running lean and i don't know why

[hardflex]

did ya hear about the engineers trying to troubleshoot the 914? the first one was a mechanical engineer, and he said "adjust the MPS". The second was an electrical engineer, and he said" rewire the ECU".

The Third was a computer engineer, and he said "close all the windows, shut it down and restart it"

you've probably heard it, but it seemed to fit this thread.

[Bleyseng]

You got that right!

Brad is totally right on though, you need to take the damn car to a dyno shop to have the exhaust analyzed. Only then will we know if its lean or what under load.
We can armchair advised you on this until the cows come home, but we are going just by the plug color. Yes, it does look alittle lean but....
If the engine is running lean why? Vacuum leaks, advanced timing? Not the right parts?

Geoff

[jkeyzer]

Who knows. Why would a vacuum leak make d-jet run lean?

I've been through all the basics, if it was something obvious I would have found it by now.

With dyno time at $75 an hour and figure at least two trips to get it right, that's a pretty expensive solution, but it's a possibility.

Never heard of a shop that specializes in emissions other than the smog check places.

PS: I'm pretty sure it's at least a little lean. If I connect the airbox temperature sender, the car runs like crap. I run with it disconnected. So right there I am starting off with needing some extra fuel to get it reasonably drivable.

[Bleyseng]

Bottom line-If the car is running lean bad things can happen pretty fast. Heads get too hot, aluminum heat cycles and gets brittle, seats fall out, valves get too hot, snap off etc.
Spend the $150 for the 2 trips to sort the car out and then drive it without worry.

Geoff

[jkeyzer]

I am borrowing a Gunson gas tester from James this weekend.

What CO reading am I shooting for at what RPMs?

[pbanders]

2.5% at 2500 rpm under moderate load. But the Gunston is no substitute for a shop gas analyzer, IMO.

[pbanders]

Originally posted by jkeyzer
With dyno time at $75 an hour and figure at least two trips to get it right, that's a pretty expensive solution, but it's a possibility.

$150 in comparison with retrofitting an aftermarket system, or buying new FI parts (e.g. MPS's, ECU's, injectors, sensors, etc.) without knowing if they are the cause of the problem seems cheap to me. The only reason my car hasn't been on a dyno is that I've been able to figure out what's been wrong, if I couldn't, I'd likely have it on the dyno, too.

[seventythree]

If you are in California, go to a smog shop and get a pre-test done for $19~$49. They won't let you make any adjustments, but at least, you will get a baseline for what your current mixture is.

[jkeyzer]

I've been busy recently but have made some more incremental progress on investigating my FI.

Manifold vacuum:

I tee'd in my Mityvac (to the MPS line) and checked manifold vacuum at idle. Last time I did this was before my first valve adjustment and I wasn't sure how accurate the results were. Here's what I got:

800 rpm (what I normally run) - 10 inches Hg, 1.5 inch rapid vibration of the needle

1200 rpm (set using idle bypass screw) - 14 inches Hg, rock solid reading.

Gunson Gastester results:
I am borrowing a Gunson box from red-beard. I don't know how accurate it is but I was hoping to get a general idea of my mixture at idle and 3000 rpm with it. This is what I found. I tested a couple times to get a general idea of the numbers.

With TS1 disconnected:

Idle - 6% CO
3000 rpm - 8 to 10% CO

With TS1 connected:

Idle - 3 to 4% CO
3000 rpm - 6 to 8% CO

So it runs leaner with TS1 connected, as expected. But in all cases the Gastester is showing it is excessively rich, yet the spark plugs are a light grey color.

I am going to look into this more with a real exhaust analyzer and a d-jet tester when I get a chance.

That is all for now.

[hardflex]

Random thoughts -------

doesn't the rapid variation in the vacuum at idle indicate a valve problem, a leaking valve or guides?

I've never checked my 914, but I thought vacuum around 14 in was about right. I'm wondering why your vacuum falls off at idle so much.

Have you checked the vacuum diaphram on the distributor? I wonder if it may be leaking and causing a faulty advance curve (vacuum adjustments affected).

Did you take mixture measurements at the 1200 rpm range, when the vacuum needle is rock steady? Were they consistent with the 3000 rpm range?


Inquiring minds want to know

[Bleyseng]

Setting the idle to 800 is alittle low, try 950-1000 rpms. You should have a good 14-15hg of vacuum with a steady needle.


Geoff

[jkeyzer]

hardflex - I usually run the idle a little low because if I set it high, I sometimes have problems with it settling high at stoplights. But to me that is a whole other issue.

I measured the mixture at 800.

I read a post by Brad a while back that said to test vacuum with the idle raised to 1200, which is why I did.

Today I drove up to Pelican Parts to use their VW 1218 D-jet tester. My d-jet system passed all tests.

I also put my car on their exhaust analyzer for a while. Their tester is saying that at 800 rpm idle I am very lean, around 1.5%. At 3000 rpm I was about 3.5%. This is with TS1 disconnected and the ECU idle control at full rich! With TS1 connected, my CO dropped to 0.5% at idle and 1.5% at 3000 rpm. I did find that if I raised the idle speed up I could get about 2-2.5% CO with TS1 disconnected and the ECU in the full rich setting.

I was very surprised by the idle readings especially, as the idle does not sound particularly lean to me and bogs down when I richen it all the way on the ECU.

HC emissions were under 400 ppm in all cases, maybe around 200 ppm most of the time.

[hardflex]

nothing wrong with your djet that hooking up the TS1 and adjusting the mps 1/2 turn out wouldn't fix. drill a 1/4" hole directly in the center of the plug, and it doesn't even affect your full throttle setting. I'd send you mine just for the pleasure of hearing the engine that's been starving for gas finally get the mix it wants (if I could spare it).

I battled lean, inconsistent running on my 1.7 after I put in 1.8 jugs for years, tried resistors, potentiometers, bigger injectors, higher fuel pressure, asked everybody who might know, until I parked it in frustation. I said more than a few curse words about the Djet. now i won't go without it.

You've done your homework and checked everything else, in 20 min's it could all be behind you.

[jkeyzer]

The only thing I don't know is if for sure there are 1.8 jugs in there!

[jkeyzer]

I was told recently that a dyno will only tell me the mixture at WOT since that is where dyno runs are made.

Is this true? Geoff, how did you tune your car on the dyno?

[Bleyseng]

I made some runs with it like I was street driving say...3rd gear at 3000 rpms to get the data. Then a run in 4th slowly starting from 2000 rpms to 5000 rpms to get a graph with the fuel mix plot across the rpm range.

You can just do it like this, start the run in 4th at 2000 rpms (basically partload) then go to WOT all the way up to 6k.
You should get a dyno sheet showing hp and torque curves but ask for a A/F mix plot. The sheet will show a plot of where you are at at part load (2000rpms) and then the plot across the rpms upto 6k. The curve should be fairly flat, starting at 13.7to1 then going to 12.5. It will vary some but what you are looking for is starting out at 14.5to1 and then going to 13.8 to 1 at WOT. That is too lean.

Geoff

[pbanders]

1.5" fluctuation at idle sounds like a possible sticky valve to me. 10" is low, too, possibly due to the valve sticking. You're in San Diego? Somebody on the list recommend a good shop for Jeff to stop by and have a look at his vacuum gauge readings.