Well, knock sensor is not damaged physically. Wiring is originally shielded. If wiring was broken - CEL would be set. Tried to disconect it, and absolutely nothing changed, just ECU threw a code for knock sensor. Timing, set by ECU at idle is constant - 12 degrees BTC, as it should be. As I understand - knock sensor just retards timing, if it detects engine knocking. Nothing like that happens, so I'm pretty damn sure there is no problem with knock sensor.

Now the MAP sensor. First of all, I took a service manual and a multimeter, and measured it's voltages. With ignition ON, but engine off, MAP gives 3V output, and that's what manual states as normal. At idle it gives 1V at the output, just like the manual says. Another thing - scantool at idle shows about 25-27kPA of vacuum in the intake, that's the MAP's voltage converted from volts to kPA by ECU. As far as I know, that's a good number too. Also, I have another known good MAP sensor, and just for the heck of it, I've swapped them, and again - nothing changed. So MAP sensor problem is highly unlikely too.

Now the TPS. At idle 0.5V, at full throttle 4.5V of output. Voltage transition is smooth. No dead spots I did find. And also, I can't imagine how TPS could cause rough idle. Jerking/hesitation - maybe, but not the rough idle. So TPS is not likely a problem source too.

Spark plugs are new, gapped properly.

So i'm really stuck with this s**t six months already. And I guess, that some sort of mechanical problem is here, and compression test does not show that. If I was sure that's mechanical-compression related - then new long block and problem solved. But i'm not, and I don't want to throw a bunch of money into it, beeing not sure that it will cure it

 

New long block has arrived. Has 80.000 miles on it. Two weeks later it will be installed into my car and I hope that I'll forged about 0.5l of oil every 1000km, jerking/hesitation and rough idle.

 

 

Well, new long block actually solved pretty much nothing. Now compression is 220PSI across the board, and this one consumes much less oil. And that's that. But idle is still rough as it was before. Car jerks/bucks very harsh on acceleration when engine is cold. After a mile when engine gets warmer, jerking/bucking on acceleration becomes less noticeable, buts it's there all the time. When in neutral, idle gets low, almost stalls, then slowly returns up to spec. After few more miles idle becomes normal, doesn't get low and is just rough. Tried a few more things - known good ECT sensor, IACV and PCM/ECM. To no improvement at all. There is no air in coolant, and no CEL. Maybe somebody has any ideas on what to check next?

 

Never saw mention of back pressure from exhaust being a concern.

 

What parts from previous engine were used on this replacement? intake manifold, throttle body, injectors, distributor?

Since problem is unchanged, you have to suspect the problem lies in the retained parts.

Has fuel pressure been checked? I didn't see this mentioned in thread but may have missed. Low fuel pressure or fuel pressure that drops when fuel demand rate increases steeply (acceleration) might cause a lean stumble.

Your determination to repair this car is admirable. It is unfortunate that fate sometimes assigns a difficult task to be resolved. Please perservere.

good luck

 

I'm not sure if removal of O2 sensor before catalyst is a reliable way to test is there an exessive backpressure in the exaust, but I did that and car acted the same. Catalyst is doing it's job - O2 sensor voltage past catalyst reads what it's supposed to when driving, and there is no CEL, no smell of rotten eggs, no smoke. Looking and listening to the exaust sound at talepipe I can definitly say that some sort of missfire happens frequently.
Intake manifold with all sensors and throttle body was used from the old engine, yes. I've tried another injectors set, to no effect. Distributor is from the old engine too, but I have another one from the salvage yard, to same results. Fuel pressure is up to spec. FPR is not ruptured, no smell of fuel at it's hose.
This car already costed me a fortune to repair and lots of my personal time, but on a bright side - I still like it, and after chasing this problem for more than a half of a year, I know about cars way more than I used to. But anyway, I really would like to get it fixed, because driving it like that will get my engine mounts torn sooner or later I guess. And probably some sort of mechanical damage, like clutch and etc. BTW, MPG has not suffered much or at all, when this started.

 

Fuel pressure may need to be checked while driving and under load to see if fuel pump is keeping up w/ demand. Accord uses a regulator on return line to maintain a constant pressure. Pressure dropping under heavy load, where you report the problem, might be a cause. You can rig a fuel pressure gauge to windshield w/ tape to allow monitoring while driving.

You've eliminated all the "usual suspects". What's left must be something very subtle.

good luck

 

I'll give it a try. One more thing I've noticed when driving with scantool connected and looking at live data. When I step on gas and jerking/bucking is happening, ECU sharply retards timing. Like, driving at constant speed, timing is, say, 25 degrees BTDC, when i sharply step on gas it becomes, say, 9 degrees BTDC and the car is jerking/bucking for a few seconds. Then timing gets back to what it was, and there is no jerking/bucking. I'm pretty darn sure, this is timing/ignition related - not fuel delivery. But machanical timing is 100% correct, timing is controlled only by ECU, distributor does not rotate, it has fixed position. And ECU is o.k. too, wire, which goes from ECU to the distributor and controls timing is o.k. too. I suppose some sensor sends some wrong data to ECU and makes it to retard timing or so. But maybe I'm all wrong, and timing is supposed to be retarded on sharp acceleration. BTW car reaches top speed, just much slower than it did 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.

On earlier models and perhaps yours also (obd1 codes appear in shop manual), jumping the service check connector fixed the timing at 12 btdc for timing check. Suggest trying this to see if makes any difference in drivability.

Also might look at your obd1 scan data for anything unusual; IAT, ECT, MAP, etc. You've probably already done this.

Some scanners have ability to track engine misses and total them by cylinder. Does yours have this capability? It might be helpful if you knew which cylinder/s was causing intermittent stumble.

I keep returning to fuel distribution somehow since this is not well monitored by obd2 (no individual fuel trims, only system LTFT and STFT). I had a problem w/ 01 Celica which had P0171 that was finally traced to fuel injector imbalance. Engine ran fine, but ECM pushed LTFT to high + (25-30%) LTFT to balance lean condition caused by one injector. Fuel pressure was fine. However for your case, an alternate set of injectors w/ same problem pretty much rules this out.

It may be time to seek out a replacement PCM. Have you checked that your unit is correct PCM for engine/drivetrain?

good luck