O.K., I'm better now... D-jet is driving me to drink.

The 1.7 starts and runs well until it warms up. After 10 min. idling smoothly @ 900 it hesitates two or three times and stops idling at all, which means I can't keep it running long enough to set the timing. I can keep it going by feathering the throttle, but my nice smooth idle is gone. I reproduced the symptom three times in a row today.

By clamping off the Vac hoses one-by-one it appears the FI components seem to be doing what they're supposed to. No other vacuum or capacitance probs. found so far. Haven't checked the injectors.

This started after I repaired the ground strap on my distributor -- must have bumped something else. CHT?

I have the FI diagnostic checklists and am working on them. If this was a carburetor I'd say it was running with a clogged idle jet . These symptoms seem specific and uncommon. At least I've not seen others posts that describe the same thing. Did I miss one?

Try this...

L Jet Manual

Could be. I have seen this exact behavior when the CHT had been disconnected ...


Does it smell rich after warmup? A disconnected CHT will dump lots of fuel into the cylinders when warm.

Thnkz Mike,
Although "my" ECU monitors mainly manifold pressure (its a dog, not a lamb) I've been looking for the L-jet document you linked. IMHO, to take best advantage of the resources offered from 914 enthusiasts U've got to be able to parse suggestions and checklists that might represent "blended" experience from the two FI management system. So I've enjoyed reading the book.

As an aside I note that in the AC VW world there's a consensus that "D-jet, bad, L-jet," good." Not so with "Teeners," as far as I can see. In fact it's somewhat the opposite. Interesting, no?

Best wishes.

Dear Sir,

I appreciated your post 'cause after reviewing Anders I'd become over excited about the ignition timing. I once worked on a car where the dizzy simply shut down when the engine reached a certain temp, and that took forever to diagnose, but I don't think that is my issue here.

Your comment reminds me that, if the hypothesis is a fuel delivery issue the only conditions relevant to that analysis are "lean" or "rich." Engine stall because of either can offer identical symptoms (besides smell). I need to determine which it is.

I wondered / hoped whether that might be CHT, but I'm working in the breeze so I can't remember whether exhaust smelled rich. I'll repeat and pull a plug right after it stalls.

Another thought was that some failure of primary injectors meant that engine would "lean out" after cold start valve shuts off; however, further reading tells me that CSV shuts down much sooner that the 10-15 minutes of idling I observe prior to the engine stall.

I check the exhaust smell and CHT continuity this AM and report back.

Thanks again!

The cold start valve does not come on till the air temp hits about 40F.

The Auxiliary air regulator however, will open up on a 60 degree morning to give extra rpms during warm up. after about 10 minutes it should close and the idle come back to 900 (or wherever its is set). You may just have the idle set to low, or the throttle position switch may need to be reset. A mis adjusted tps will render the idle mixture knob on the ecu useless.

Great info! thanks. Assumed incorrectly that without extra fuel delivery, an open AAR would over lean the idle mixture -- ECU must adjust for proper mixture some other way.

My cold engine idle is first 800, then rises to 900 in a discrete stage. The third stage is stall. Are you suggesting thesel RPM values may be too low? I.e., if the TBP is mispositioned and the idle set too low, discontinuation of the AAR function would bring it down to a stall condition.

I'm going to check Throttle position sensor and re-check AAR and report.

The CSV does not come on unless the air temperature is Below 40 maybe even colder. Unless it is leaking, it should not be a problem here.

Great minds think alike. Needed to know that. Thanks.

Yep, i'll bet AAR myself. Should idle about 1500+ for 5 minutes or so, then come down. Base idle is too low.

Unplug the hose to the AAR at the air cleaner and plug it. Then try to start the engine cold. If it won't start you "may" have found the problem. open the air bleed screw on the throttle body a couple of turns, and try it again. The tps and ecu idle mixture must be set or it still won't work as it should. It'll be better, but still not 100%

From above;
Great info! thanks. Assumed incorrectly that without extra fuel delivery, an open AAR would over lean the idle mixture -- ECU must adjust for proper mixture some other way.
The ECU adjusts fuel mixture based on several inputs, one of which is RPM's.

I think vsg914 has a good test for you, although I would do it a little differently. Start car normally when cold, after about one min, close off the hose to the AAR from the air cleaner by pinching it with some pliers. If the engins stalls, you have the idle set too low. As others have said, with the AAR operating and open, your idle should be 1500 or so.
Tom

Tom, is it really Magala Red?

Tom, thanks very much for that clarification. It must seem obvious to you and now it is to me too.

Along those lines, do you know of a document that describes what the ECU "does" with the input from each pin. Manuals and checklists I've seen tend to focus on a description of what each device in the system measures with less detail about how the ECU responds to each signal in order to maintain a "stoichiometric air-fuel ratio".

The TPS must need attention. I think the idle screw is at the end of its travel for it would only turn 1/2, and as vsg914 predicted, that had little effect on idle speed. Though the engine now will continue to run after warm-up, there is little throttle response after the AAR kicks off

Synopsis:My problem was too slow idle during warm-up and then a predictable and sudden stall at 10 minutes. Plugs were fouled. CHT lead had fully or partially ground out. Fixing it (and /or changing plugs) brought the cold idle up from 900 to 1500. Even so, clamping the AAR stalled a cold engine, but a warm engine continued to run without AAR, albeit unacceptably rough and slow (4-500 RPM.) I know that the vacuum system leaks -- the hoses are 25 years old or more. I don't know why the air screw would not turn in more than 1/2. Perhaps it is at the end of its travel and per VSG914 the TPS needs work.

Thanks for all the good advice. There was a lot of it It really helped.

Even though this problem started after a specific repair... when was the last time you adjusted your valves? Generally with D-Jet systems you want to start with basic, simple (and free) things first. Vacuum for the rest of the system is affected by proper valve adjustment so if your valves aren't right you'll chase your tail on a bunch of other stuff because the foundation isn't solid. Besides that, I've personally experienced that valves which are too tight will run great when cold, but produce a crappy running and stalling engine once they warm up and close up the gap.

That's true for any diagnostic procedure. Thanks for the reminder about the valves. How are things in Flag? You're lucky to live in the city of my favorite airport (free parking).

You must fix all vacuum leaks. The idle adjustment screw is a "controlled vacuum leak". It cannot have a proper effect if the system leaks elsewhere. Fix the leaks first! and adjust the valves.

The information I have is Paul Anders' stuff and Jeff Bowlsby's site at The Classic Porsche 914. He has a lot of info there and lots of wiriing diagrams. You can figure out which wire at the ECU goes to each sensor from the diagrams there, but the best description of what the ECU does is on Paul Anders site.
Tom

Unplug and cap all vacuum lines at the plenum.
the only line that should be plugged into the plenum is the MPS vacuum.

How is the idle then?
Set valves, dwell, and timing.
How is the idle?

If its still bad, and it still shows signs of dropping off after its hot...check your CHT for the proper Ohm readings for your ECU.
Anders site has the correct readings for your ECU and your CHT...
Each year is a bit different.

If that fails...there is are three more things that affect idle.
Failed MPS...runs to rich and stalls
Failed Cold Start Injector , always open, runs rich and stalls
Poorly adjusted TPS...runs rich and stalls.

You really need to ensure your ignition system is set properly before any FI diagnostics...you will chase your tail if the valves are adjusted wrong, or your dwell and timing are off....even just a little bit.

Djet is very sensitive to proper dwell and timing for the idle speed.

Djet and Ljet are both early systems and are both a guess at best for the proper A/F settings.
Both systems suck.

When the LH jet was introduced...that has the O2 sensor and the feedback made the system self adjust...then they had a decent system.

Rich

Thanks so much, Tom. I'll revisit Mr. Anders and Mr. Bowlsby. I appreciate everyone's input.

I agree with the posts that urge dealing with vacuum leaks now. My original inquiry was inspired by a sudden change in engine performance and whether the new symptoms indicated a specific condition. Is that backwards? Anyway, I'll have to take a different approach now since I can't time it in the present running condition.

Before I adjust the valve lash: Are the values stated by Haynes, etc. still thought to be the correct specification? For my 1.7 that would be .06/.06. I ask because at some point it's been agreed that VW type 1 (magnesium engine case) valves should be set slightly over recommended factory spec. I think I'll start a new valve adjustment thread for some questions I have on good head hygiene.

Despite the potential vacuum problems the engine is running much better than before I started this post, so thanks again to everyone who took the time to help me.

look in the classics forum here for Cap'n crusty's valve adjustment process...its simple and works every time.

Rich

Rich, I appreciate your thoughtfully honest comments, that info is gold, e.g., now I'll be able to set the timing.

"When the LH jet was introduced...that has the O2 sensor and the feedback made the system self adjust...then they had a decent system."

Does that describe what some call an "open loop" system?

Have you used this information to customize fuel delivery on your car(s)?

Despite the continuos frustrations, I really enjoyed learning carburetor theory a few years ago, now I'm loving "Fi 101."

PS: Ill go through those procedures, set valve clearance and report back

FI is simple.
You are still measuring air flow.
ALL EFI systems measure air flow.
Based upon air flow, fuel is metered out to match the airflow.

There are two schools of thought on measuring air flow
Manifold pressure, and air flow volume.
Both measure air.
Both still exist in one form or another.

With the introduction of an O2 sensor, the computer can see if it guessed properly and do slight tuning based upon the exhaust gas mixture...

Heat sensors, specifically head temp sensors, tell the computer if the car is to hot, and it adds fuel to cool the motor.

The rest of the sensors are for even more fine tuning.

Rich

Open loop in modern EFI systems refers to the engine running with no feedback from the O2 sensors(and other components) while it warms up. Once the operating temp is achieved the system enters closed loop mode which uses the various EFI components to obtain the maxium F/A ratio. The vehicle will also enter open loop at idle warm or when the engine is not under load.

Thanks benalishhero, So the word "open" as used in open loop, signifies the something similar to "open circuit," i.e., "unconnected?"

Here's what I am presently dealing with:

-1.7 D-jet

-Good compression all 4

-Tune-up, dwell, timing and valves set to specifications per Haynes.

-No obvious vacuum hoses leak (air/idle screw adjustment now affects idle speed.)

-Starts at 900 RPM, after 5 minutes it rises to 1500 RPM. Idle is steady and smooth.

-Clamping the AAR hose on supply side bring idle back down to 900.

-Disconnecting and plugging manifold connections has no significant effect.

-At 15 minutes run time, the engine begins to miss and then stalls after about 2 minutes. Reving motor and/or unclamping AAR doesn't prevent stall.

-Exhaust smells rich when engine stalls--assume flooding.

Per Haynes FI trouble shooting section: "engine starts cold then stalls:" = MAPS or CHT.

This is consistent with the opinions of several members here.

Tomorrow, I will:

-Check CHT resistance value

-Go through diagnostic procedure on MPS.

Anything else?

Where will I find a serviceable MPS if mine's kaput?

Besides, starting with the basics, the other thing you'll often hear is that you should check everything else first before you look at the brain because "they never go bad". Well... almost never.

Your symptom of smelling fuel when it dies sounds alot like what happened with my son's 1.7L at RRC 2 years ago. We went around a corner, the engine suddenly died and it was obvious that it was flooded. After we got it home we went through the entire system methodically like you are doing. Because we had known good parts that we could swap we changed out AAR, MPS, etc. Finally, there was only one part left - the brain. Sure enough, after changing out that the problem went away and has never returned.

If you can lay your hands on a brain that is compatible with the rest of your system, it may be worth swapping it to see if it makes a difference. It's sounding a bit to me like a solder point or circuit board that is breaking a circuit once it heats up.

Thanks, ThinAir:

Those fit the symptoms. At any rate, something specific happens when the engine reaches a certain temperature. There's a clear transition and whatever the problem, it causes the engine to stall within a few minutes.

Not knowing much about FI, I hoped someone who did could ID these symptoms as indicating failure in a particular system. I now understand why it would rarely be that clear-cut.

Lessons learned by anyone who, like ThinAir, has experienced something similar and is willing to share might help me avoid a lot of unnecessary work and expense, as I attempt to source replacements for those expensive components that may be bad.

Update:

Pulled ECU and tested resistance of CHT and MAP at pin harness per Anders.

CHT resistance 2.54K (within a few OHM's of spec.) MAP was right on spec.

My brother suggested an unseen leak at throttle body or manifold that would expand with engine temperature.

Next on list is to test TPS and inspect throttle body for unseen leaks.

I'll keep digging.

All these static tests leave me wondering...

To recap, I'm looking for the condition that causes my 1.7 FI to stall out at a certain operating temperature. The idle is good for about 15 minutes and then it starts hunting, the RPM's drop and it stalls. After checking the plugs, and smelling the exhaust, I believe it is a flooding condition. The functions and behaviors of the FI components (the "likely suspects") also argue that any massive failure of this kind would result in flooding.

The possibilities are almost limitless and I wont discount any of them, but I feel it is one of three: 1. CHT thermistor never warms up. 2. ECU issue. or 3. TBS is doing something I don't understand.

As I said, the static testing only gets me a spec at room temp, and especially for the CHT, this is not sufficient. If it is not decreasing resistance as the engine approaches warm, the mix could conceivably become too rich to run -- I guess.

Its the sudden onset of the problem that really confuses me. it would seem that any falure of a FI component brought on by heat, would show some progression. This problem comes on all at once.

I'm considering splicing a pigtail to the CHT wire and grounding that when the engine is about to stall. Grounding the CHT would essentially render it invisible to the ECU. Good / bad idea?

You should measure the value of the CHT when it is warm.
Tom

Thanks, Tom

my last "idea" -- to ground the CHT when the engine was about to stall-- was a round-a-bout way of checking it warm. Your note brought me back to reality .

Unconsciously, I think I was trying to avoid the trouble of removing the ECU when it was hot. (I've so far done my resistance checks at the ECU 24 pin harness)

I'll check the warm CHT at the ECU pin and also check the warm MAP values at the same time.

I'm going to do this now and report back later.

Appreciate the help!

Started engine, ran well for 15 minutes, hesitated and stalled as usual.

MAP and CHT resistances are within spec at "warm" (=/- 200 F). These values are from the ECU connection pins. I couldn't detach the CHT sensor wire from the harness. There is a terminal block but the spade would not pull out with moderate force and I was afraid of breaking the plastic cover.

Don't have a vacuum tool, so I can't test MAP "leak down."

Next check is the throttle/air distribution manifold assembly looking for something out of order (like a big a-s hole) underneath. Also the cold cold start valve for leakage.

I have heard of cracked plenums causing a vacuum leak. Going to be a hard one to check. Maybe someone on here will give you a good way to check the plenum for vacuum leaks.
Tom

Indeed. I've not seen a lot of info on how the plenum/throttle assembly goes together or, more to the point, comes apart.

ThinAir, did you do more work on your son's 914 since this post?

http://www.914world.com/bbs2/index.php?s=&...t&p=1129848

It indicates that a problem with the MAPS electrical connection caused the car to suddenly flood-out.

Did you later determine an ECU fault? Per:



Thanks a lot!

This actually all started with this: http://www.914world.com/bbs2/index.php?sho...c=80367&hl=

I thought I had posted my solution, but I can't find it.

Yes, I solved the ultimate problem by replacing the ECU. The car would randomly flood and die immediately. We took the approach that the ECU "never goes bad" because they hardly ever have a problem, but in our case it was the only component that we hadn't changed for a known-good unit so there was little choice left. Since we had one available, it was worth a shot. Problem has never returned and the car made a very successful trip to RRC last Fall with nary a hint of a problem.

ThinAir, Thanks for the update. Reading about your experience helps a lot. Sorry you had to go to so much trouble back then. Great work sticking with it to fix the car. Helps me consider my issue from a more positive perspective. Stuart

Rich, I forgot to thank you for the reference to Cap'n crusty's valve adjustment process. Heard of such, but never before in sufficient detail to be useful.

Found it the thread and it worked well for me.

As I was underneath the car, it struck me that the different methods for setting lash seem to represent a spectrum of "automechanica" that starts with a standard Haynes approach. In the middle of the spectrum is the modified method found in "How to Keep Your Volkswagen Alive." and at the end is Cap'n crusty's way, which is the most logical, but requires visualization of the operation of the motor's top and bottom end.

Follow up and diagnosis. Ultimately it was the fuel pump...

This thread helped fix the problem of engine dying from idle when it reached a certain operating temperature. At the point of stalling, no amount of feathering the accelerator would prevent it.

After testing the FI components the important ones gave good readings. I spliced into the harness to test the resistance of the CHT and some of the others over time. While it seemed to be operating to spec, I discovered CHT would momentarily ground out -- flooding the engine, but not killing it. I repaired the skinned wiring leading to the CHT and fixed some vacuum leaks. Tune up was hard because I only had about 5 minutes to set the timing between warm-up and stall. Doing a valve adjustment gained another 5 minutes. Strange. Determined the AAR was remaining open all the time and made allowance in the throttle screw and the idle mixture knob on the ECU until the motor would start and idle appropriately.

After all that I had a much better running engine until it stalled. Every time after about 15 minutes of idle from cold. At least it wasn't running rich anymore. That turned out to be an important clue.

Brought a length of fuel line into the driver compartment and hooked up my water heater/fuel pressure gauge. Went through the ritual of starting car, allowing it to warm up and waiting for it to die. For the first time after two weeks of fooling with this issue I noticed the pressure drop in conjunction with the hesitation that preceded stall. I'd looked for this before from the engine compartment but I think all the other work prevented the engine from dying so suddenly, so this time I was able to notice some pronounced flickering of pressure right before it stopped running.

The next day I fried the ignition harness. Ironically it was not the wires I was splicing into to test resistance that failed. Ordered a nice harness from Jeff Bowlsby and rigged up a temporary connection with the starter, coil and oil pressure gauge so I could continue with the stall issue.

One problem with d-jet is you can't casually swap suspect components for known good versions to track down a problem. Its either impossible to find them quickly or they are prohibitively expensive to buy just for that purpose.

Found a $31 Bosch in-line electrical fuel pump at NAPA that specified correctly. Maybe it won't last a week but I wasn't sure whether an electrical problem was discontinuing signal to the pump, or whether the pump was failing. Now I know. Money well spent even if it prematurely fails. The old pump was failing. It showed good pressure until it didn't. The new pump works great. Engine starts easily/runs better now

Up on JS right now so no test ride but the car idled relatively smoothly for 45 minutes before I turned it off. Every once in a while I'd rev it to keep the oil flowing and the gas off the cylinder walls. Runs 50 degrees cooler at the Ch tins now. I like to think the bad pump was creating a lean mixture and heat. Big difference from when I thought it was flooding out. The ambient is 20 degrees cooler than before, so that's probably the real explanation.

The "idle quirk" problem appears solved, now I can go back to testing FI components, fixing vacuum leaksand dialing in the tune.

THANKS FOR YOUR HELP.

give us the part # of that pump!

Began a new thread -- no idea if it's any good for the long run.

Congratulations on figuring it out! Persistence pays off.