Lean intake backfire

[TheSushiGod]

Good morning Samba,

My 84 1.9 has a lean intake manifold backfire pop if I accelerate hard. It bogs, and pop right through the air filter. This happens after it warms up. If I disconnect the t2 sensor, it runs perfectly and the bog and backfire are completely eliminated. The engine really likes the rich condition of unplugging the t2.

I first suspected a bad t2, so I tested the ohms from the Ecu as well as the sensor itself. The readings were spot on (digijet pro manual). I then tested the AFM from the Ecu and the unit itself (EXCEPT for the potentiometer arm) and the ohms checked out. Air temp sensor too.

My thought was this: somewhere I must have unmetered air causing a lean condition. Performed starter fluid spray check. Nothing. Changed out the throttle to plenum gasket, completely cleaned the throttle body out and noted that the plate sits very nicely with very little gap. Plugged the boot holes for the aar and oil tower to eliminate any issues there. Replaced all of the vacuum lines... All of them. Plenum to intake boots replaced with 1.5 ID hose and clamped down. New intake gaskets and fuel injection rings. New injectors too. Bypassing the idle controller or disconnecting the O2 does nothing. Those are new too. New FP regulator and fuel lines. Grounds cleaned and new ground straps.

Am I missing something? Could the AFM be causing this? I've continually checked for vacuum leaks, I can't find any. After I plugged the s booty idle improved a bit but I think that's because the oil tower was gutted. I inspected the S Boot as well, didn't see any cracks or holes. I'm not sure where I should go from here

[Bassyaks]

Look to see if the throttle plate has any vertical movement, it can look good but pressing on the throttle shaft you can get almost 3/32th" movement that on top of the horizontal movement may cause your problem

[Wildthings]

Checked your fuel pressure and fuel volume according to the manual yet?

Adding a bottle of FI cleaner to your gas tank isn't a bad place to start either.

[TheSushiGod]

I replaced the fuel T with the nut with a brass T without the nut when I re-did the fuel lines :/ the 1/4 inch fuel line wouldn't squeeze onto the older fitting. In retrospect I wish I had gone with 5/16th.

[hdenter]

The fact that it runs better when something is disconnected that puts it in open loop would suggest that you test all the FI components. Test the O2 and it's wire. Usually it fails in the rich direction, but stranger things have happened.

Hans
_________________
'79 triple white convertible bug
'84 sunroof vanagon
'85 weekender

[TheSushiGod]

Brand new O2 sensor; not the splice kind. When I disconnect it nothing changes. 😞

[hdenter]

Did you take your test readings for the different sensors directly or at the plug to the ECU?
_________________
'79 triple white convertible bug
'84 sunroof vanagon
'85 weekender

[TheSushiGod]

For the temp 2 , both. Same ohm reading within spec. For the AFM, same ( just not the arm moving test )

[Howesight]

The most common causes of backfiring through the intake manifold are:

1. Retarded spark timing; and
2. Sticking intake valves. In this scenario, the intake valve sticks open due to varnish buildup on the valve stem and the fuel/air mixture is then ignited by exhaust gases during the time the exhaust valve is open.

Was there old gasoline in your tank at any time?

As for the change to "perfect" running when the Temp 2 sensor is disconnected, that might simply dump in a sufficiently rich mixture to suppress ignition of the charge in the cylinder by the exhaust gases (ie: in the sticking intake valve scenario).

Regarding the temp 2 sensor, you also need to check for the resistance between the sensor wires and the ECU. If that wiring is compromised or the grounds are compromised, then the resistance noted by the ECU will be incorrect and will adjust the mixture incorrectly. Note that the oxygen sensor signal is ignored completely by the ECU until the Temp 2 sensor (and its wiring) deliver the CORRECT signal to the ECU that the engine coolant is at the threshold value. I do not recall what the threshold value is, but it is close to 80 degrees Celsius, IIRC.

Hope this helps.
_________________
'86 Syncro Westy SVX

[TheSushiGod]

I tested the ECU to the t2 first, and then the t2. Identical OHM readings at about 1000 ohms. Engine was warm, but not hot. I don't have a thermal sensor, but it seemed to be in the ball park of 104 degrees. I did replace the pigtail connector to the sensor as well.

I set the timing according to tencents total timing method. I've tried the Bentley method too. Nothing changes.

One odd thing I noted and someone else commented on: looking at the distributor (brand new) from the top down, my vacuum canister is at about 2-3 o'clock. Pics on the web, and others from a vanagon FB page i'm on, all have their vacuum advance cans pointing almost to 12 o'clock. I've tried twisting my dizzy to 12 o'clock, but it begins to run like crud. As soon as i return it to the correcting timing mark with my gun, it runs better again.

Is there a natural "hissing" sound that comes from the engine? I swear i hear one coming from around the top of the intake plenum near the transmission. Yet, I've sprayed a crap load of starter fluid there and not once has it revved. If i cup my hand around the north part of the plenum, the sound diminishes. yet again, i could be hearing just the natural movement of air through the plenum and intake and my hand is simply just absorbing some sound. Is there a better method to detect vacuum leaks besides spraying starter spray?

[Wildthings]

[TheSushiGod]

Understood. Would the 1/4 inch fuel line perhaps be restricting fuel pressure some how? I've read of others using 1/4 and not having problems. I did use a brass barb style T instead of the other type ( i couldn't squeeze that 1/4 onto the original with the hex nut to save my life). Brand new regulator and the bosh fuel pump looks really new as well.

Wildthing, does the plenum area have a natural air rushing sound to it? I swear this thing runs like a champ with that t2 disconnected. I'll ohm test it at the ECU "cold" tonight. I think its around 68ish here in Seattle now.

[Wildthings]

The plenum is pretty quiet, if you had bad shaft seals on the throttle body you might be able to hear air passing through them. Use a length of hose to listen for possible leaks. I would guess that 1/4" hose would be okay, but the original was about 7mm IIRC and 5/16 (~8mm) seems to work fine. You really need a way to check your fuel pressure, so I would suggest making the swap to 5/16" hose so you have a test port again. One little piece of dirt and a new FPR will no longer work correctly.

It runs well with the TSII sensor disconnected because the ECU believes the engine is dead cold and is dumping in lots of extra fuel as a result.

[Howesight]

Starter fluid is not a reliable way to find vacuum leaks - -especially if the idle stabilizer is still connected. A smoke machine is better. There are known vacuum leak sources that commonly arise on WBX engines. Replacing the intake manifold connector hoses, the throttle body to plenum "gasket", and removing the "S" hose to visually inspect it are a good start. The hissing sound is not normal.

Were you able to check for any sticking intake valves?

One other cause of backfiring in the intake tract is a bad/cracked distributor cap. Such cracks are hard to see. Do you have a known-good spare cap and rotor to use to check?
_________________
'86 Syncro Westy SVX

[TheSushiGod]

Good morning!

* intake runners replaced as of two days ago with 1.5 ID hose, attached with hose clamps
* S - hose, new one in the mail. I've inspected it and do not see any cracks. i've also twisted it a bit etc while running to see any change; none.
* Sticking intake valves, would that be a compression check? I'll do that today!
* Cap and rotor are new

UPDATE: did an ohm check of every injector on the van via the ECU. 22 ohms!!! the manual said it should be no more than 16.8 could that be causing this? * ECU disconnected testing at the harness.

[Terry Kay]

<<Sticking intake valves, would that be a compression check? I'll do that today>>

This is the closest thing to reality that you have thought of.

Burnt valve would be closer to reality.
Do the compression check & see that more than likely the problem.
_________________
T.K.

[TheSushiGod]

Compression results 120 for 3 cylinders, 110 for cylinder 2

Any thoughts on the high ohms for the injectors ? The injectors themselves = 16 ohms on the dot

[TheSushiGod]

Sorry the compression tester wasn't screwed in tight enough on cylinder two it is also 120 compression

[TheSushiGod]

I'm currently re-wiring my grounds, i'll keep you posted.

[TheSushiGod]

Now all injectors read 17 ohms on the dot via the Ecu! Unfortunately, I'm still getting massive hesitation with the t 2 connected and runs great disconnected.

I pinched the fuel return line with pliers but leave t2 connected, it runs crappier but 50% of the hesitation subsides and it stops backfiring ...