAP, STFT and LTFT, engine load, revs, TPS and ECU voltage. Everything loooks normal to me, as much as I'm able to interpret that data.

Unfortunately I can't see raw data on misfires with this scantool (which is a simple bluetooth adapter+torque software on my phone). All cylinders do work, I can clearly see this by doing a power balance test by removing sparkplug wires one by one. Interesting thing here is, that when I remove sparkplug wires from cylinders 1,3 and 4 engine does not stall, just starts to vibrate much more, as it should. But when I remove sparkplug wire from cylinder 2 - engine stalls almost instantly. The same thing was happening on old engine too. I don't think that's normal.

I've tried known-good replacement PCM two days ago. Car acted the same.

 

An alternative possibility to look for ignition misfires is to use a timing light rigged on No. 2 (seems likely place to start). Hope would be to detect an interuption of flash (no spark) when bucking/jerking occurs. Here again, I think only way to rig is to securely tape the timing light (switch closed w/ nylon tie wrap) to windshield.

If spark is reason for missing, then you should be able to see an interuption in the spark sequence. May need a good helper to drive for you.

Did you check fuel pressure while driving under load? This seems more likely than spark since you've tried two distributors. Fuel inlet filter is in fuel tank (at least for 2.3L Accords for which I have manual).

good luck

 

When the problem first occurred, did it start out of nowhere? Did you do any work (major or minor) on your car right before the problems started?

 

As far as I remember - just drove home from work and noticed that some power is lost, and jerking/bucking effects are present. Opened the hood, and saw that engine on idle is vibrating, especially when AC is turned on, or any big electrical load. I mean, the more load on engine when it's in idle, the more it vibrates. But rpm's stay about 750, when engine is up to temp completely. Some work on this car was done a month ago - it was LPG converted, BRC sequent 24 system was installed (you can see that in the last photo). But car acts the same on both fuels, so chances are slim that fuel delivery is the problem.

Thank you for your time, I really appreciate it.

 

Is it missing at idle or just vibrating?

Vibration at idle is usually result of idle air control valve not providing enough bypass air to compensate for load. This can be caused by dirty IACV and/or dirty throttle plate/bore. There is a bypass bleed screw that is adjusted w/ IACV disabled to 550 +/- 50 rpm. Unsure if the 1.8L Accord is same.

Are you sure engine is idling at 750 rpm? The first markings have been a source of confusion in some cases, but first large mark above zero is 500 rpm on gauge.

Any chance that intake penetrations for LPG injection are somehow allowing additional air?

good luck

 

No misfire codes, all cylinders work, so I guess it's not real misfire, more likely vibration.

IACV looks good, like new. Cleaned it (it was pretty clean actually), no luck. Bought another from salvage yeard, cleaned it and installed - no luck either. With IACV disconnected idle becomes bouncy, car becomes pretty much undrivable then. Throttle bore is clean and shiny like new. Well, scantool shows ~750rpm at idle, rpm gauge shows that too.

No chance of air penetrations anywhere around intake manifold and hoses, injectors, etc; I've checked it with brake cleaner, then two shops checked for that when trying to diagnose the problem.

Did some voltage measurements at ECM/PCM harness today. Noticed one strange thing. TPS voltage at fully closed throttle, that enters PCM is 0.391V. At full opened throttle - 4.59V. Manual states, that there should be approx. 0.5V and 4.5V. Scantool at throttle fully closed reads 7.8% and at fully opened 93.3%. Manual states there should be approx. 10% and 90%. I guess it's time to try to change throttle body along with TPS. Or just TPS maybe? Or I'm wrong and these readings are completely normal?

 

Those TPS voltages look good to me. Although no tolerances are given for such measurements, a variation of +/- 10% is typical for voltage measurements for most electrical circuits. +/-5% is considered precision.

You might want to review some of the videos on youtube by Schrodingers Box. He advocates developing a theory, tests to verify/invalidate the theory.

However in your case, although the usual suspects seem to be ruled out. Someway you need to capture engine data when problem is occurring to determine source of problem. Missing is either fuel/air mix, ignition, or timing. You've ruled out compression pretty conclusively. Timing seems equally unlikely. That leaves fuel/air and ignition.

The Honda HDS system will capture extensive data from your engine. A "clone" version is avialable on ebay (search "V3.002.032 hds him diagnostic scanner") for $130. That may be an option.

Most of us have occasionally run into very intractible problems. Best wishes for a successful resolution. Please advise your final resolution.

good luck

 

0.391V is 21.8% off from 0.5V. I quess I'll give a it a try and swap throttle body along with TPS. And as far as I know timing has alot to do with TPS

 

Maybe this tells something?

 

I'm sure it is me and my computer but I can't seem to get past the ~37 second mark of the video