Still can't find what causes my problem Options

[andreic]

Well, here we go again. A month or so ago I posted about the troubles I am having with my 1.7L D-Jet engine. I did a lot of testing since, but still can't find the root of the problem. Help!

Here is what has been going on for the past few months; before that the car was running perfectly.

When starting the car cold it runs perfectly for a couple of minutes, idling high as expected. For these few minutes the car drives well. Then the problem starts. The idle drops to the point where the red alternator light comes on in the dash. The engine bogs on transition from idle to power, sometimes backfiring. If driving, it has low power, and it constantly stumbles pretty badly. As it warms up things get better, to the point that after about 15 minutes of driving it is almost ok (you can still feel a bit of stumbling at all powers, but it gets much better than before). The idle comes up to almost where it should be after warming up, but it is never smooth, it hunts and feels as though every once in a while not all cylinders are firing. It is impossible to bring the idle up above 850-900 rpm by unscrewing the bleed air screw; after a certain point it does nothing. Turning the idle mixture screw on the control box has almost no effect on the idle (unless turning mixture very low, when the idle drops further).

Diagnostics performed:

1) Fuel pressure checked 28 psi, very steady. All hoses in the fuel system replaced, new fuel pump before the trouble started (and ran well after).
2) Compressions uniformly 120-125 psi.
3) Valve gap adjusted three months or so ago, a few weeks before problems started.
4) New spark plugs, new points, rotor, and cap. Dwell angle and timing checked several times.
5) MPS checked: holds vacuum well, resistances checked good.
6) CHT sensor measures right where it should be when engine cold; wiring to the harness good.
7) Removed plug from TPS; drivability problems still there, or slightly worse; idle slightly lower.
8) Removed D-Jet harness from the control box (brain), measured resistances between various pins as described in Anders' instructions, all within specs.
9) Replaced coil and spark plug wires with others I had in my garage, no change.
10) All vacuum lines replaced with new.
11) Checked the injection points in the distributor by taking them out (they looked to be very clean) and checking that they do make a contact while turning the distributor cam (I don't have an oscilloscope to test them further).

To me it feels as if the engine either does not always have a spark (but then why would it run consistently well in the beginning, and then be much worse, then improve as it warms up?), or that it is running very lean, for some reason.

Things I could still do:

1) Replace condenser on the distributor.
2) Check that distributor canister holds vacuum on both ports.
3) Check fuel pressure with engine cold, when the problems I mentioned are worse.
4) Send the injectors to be checked and cleaned.
5) Recheck the valve gap.
6) Try to clean up the various places where the harness grounds are.
7) Other things??

I don't know what order to try to do the above things, or what else to try. This problem has driven me nuts! Please help with any suggestions I may not have thought of.

Thanks,
Andrei.

[mepstein]

If you send out your injectors, Consider using mr injector. He's done over 150 injectors for me and makes them look and run like new. Before and after flow sheets, new seals and hose. Quick turnaround. $17.50 each.

Good luck with troubleshooting and don't give up. Stock fuel injection is a wonderful thing - when it's running right.

[TheCabinetmaker]

From your location, I assume it's cold there. The aar is powered by 12 volts. Even without power it will still close from engine heat, but it would take a long time to close, and maybe never completely in low temps. Did you set dwell before timing, and disconnect and plug diz vacuum hoses before checking timing?

[914Sixer]

Go for the distributor condenser, IT can cause weird problems. Cheap and easy to replace.

Note: Non all condensers are equal. Cables on the new Bosch may NOT reach from mounting point to on clip to hole in distributor. Go for a NOS one if you can find one. I might have one in the stash.

[dr914@autoatlanta.com]

I agree, check the dwell and then the timing at 3500 rpm, and if the condenser is old replace it. Also make sure that the points are good
Regap and clean the trigger points and check the ground leads at the back of the block

Check the head temp sensor resistance as the engine warms up and monitor, at 40 degrees it starts at 2750 and ends up fully warm at about 50, monitor the change

Check the aux air valve for being open cold and uniformly closing until fully closed after about ten minutes

Make sure that your thermostat is connected and that it opens as the car warms up

Finally make sure that the valves are in correct adjustment and not tightening up as the car warms because of an improper adjustment

From your location, I assume it's cold there. The aar is powered by 12 volts. Even without power it will still close from engine heat, but it would take a long time to close, and maybe never completely in low temps. Did you set dwell before timing, and disconnect and plug diz vacuum hoses before checking timing?

[brant]

2nd for checking the CHT hot
in fact... I suggest pulling the sender out of the block and testing it
I've had senders that seem to corrode and loose the ground

I once took one out, cleaned it, reinstalled it. and solved a problem on mine.
screwing it out and back in, will chase the threads and improve the grounding.

[914 Ranch]

It has been my experience that if the idle screw does nothing there is a vacuum leak somewhere.

[Spoke]

It has been my experience that if the idle screw does nothing there is a vacuum leak somewhere.

(IMG:style_emoticons/default/agree.gif)

Sounds like you have a vacuum leak. If you haven't already done this, get a can of carb cleaner and with the engine running, spray it on every hose, every intake joint like manifold to head.

You'll be listening for a change in engine note when you hit the place where the vacuum leak is.

This test is easy to do and the carb cleaner will also clean the engine. (IMG:style_emoticons/default/beerchug.gif)

[andreic]

OK, I did some more testing today. I got a 5K Ohm potentiometer from a local Radio Shack (this was the hardest part -- most local stores have closed and the closest one is 30 mins away...), soldered some wires to it, and replaced the CHT sensor with the pot. Now I can manually adjust the mixture from inside the car, while driving!

And indeed, with careful adjustments (which required changing the pot setting every few minutes as the engine warmed up) the car runs like a dream! Lots of power, no stumbling, no backfiring either cold or warm.

So my theory that the engine was running lean proved correct; or at least I know that the problem is one of mixture, and not one of spark or fuel pressure.

My next step will be to remove the CHT sensor and clean it up and its threads thoroughly and reinstall. I strongly suspect this will turn out to be the culprit.

For some reason I don't believe the problem is a vacuum leak. First of all, since this is an MPS system, if I had a vacuum leak it would only lead to the idle being high, but the MPS should adjust the mixture correctly. (It is L-Jet that would run lean if it had a vacuum leak, because it would let in unmetered air.) Second, all my vacuum lines have been replaced with new and fit very tightly, so my guess again is that there are no leaks. But if the CHT sensor is not the problem, I'll go down this rabbit hole too.

[Dave_Darling]

Be very very very careful removing the CHT sensor. It is steel, and the head is aluminum. It loves to bring the threads with it out of the hole... Repairing without removing the engine is a pain in the butt.

--DD

[Spoke]

My next step will be to remove the CHT sensor and clean it up and its threads thoroughly and reinstall. I strongly suspect this will turn out to be the culprit.

Good point about the MPS system. A vacuum leak is like opening up the throttle.

If you want to check the CHT sensor, it's best to leave it on the engine. The CHT sensor is a negative temperature coefficient resistor which drops in value as it heats up. As long as the CHT sensor is on the head, it will attain the same temperature as the head as it is supposed to do. I think it goes from something like 2.2k ohm cold to a couple hundred ohm when warmed up.

Since it's just a 2 terminal resistor, one terminal is the case and the other is the single wire. If you want to check the case, measure resistance from the case or body of the CHT sensor to the engine case. It should be zero ohms.

To test the CHT sensor, put an ohm meter from it's one wire to ground to the engine case and check the resistance. Watch the resistance as the engine starts and heats up and write down the values. I think Banders has a graph of what the resistance is over temperature.

[andreic]

OK, so what is the best way to take out the CHT sensor without causing damage? I can do all the tests suggested by Spoke, but if the grounding of the sensor is intermittent as some others have suggested I need to do something about it.

Should I soak it in penetrating oil before getting to it? Any other tricks? I really would not want to have to drop the engine to fix threads in the head...

[mepstein]

Buy the longest/deepest socket you can find. Doesn’t have to be high quality. $5-6 usually. Coil the wire in the socket and remove the sensor.

[dr914@autoatlanta.com]

tap the head of it to unfreeze the threads, use a 3/8ths short extension and cut a v in it to pass the wire through, and then use a 11mm socket

be very careful you do not want to break it off in the head

[Spoke]

OK, so what is the best way to take out the CHT sensor without causing damage? I can do all the tests suggested by Spoke, but if the grounding of the sensor is intermittent as some others have suggested I need to do something about it.

Should I soak it in penetrating oil before getting to it? Any other tricks? I really would not want to have to drop the engine to fix threads in the head...

Did you check the resistance of the CHT sensor? Not sure by your post if you've tested it.

[914_teener]

It has been my experience that if the idle screw does nothing there is a vacuum leak somewhere.

(IMG:style_emoticons/default/agree.gif)

Sounds like you have a vacuum leak. If you haven't already done this, get a can of carb cleaner and with the engine running, spray it on every hose, every intake joint like manifold to head.

You'll be listening for a change in engine note when you hit the place where the vacuum leak is.

This test is easy to do and the carb cleaner will also clean the engine. (IMG:style_emoticons/default/beerchug.gif)

Yep and my money is on the vaccum canisters.