Hot Start Issues - FI bus super rich when hot **FIXED**

[canadianveedub]

So, this is another hot start issue thread, however, I have narrowed the problem down to being way to rich. Here is where I am at.
I have a 1979 Cali bus with EFI and electronic ignition. I have only started experiencing this in the last 4 weeks.
Like most others, I do not have the problem if I shut the bus down for a few minutes, or if I let if cool completely. It only has issues starting when it has heat soaked for a bit. The ignition is working and the timing is fine.
I have narrowed it down to the bus being super rich when it is trying to start. If I go to start it, it may kind of fire OK, but then shuts down. No amount of cranking will get it to go, with or without the throttle depressed. I found some previous threads on this problem that caused me to unplug the cold start injector, which almost worked, but led me down the path of trying to unplug all of my injectors. Crank it over, plug them back in and it will fire up and run.
So, my question is, what would cause it to crank really rich when hot?

Thanks in advance.

[Tcash]

http://www.thesamba.com/vw/forum/viewtopic.php?t=602704

[canadianveedub]

The fuel pressure is one of my thoughts. I will check it out ASAP

[old DKP driver]

Is this 1979 CA Bus still original with
o2 SENSOR and CATALYTIC CONVERTER?

I will assume you have a Bentley repair manual and that you have
already verified the cylinder head temp sensor is within specs for both
cold and warm engine temps.

If the o2 sensor is faulty it will contribute to a rich running engine as soon
as 5 minutes running time
_________________
V.W.owner since 1967

[SGKent]

[canadianveedub]

The bus does not have an O2 sensor, EGR or Catalytic Convertor. What I am having trouble with is why it won't start until after I unplug the injectors, crank to clear the engine and plug them back in. It seems as though it starts off cranking rich, leading me towards fuel pressure issues

[raygreenwood]

Or fuel leakage issues. This will be a simple check with a pressure gauge even without the engine running. Hook the pump up direct to pressurize the fuel loop. Pull all five injectors out first (actually get some fresh injector seals first).....and verify that none of your injectors is leaking into the engine. You can check fuel pressure at the same time. Use a hand vacuum pump to take the fuel pressure regulator up to 15"hg and make sure its operating correctly. Ray

[canadianveedub]

So, after a massive goose chase I have found the issue. It was a faulty AFM. Here is how I got there:

When you have a rich condition, there can be several issues, and you have to troubleshoot rather systematically to narrow it down. In this case, I needed a multimeter, a fuel pressure gauge, a bentley manual and some scotch.

I started with checking all of the sensors at the ECU plug. I checked the TSI, the TSII and the AFM resistance, which actually checked out normal. I checked the fuel pressure while cranking and it was a rock solid 36PSI. I moved then to the coil, which was WAY out of spec, so I replaced it. This did not solve the issue.

I checked for leaky injectors by doing a fuel pressure leak down, which was normal. I checked all hoses and they were all good condition with no tears or leaks. I still had the same issue: when I cranked while the engine was warm, it flooded. It could be cleared by unplugging the injectors and then the engine would start. I moved to the last piece of the puzzle. The AFM.

I pried the top cover off and was greeted with severely worn tracks inside the AFM. I adjusted the wiper to clean track and cleaned the tracks with contact cleaner. (I fiddled with it for several hours trying to fix the problem, which I think may actually be an issue with the contacts at the top of the wiper arm)This did not fix the issue, so I decided to probe a little deeper.

The AFM uses an unconventional 7.5V reference signal, as opposed to most that use 0-5V. As you sweep through the flap positions, you should get a linear output that increases steadily. I turned on the ignition and moved the wiper arm while probing pin 6 (gnd) and pin 7 (signal out). The wiper should be 1.75-7.0v. As I moved the arm through its movement, I noticed that it would go down from 7v to 3v in a linear fashion and then plummet to 1.5v for the last entire section at the bottom...right where you are cranking. It was missing several volts of signal. This really tied everything together for me.

For the life of me, I could never get my bus to start when cold without giving it a bit of throttle. I never really thought about it, but moving the throttle moved the flap enough to get it out of the AFM dead zone into an area where it was getting enough voltage to ask for some fuel. When hot, however, it would go from no fuel to flooding out when I touched the pedal as a hot engine will not handle being too rich. I ordered a refurbished AFM from German Supply and got it this afternoon. I crossed my fingers and the bus fired right up with a tiny bit of throttle when cold and like a dream when hot. It runs smooth all throughout the RPM range, although I think it has lost a bit of torque.

For all of the FI guys with hot starting issues, I would suggest you look at the voltage output of your AFM as a troubleshooting step as the resistance checks in the bentley manual checked out fine for a dead AFM.

[SGKent]

Glad it is fixed - FYI - the AFM is not linear. Think of it like a runner whose distance is being measured to six people standing at different spots on the road. Each person is getting a different measurement from the runner both coming and going. Those measurements are fed into the circuit which determines where the door is and how fast it is changing. Both those are needed to know how much extra fuel to add. If for example the door is moving quickly by a spot then the amount of fuel needed is greater (think accelerator pump on a carb) than if the door is static in that spot. The bottom line is that it does not read just one voltage signal, it reads six or seven. And moving the board in any plane changes the calibration which is needed to get the measurements right - even though lots of places say you can fix a worn AFM that way. You can but much the same way bailing wire will hold a door closed for awhile although it is not really a latch.
_________________
“Most people don’t know what they’re doing, and a lot of them are really good at it.” - George Carlin

[Tcash]

First thank you, for posting what the fix was!
Just to clarify.
When you sweep the AFM flap open you should see a linear reading from 1.75-7.0v.
As you swept the arm closed, you noticed that it would go down from 7v to 3v in a linear fashion and then plummet to 1.5v for the last entire section at the bottom...right where you are cranking.
Thanks again
Tcash

[SGKent]

TCash and Canadianveedub -

The AFM is not a linear potentiometer with one voltage output. It is a potentiometer with multiple taps. Each of those taps sends a separate reading to the ECU.

The purpose of the AFM is:

1) tell the ECU when the bus engine is pulling enough air to turn on the fuel pump (contact points in the AFM)

2) tell the ECU the temperature of the incoming air (TS1)

3) tell the ECU how much air is being ingested by the engine (multiple taps)

4) tell the ECU how fast the engine is accelerating by how fast the sweep arm is moving (which the ECU will inject a little extra fuel like an accelerator pump does.)

Unless someone has the ability to measure all the signal voltages SIMULTANEOUSLY and compare them to the original specs by Bosch for each air flow range, (which no one can find) one cannot know if the ECU is truly working to spec or not. There are resistance measurement tests for for each of those taps (segments) in the Fuel Injection manual. In later models Bosch added a ground wire to the sweep arm after they discovered that corrosion in the sweep arm crimp interfered with correct readings sometimes. As they were built, a laser was used to tune each circuit board in 2 dimensions AFTER it was mounted and screwed down so it read the correct curve on each tap at different flows. One cannot measure one resistance or voltage and declare the AFM is good or bad. Nor can they move or replace the board without destroying the tuning Bosch did to each AFM. It is a bailing wire and bubble gum fix putting your engine heads at risk. Look at the manual for the correct tests. http://www.thesamba.com/vw/archives/manuals/afc_fi...Manual.pdf

Example of test:

_________________
“Most people don’t know what they’re doing, and a lot of them are really good at it.” - George Carlin

[canadianveedub]

[SGKent]

each one of the segments is tuned at the factory with a laser (see the black burn marks) so that the board follows the map Bosch had in mind when they designed it. There is more going on than what you profess you understand to be going on. Many people have monitored voltages thinking that as long as it went up and down they understood how it worked. I did the same thing way back and guess what - more is going on. All I can tell you is that as time passes you will discover that unless someone finds a schematic for the 75-78 ECU, and also the 79 - 82 ECU, it is unlikely anyone will ever completely understand how the board map works. For all either of us know, the circuit may also be monitoring current flow into that board as well.

_________________
“Most people don’t know what they’re doing, and a lot of them are really good at it.” - George Carlin

[VWCOOL]

Problem found, problem solved!

Good work and thanks for sharing

Does the injected Beetle use the same hardware?

[raygreenwood]

[canadianveedub]

There is a great desciption in the bosch l-jet technical manual on page 17.
http://www.google.ca/url?sa=t&rct=j&q=&...mp;cad=rja
I will agree that the voltage output is non-linear, it is actually logarithmic, however, the output that the ECU receives and uses for fueling calculations is a voltage, not reading simultaneous resistances on all of the pins.
We can get into all sorts of semantics about this, however the bottom line is that the output is a voltage that represents the flap angle. If this voltage output has dead spots or dropouts, it will not function correctly.

[Randy in Maine]

OK I will bite....

Why not just install a cat converter and use an O2 sensor to all the engine to lean out like it is supposed to?

[raygreenwood]

[SGKent]

[raygreenwood]