Does this engine have an Intake Air Bypass Control Valve?

 

Yes it does

 

Does it make a difference if it's disconnected? Just throwing that out.

 

Nope, nothing changes

 

I noticed you replaced the distributor cap and wires, then the entire distributor. Was the distributor cap changed with the distributor change? That is, did you get an entire distributor with the cap to swap out?

 

Ignition system is o.k. for sure. First I've changed one by one: plugs, wires, cap, rotor, coil, ICM. That didn't cure the problem. Then just tried whole another distributor with cap and everything from salvage yard to same result. Spark is good and strong, but engine dies when sparkplug at cylinder #2 is removed at idle. But if I give it some gas, like, I remove it holding engine at 2500rpm, ant then remove sparkplug at #2 and let it idle - it does not die, just vibrates as cylinder #2 is not contributing at the moment. On other three cylinders - #1, #3, #4 I can remove sparplugs at idle and engine does not die. It's clearly seen in teh video. Maybe it's pretty normal, I'm not sure. But if I do injector balance test and disconect injectors one-by-one - engine does not die when I remove cyl. #2 injector connector. Test goes as it shoud, with every injector disconnected engine starts to shake as particular cylinder is not contributing. Compression is about 220PSI on all cylinders.

 

Aside from jerking/bucking on acceleration and a bit rough idle, thing that annoys me most of all is, that when I start up completely cold engine after sitting over nigh, it fires up ok, idle is a bit higher as it shoud be, say 1500rpm, and after a hundred meters of driving, when I switch it to neutral, idle drops to 400rpm or lower, but never stalls. This happens until engine gets fully up to temp, for a few kilometers or so. Then idle gets normal and does not drop lower than 750rpm. There is no FITV, IACV only. IACV is clean, throttle body also. Another IACV was tried, to no improvement. I suppose there migh be some vacuum leak, that kinda seals with higher engine temperature? Or there are other posibilities? I guess all my problems are interconnected

 

Thinking back on the old days with mechanical and vacuum advance distributors, what you describe when you step on the gas is a problem. Timing should advance with both engine speed and with increased throttle opening. But it's not. At least when you experience the bucking, it's the opposite: severely retarded timing.

Going back to holmesnmanny post with the missing knock sensor wiring shield, the KS retards the timing via wiring to the ECU. So if the KS is taken out of the circuit then the only way the timing can be retarded by the knock sensor circuit is either by a bad ECU or bad wiring. You replaced the ECU with no effect. That leaves the wiring. Looking at the 6th Gen FSM (F23 engine), it looks like the KS sends a ground signal to the ECU when it senses engine knock. It must be a high impedance circuit since the line is shielded from RF interference. If there's a break in the shield and interference on the KS circuit retarding the timing, it may not be noticeable under low throttle/low timing advance conditions but very noticeable when the engine needs a lot of timing advance. No KS on 5th Gen cars, but as I recall the ignition sensors all share a common shield, and that shield is grounded at a single point on the engine. I wonder if the KS shield lost continuity to ground? What if you were to find and ground the shield close to the ECU plugs to see if it made a difference?

 

"Timing is advanced with increasing RPM and retarded with increasing manifold pressure (decreasing vacuum). When you ease the RPM up, timing will advance; when you snap it up, the first thing that happens is that it retards as the manifold pressure increases then advances as the RPM becomes the dominant condition." Exactly this is happening.

Timing advances with rpm's going up, but retards when I hit the gas pedal and manifold vacuum drops. So I guess that's normal, and unrelated to bucking. TexasHonda said the same thing before: "Timing should retard during acceleration. I don't know numbers and there is no map in the shop manual, but those numbers don't sound unreasonable."

And one more thing - as I said before, disconnecting knock sensor makes no difference in bucking, just lights up check engine light. Sensor has only one (signal) wire, engine block serves as a ground. All grounds that go from engine block to the ECU are shiny clean.

"KS sends a ground signal to the ECU when it senses engine knock." That's not true as far as I know. Knock sensor is a sort of a special microphone, which sends some acoustic signal thru shielded signal wire when knocking occurs, and ECU retards timing then. When no knock is detected, KS does not affect timing in any way. So when it's disconnected it should not affect timing

 

I have never seen timing retard when either goosing the throttle and/or rev'ing the engine. The old vacuum advance units used ported (not manifold) vacuum so when you hit the gas the vacuum increased, advancing the timing to better burn the suddenly richer mixture, with the mechanical advance taking over as the RPM increased. Retarding the timing 16 degrees (25 - 9) when sharply opening the throttle is, IMHO, exactly the opposite of what you'd normally see, and would cause severe drivability problems such as bucking/major hesitation.

I'm speculating that the problem is in the KS wiring - as HomesMany suggested - vs. a problem with the KS itself. The KS IIRC is just a piezo voltage generator. Looking at the 6th Gen FSM it seems as if the output is inverted, but maybe not. You may well be right that a positive voltage is input to the ECU. In any case it's a high impedance circuit and is easily upset by stray RF; RF shielding is crucial.

But of course I could be completely wrong.