running lean and i don't know why

[jkeyzer]

I tried adding a 220 ohm ballast resistor to the CHT just to see what the effects were. It made the idle too rich, even with the idle adjustment all the way CCW (lean). I also had some really bad idle-transition performance from a stop when I first started the car after it sat for 10 minutes - the car wanted to die until I rev'd it past 1500 rpms.

Despite all of this, I still had mostly white plugs on #2 and #3, although #1 is intermittently getting black deposits. Could be an issue with that injector or spark plug wire, etc as I saw that earlier when I swapped MPS's too. #4 is a pain to check (VDO CHT gauge thermocouple is on that spark plug) so I have not looked at it since changing the ballast resistor.

I also tried 150 ohms, that still caused my idle to be too rich but didn't really do much else. And I tried 550 ohms, that made my plugs nice and BLACK, and the car hesitated very badly, on the verge of not drivable.

[chrisreale]

Are your grounds nice and clean and secure?

[pbanders]

Jeff, check the trigger points, there's an outside possiblity they're bad.

As for general reasons why you'd be lean (most/all you've checked):

1. Higher displacement than expected
2. Wrong injectors
3. Low fuel pressure
4. Low fuel supply
5. Defective MPS
6. Defective ECU
7. Shorted CHT sensor (but would only be lean during warmup)
8. Totally screwed up trigger points, contacting only intermittantly, causing average mixture to be low (likely you'd have all other kinds of sputtering problems)

If you're really stumped, see if you can rent the VW1218 tester from PP. It's my best tool for immediately evaluating the state of the FI system. Checks all components, including all ECU functional blocks, MPS operation, and dyanamic testing of the temp sensors, as well as a full harness checkout.

[Dave at Pelican Parts]

Originally posted by pbanders
[Dave contends that the FI makes assumptions about the engine size...] I don't think this is true. ...
So, how does D-Jet vary mixture by specific displacement? By changing the flow rate of the injectors.

That sounds, to me, like an assumption about the size of the engine. It isn't built into the electronics, no. But it is built into the system--they sized the injectors to the engine.

--DD

[pbanders]

Originally posted by Dave at Pelican Parts
That sounds, to me, like an assumption about the size of the engine. It isn't built into the electronics, no. But it is built into the system--they sized the injectors to the engine.
--DD

Sorry, since Jeff was talking about the ECU, I thought when you said "the FI", you meant the ECU. My mistake.

[hardflex]

school me here, it's been my understanding that Djet and Ljet are similar systems, the difference being that Djet calculates air flow through the system based on pressure change and volume of the container, in this case the intake manifolds and displacement of the engine, so engine size would be a major factor if it was now %10 larger.

The Ljet actually measures airflow past the air flow meter and injects the appropriate amound of fuel. It is therefore much more tolerant of changes in engine size and can make the adjustments on it's own, to a point.

Am I wrong on this?

[jkeyzer]

So if I am going to adjust my MPS to try and change the mixture, how dark of a tan/brown am I looking for on the plugs?

Also, can I measured exhaust gas composition with the engine rev'd to 3000 as a general indicator of the overall mixture? I think PP has a gas analyzer I can use if I drive up and I'd like to check things out in addition to using the d-jet tester.

[dharder]

I had a little time yesterday to fiddle with my friend's 76 that was (is) displaying some of the same symptoms that your is. I tried swapping his distributor with mine, which has brand new injector points, and it made no difference. Here's what I found:
1) He had his coil hooked up backwards and when I fixed that it ran considerably better.
2) #1 cyl had zero compression. This would explain the gradual overheating everytime he drove it and the basic poor performance. His was a brand new engine rebuilt by AVP in Sacramento with hydraulic lifters and cam and I suspect that there has been a valve train problem since it was new because it was displaying the same symptoms from the get-go.

It doesn't sound like any of this would relate to the problem you're having but you might want to do a compression check.....

[pbanders]

What dharder points out is something I emphasize repeatedly on my web page - you MUST be 100% certain that the mechanical condition of the engine and the ignition system are in perfect working order before you take on the FI system as the cause of whatever problem you're experiencing. Often, the best way to do this is to have a new set of eyes look at your car, as they may see the problem that is staring you in the face.

I'm not saying that Jeff here hasn't already done this, just that it's the "zeroth" step of analyizing any FI problem.

[hardflex]

make sure you've read Brad's page on the adjustments, and understand it thoroughly.

if you're adjusting without a load or gas analyzer, when you turn the inner screw it's just like adjusting a carb idle mixture screw, just adjust for highest rev's at a constant throttle position. I started about 2500 and the rev's increased when i got the mixture right. I then verified the slope of the transition by unplugging the TPS (you don't want the idle circuit to kick in) and checking the setting at idle speed. Same thing, adjust it for highest speed, smoothest running. If there's a variance in the settings, you'd have to adjust the outer screw accordingly (I didn't have to do that.

I have yet to adjust the full load setting. I was going to go full load up a hill for a sustained period, pull a spark plug, and examine the mixture. Crude, but it might work.

Of course, adjustment with a dyno and gas analyzer would be preferred.

[Bleyseng]

As dharder says, you adjust the small inner screw for the right mix at 2500rpms. Do not adjust the outer screw. Check the A/F at 2500rpms with the analyzer as it should be at 13.7. I have found that adjusting by ear gets you close but still lean.

To set the WOT you can measure the stop plug first before you disassemble it. OR the correct way is to use a inductance meter like the Wavetek to set it. The stop must rest against the diaphram at 0hg otherwise it will stress it and crack it. (try to find a replacement these days).

Geoff

[jkeyzer]

I still wish there was a way to change the mixture inside the ECU. A lot cheaper than getting a $175 LCR meter. And I'm an electronics guy who usually is in favor of spending money on test equipment!

I think if we fully understood the ECU we could find a way to lengthen the base pulse length from the pressure sensing loop, regardless of temperature or pressure, and change the mixture over all RPMs.

[Alfred1]

It would be great if someone could "reverse engineer" the old analog ECU and make a new (and better) digital one. Could put it inside the old ECU housing to keep the stock look.

[Dave at Pelican Parts]

Brad Anders has about as thorough an understanding of the analog ECU as anyone does. Searching through his site will show you (at least some of) what he has figured out. There are also circuit board diagrams there, and also availble here on this site.

As for a digital equivalent of the D-jet ECU, check into Megasquirt. Uses the same general concepts as D-jet, but uses modern digital electronics and cheap off-the-shelf parts. It can be programmed to emulate D-jet pretty well; one of the BBS members has done the job already. Do a search on "megasquirt" here on the BBS for a link to his page.

--DD

[Bleyseng]

The ECU's controls the injection pulses and the MPS senses what the level of vacuum in the engine is. Low vacuum equals high flow of fuel, high vacuum equals low fuel flow. The MPS's where set to a calibration for each run of engine type, 1.7, 2.0L etc.
Easy just to adjust the MPS.
Geoff

[pbanders]

Originally posted by jkeyzer
I still wish there was a way to change the mixture inside the ECU. A lot cheaper than getting a $175 LCR meter. And I'm an electronics guy who usually is in favor of spending money on test equipment!

I think if we fully understood the ECU we could find a way to lengthen the base pulse length from the pressure sensing loop, regardless of temperature or pressure, and change the mixture over all RPMs.

The ECU is understood fairly well. If you want to add an "overall mixture control", in addition to the one that already exists in the MPS, you can do so by modifying the speed control circuit on the daughter card to provide an overall speed-independent offset. This is essentially what the idle mixture control does, except that its only active when the idle switch is closed on the TPS.

You could very easily experiment with using the idle mixture control by disabling the idle switch in the ECU. While you would lose independent control of the idle mixture, this may not be an issue. You could also duplicate the circuit used in the idle section to duplicate the level offset for non-idle conditions (e.g. get a second pot) if you wanted to.

But, I'll say it again - I think all of this is a band-aid and doesn't get to the root cause of your problem. You need to first find out why you're running lean. If it's displacement, and you're well over 1.7L but still using 1.7L injectors, you should get 2.0L injectors. If there is some other hidden problem (e.g. compression, vac leak, timing issue, etc.), putting a pot on your ECU may fix the symptom, but you'll still be running sub-optimally.

[pbanders]

Originally posted by Alfred1
It would be great if someone could "reverse engineer" the old analog ECU and make a new (and better) digital one. Could put it inside the old ECU housing to keep the stock look.

We've discussed this topic extensively on this list and on Rennlist. The problem here is that the MPS is a highly integrated component with the ECU, you cannot easily replace either component without replacing both.

Dave's suggestion of using a Megasquirt (or some other aftermarket FI system) is a very good one. It requires a fair amount of engineering to do it correctly, but now there are a couple of good examples out there. Please note - don't expect that simply switching to an aftermarket FI system will result in some big gain in performance or driveability over a good working stock D-Jet FI system. Converting is more of an issue of addressing the long term maintenance of the car, and of a cost/benefits analysis of whether it is cheaper to fix your D-Jet system or to replace it.

[Alfred1]

Brad, this aftermarket system looks very interesting to me. I like it because it's designed for an air-cooled VW motor, but, I would like to modify it to work with the d-jet manifold and runners so that it looks stock.

Alfred

----

REDLINE WEBER Fuel Injection

Our System Includes:
Billet or Cast throttle body.

All 5 sensors 1: map sensor in ECU, 2: Engine Temp sensor, 3: air Temp sensor, 4: Throttle Position Sensor, 5: O2 (Oxygen) Sensor. All fully functioning and mapable from simple desk top dash board system.

Full wiring harness with pre-assembled, plug & play.

ECU ( basic map pre-installed, based on engine size & horsepower. )

Injectors for all power ranges.

ECU will operate high & low impedance injectors

User friendly tuning software with six viewing options includes tuning CD and instructions.

Windows based software, Windows 95 or better.

Fuel rails and fuel regulator.

Performance high pressure pump.



ECU features:
Built in Map sensor ( for altitude & atmospheric changes. )

Direct P.C. connection for easy lap top tuning, uses windows based software with real time interface - Tune while running

Tuning views from simple dashboard controls to full mapping for professionals.

Ability to data log/download and monitor all sensors, systems, maps and download or send over internet to tech support for tuning assistance.



Benefits:
Cold start and warmed systems for better starts.

Crisp idle and throttle response

Map sensor for better fuel control on long trips or mountain rides. ( altitude compensation )

Cleaner mid range, better torque.

Better range of usable power, not as sensitive to atmospheric change like carburetion.

Most of all more reliable, no needles to stick, no flooding or shut down.

[Dave at Pelican Parts]

Sure, no problem! What is it, $2500 or so? Pocket change!!

If you're actually looking for a useful substitute for the D-jet EFI, I very very strongly recommend researching Megasqurt. Lame name, good system. It'll cost you a couple of hundred to do, if you already have a laptop Windoze computer and soldering equipment--and a complete D-jet system. More if you need to buy other parts. Dave Hunt has a running system, and he has screen saves of his parameters. It will be less tuning than trying to adapt a Type I oriented system like the Redline one.

--DD

[pbanders]

One more thing in this discussion - to date, if I'm not mistaken, we're relying totally on reading plugs for the mixture information. While this is a time-honored technique, if you're talking about spending big bucks for aftermarket systems, ECU mods, etc., to fix the problem, I suggest putting some numbers around "how lean" first. Get the mixture tested under load with shop quality equipment so that you can see how far off you are. This can be done either on a dyno with the car hooked up to a gas analyzer, or with in-car analysis (there are some 5-gas analyzers that work off 12V and can be connected to a car while driving). I would go to a local shop that specalizes in emissions analysis for this kind of service.