Need some help: FI hard start, dies on idle

[kooper271]

[vwwestyman]

[busdaddy]

The spoonfull of gas trick will expose a no fuel condition, if it's not getting any fuel at all it'll run a few seconds, if it's getting fuel but too lean to start it'll start almost instantly.

A small squirt bottle works well for this test, pop a breather hose out of the side of the S boot and give the bottle a squeeze so ~ a tablespoon (15ml) of gas gets into the plenum, then replace the hose and attempt to start it.
A trigger sprayer also works, or even a length of fuel or vacuum line dipped deep into a gas can, seal the other end with your thumb and release it when you have it aimed into the boot.

If it makes it even harder to start it's likely you have a rich condition or bad spark.
_________________
Rust NEVER sleeps and stock never goes out of style.

Please don't PM technical questions, ask your problem in public so everyone can play along. If you think it's too stupid post it here
Stop dead photo links! Post your photos to The Samba Gallery!

Слава Україні!

[furgo]

Perfect, thanks for the clarification. I now understand much better the spoonful of gas trick. So as you said, with it I can determine one of the 3 conditions:

1. No fuel (I don't think it applies here, as I could eventually start the engine and get it running for a bit)
2. Lean
3. Rich

Say if I manage to determine if it's lean or rich, what can I do with this information?

I could think for instance it would help me determine which direction I should turn the idle speed screw to, correct? What else I should be looking at, armed with this knowledge?

One more related question:

[Wildthings]

[busdaddy]

[furgo]

Thanks. Ok, so to recap the tests in my head:

• Grounding TS2 test => leans the mixture
• Unplug (manifold?) vacuum line => enrichens the mixture
• Spoonful of gas test => enrichens the mixture

Some further questions on the fuel pressure regulator to see if I can rule it out:

• Given that a) while I was manually actuating the pump, the pressure was 2.7 bar solid (39 psi), b) there was no pressure build up, and c) it's rated at 2.5 bar, I assume that it's doing its job in terms of regulating the pressure, correct?
• In terms of its vacuum function: how much vacuum should I apply for it to return fuel to the tank? As mentioned before, I applied 15 inHg without apparent drop of pressure (and the vacuum held). I can go up to 30 inHg with my manual vacuum pump, but I wasn't too sure if that was too much and if I could damage the membrane.

Or otherwise, how should I be testing the FPR?
_________________
'79 Westy, P22 interior, FI 2.0 l Federal, GE engine (hydraulic lifters)

Decode your M-Plate

[Wildthings]

[SGKent]

If you haven't done it already please glance at the two pages 10 & 20

https://www.thesamba.com/vw/archives/manuals/afc_f...Manual.pdf
_________________
“Most people don’t know what they’re doing, and a lot of them are really good at it.” - George Carlin

[furgo]

[vwwestyman]

There's a lot of info on the forums about building a smoke tester to help find vacuum leaks.
_________________
Dave Cook

President, Wild Westerner Club

1978 Champagne Edition Westy, repowered to '97 Jetta TDI
1973 Wild Westerner
My Thing

[furgo]

I admit defeat. For now. It'll have to be plan B for holidays this year :/

But I didn't go down without a fight, and I'll be back to it after vacation.

TL;DR:

• I narrowed down the symptoms: hard (cold) start and rpms creeping down until the engine dies
• I think they're two separate issues, the decreasing RPMs being the one to worry about. I've now learnt to temporarily work around the hard coldstart by half-depressing the gas pedal while cranking. I found some references with others having a similar issue. Unfortunately, I could not find a report of whether it was ever fixed.
• I did the spoonful of gas, TS2 grounding and vacuum line removal tests. See below for the results, but none of them made the situation worse. Removing the vacuum line seemed to make the engine happier by slightly increasing the idle RPMs
• Timing is off by 5.5°: it's 2° BTDC when it should be 7.5 BTDC
• Contact breaker point gap is narrower than spec: ca 0.15 mm vs the expected 0.4 mm

Pics

The previous pictures of the fuel line replacement seemed to confuse some folks as connectors and other bits and pieces were missing at the time of taking them. Here are some pics of what the engine actually looks like:

Initial test

I wanted to have a baseline to compare against, so I tried to just start the bus first. I also connected my multimeter to the engine speed signal (coil -) that goes to my analog tacho, as the digital multimeter has a better resolution (10 rpm vs. 200 rpm). I could then see a more accurate figure for rpm and dwell. So here's what I did:

1. First start with gas pedal half depressed. Went well, not too much cranking or too hard a start.
2. Revs idled to ca. 800 rpm for a bit. A bit on the low end, but still within spec.
3. After a minute or two revs started creeping down to 700 rpm and gradually to 600 rpm. Then the engine died

I also noticed there that a steady foot on the gas pedal helps keeping the engine alive, but sudden acceleration results into a quick engine death once the foot is off the pedal and rpms go down.

4. Second start, without depressing the gas pedal. Much rougher, more cranking was required.
5. Same as before, after a minute or two revs start creeping down. You can feel the engine struggling when it gets down to 740-760 rpm. It all goes downhill from there.

6. Third start much rougher, I couldn't get the engine to start and gave up for the time being.
7. Checked the battery: still 12.6 V

As a note, I did hear a loud hissing sound while the engine was running, but I believe that's normal.

Spoonful of gas test

This one is supposed to enrichen the mixture by adding extra fuel.

I used this:

To put some gas in here:

I essentially unplugged the crankcase breather hose and then squirted the contents of the pipe twice into the S-Boot.

Results:
• It seemed to start better as before, in any case, not worse
• Revs around 800 rpm, but still creeping down after a minute or two
• I happened to look at the fuel pressure while running: 2.5 bar, higher than the usual 2.3 bar I had measured and higher than spec with engine running

Vacuum line removal test

This one is supposed to lean the mixture by creating a vacuum leak. I figured it would be better to use a manifold vacuum line, so I unplugged the decel valve's vacuum control pipe:

Results:
• It started and ran much better
• Higher revs: around 900 rpm, engine gives a much more confident feeling
• It holds the revs up for longer than before, but eventually they will creep down to the critical 740 rpm point where the engine starts to gasp for breath.

TS2 grounding test

This one is supposed to lean the mixture by bringing the ECU into thinking that the engine is warm and does not require extra fuel for coldstart.

I grounded the ECU wire (terminal 13) to the free connector on the FI star ground below the 3/4 intake runners (the purple wire on the picture).

Results:
• It also started better (no need to depress gas pedal)
• Revs slightly below 900 rpm
• But they creep down quicker than in the vacuum line removal test

Test conclusions

• Revs creeping down seems to consistently happen in all tests. To me, that's the critical part. The hard start not so much.
• In all tests a sudden acceleration has the effect of revs going up as expected, but afterwards going down much quicker than gradual changes in acceleration. Most of the times it leads to revs going as down as to kill the engine.
• None of the tests apparently made things worse, which makes it difficult to tell the state of the mixture. As such, I'm not sure if they're too conclusive.
• Regarding the improvements in starting, it might just have been that the motor was warm already when I started the tests. This engine has chronically not started well cold (but also not so roughly as now), but it has always started well when warm. So any improvements to cold start might have gone unnoticed.
• The vacuum line removal test seemed to give the best results. Does it then mean that the mixture is too rich?

Measurements

• TS2 resistance at the ECU connector, pin 13: 550 Ohm (ok, the motor was warm)
• Double relay and series resistor connector pins ok (e.g. not pushed out)
• ECU connector pins ok (no pushed-in pins)
• Dwell over 50°, i.e. longer than spec. It was ca. 53°, but difficult to tell, as the values in the meter were a bit jumpy.

Timing

It was the first time for me to do this at all, so I tried static (test lamp) and dynamic (strobe light) timing. I only did the measurements, I did not do any adjustments at this point.

Note to self: remember to turn the ignition on before wondering why the test light does not turn on

This one puzzled me a bit, I might have not done it the right way round (CW or CCW)

The first time, turning the engine CW, the light was turned on at 4° ATDC:

When turning the engine CCW, the light stayed on until reaching ~1° BTDC

I then did the dynamic measurement with the stroboscopic light, and the timing was at 2° BTDC. That means:

• Nearly same measurement as CCW static timing
• 5.5° off from the expected 7.5° BTDC

The strobe light did not turn on at times. I'm not sure if due to a crappy lamp (it was the first time I used it) or if this had something to do with an uneven spark.

Contact breaker points

As mentioned before, the dwell was ca. 53°, outside specs (47°+/- 3°). The breaker points were quite tight: 0.15 mm instead of the expected 0.40 mm. However, one thing that I just realized later was that the points were slightly misaligned: I was expecting them to be more or less parallel when open (I double-checked this with a set of new points), but the opening was slightly larger at one end:

Time to replace the points, I guess.

While I was in there, I cleaned up the rotor contacts and added 3 drops of engine oil to the pad at the top of the shaft. I actually have an overhauled distributor at home I want to replace the one on the bus with as a spare, but I didn't get to swap them yet.

What next

• I'm wondering if I should adjust the points and timing at this point, or if I should just leave as they are to keep the baseline until I've figured out the actual problem. I assume the advanced timing does not have an effect on the decreasing rpms
• Do the 5.5° additional advance have got a significant impact on the ignition timing? I'm asking as I don't have a feel for which type of figures start having a noticeable effect.
• Any ideas on possible causes for the decreasing revs?
_________________
'79 Westy, P22 interior, FI 2.0 l Federal, GE engine (hydraulic lifters)

Decode your M-Plate

[Wildthings]

Get the points set correctly to 0.16-0.20". They can't have any little tits if you are setting them with feeler gauges, you need either new points or to file the tits off the old ones. Now set the timing to 7.5° BTDC when turning the engine CW. It should now be very close to being properly timed. Having your timing set 9-10° too far retarded is going to cause starting and running problems.

If you buy new points DO NOT BUY BOSCH, but go to NAPA and buy Echlin. Make sure you lube the wear block and cam for the points with the supplied lube.

[hazetguy]

replace the points and set them to the correct dwell. i almost never use feeler gauges to set the gap, and almost always a dwell meter, points set to the lower end of dwell spec, because dwell will increase as the rubbing block wears. then set correct timing.
also make sure your vacuum canister is working (no leaks) and that the points plate is actually moving freely and not gummed up.
_________________
thebucket: I invested in hoodride, now DBD won't return my call?
hazetguy: invested?
thebucket: Yeah Haze, its where people put money into a company in hopes of a return on their money

[furgo]

[vwwestyman]

The reason that you don't want the Bosch points anymore is because the rubbing block is an L shape piece of plastic that is riveted to the points arm. The plastic is prone to breaking-I had a brand new one break within maybe 5 hours of install in my '73. I had forgotten that I had the exact same issue with points in my '78 6ish years prior TWICE on the same trip. (First pair broke, and I installed the spare pare and it broke in short order.)

The ones I got from Autozone (whatever their house brand is-Duralast I think) in both instances (coincidentally both from the same store in Nebraska, as I was near it both times) had a pyramid shaped rubbing block that can't break in the same manner.
_________________
Dave Cook

President, Wild Westerner Club

1978 Champagne Edition Westy, repowered to '97 Jetta TDI
1973 Wild Westerner
My Thing

[furgo]

[Wildthings]

[airschooled]

[raygreenwood]

[quote="asiab3"]