Calling MFI guru's, help needed

[motomiata]

First things first:
Yes, I have downloaded, read and followed the CMA
Yes, I have searched and read all relevant threads about MFI

Engine is a 72T MFI, 48.000 miles, to my knowledge never rebuild
The car has ben restored over the last 7 years and not driven
Compression is 140 psi plus/minus 4 on all 6
Cam timing checked and double checked
Valve adjustment done twice, just to be sure
Rebuild dizzy, new cap, new cables, new plugs, dwell and timing OK
Throttle bodies cleaned and checked, looks good with no signs of wear at the shafts
New linkage all over
Throttle plates adjusted on a bench using a flow meter, all flowing 4 kgs/hour at the stops.
New gaskets on throttle bodies and stacks
Linkage very carefully adjusted for correlation. All 6 flows the same at idle and 3000 rpm
MFI pump fully rebuild and calibrated by well reputed MFI guru in the UK
Injectors ultrasonic cleaned and checked, matched to the pump
Warm-up regulator replaced with a set screw
Fuel tank cleaned and POR15, new screen, new fuel filter, new lines, fuel pressure checked and OK

Sounds like the recipe for a smooth and well running engine, but no so.

The engine starts easily but runs very rich at idle and when revving above 2000 rpm. It idles nicely and very stable at 500 rpm and up to 1000. Increasing the throttle, the engine very rapidly goes very lean (AFR 15+), stumbles and misfires up the stacks. When the engine revs past ca. 2000 rpm it clears up and can be revved all the way to redline. Lifting the throttle, it seems as the engine hangs in the problematic rev range for a few seconds before dropping to idle.

If I enrich the pump using the set screw for the warm up regulator I can almost get rid of the lean transition, but the idle and top then gets even richer (around 10 AFR). Leaning out the main rack seems not to help either.

Here is a video which shows the AFR versus RPM: 2016-08-12 16.15.13 by viggoman, on Flickr

When pushing the outer control rack on the pump (the one pushed by the throttle linkage), the main rack inside the pump only moves approx. 3mm as seen in this video:
a - 1 by viggoman, on Flickr

Does this look right?

And when the outer rack is released (equal to lifting the throttle completely) it does not return to stop on it's own, but needs a push (the return spring in the throttle linkage do return the pump to the stop when fully installed)
a - 1 (1) by viggoman, on Flickr

Is this OK?

Unfortunately we only have one MFI guru here in DK and he has a 4 month waiting list

Please chime in with suggestions for how to crack this one.

Thanks

[matthewb0051]

Take further look at the injectors. I've found that simply doing an ultrasonic clean does not completely resolve any issues. At the end, there is a little spring that allows the valve like piece to move in/out and atomize the fuel. There can be carbon build up around the tip and/or corrosion around the spring.

Unless you have seen the spray pattern from your injectors, I would not claim that they are good.

Go online and watch the Wheeler Dealers episode with the Mercedes 560SL from this most recent season. Edd used a tool that looks like a bottle jack that can pressurize fluid similar to that your injector will use. I bought one of those tools and was blown away how bad my injectors were.

Wheeler Dealers S13 E01 HD Mercedes 560SL 1986 - Video Dailymotion

The tool is made by: www.mercedessource.com



But by the time you have bought the tool to test you could have paid for at least 1 new injector. It is a true cost benefit analysis

[manbridge 74]

Fuel pressure and flow checked from rebuilt pump?

[motomiata]

Pressure and flow checked.

The injectors have been flow tested and spray patterns checked as well.

[ant7]

Re-check pump, and throttle body alignment.
A...

[ed mayo]

Pump control rod set to 114 mm ? All other linkage is then set for no preload relative to the pump control rod.

[kent olsen]

Email Mark Jung at 356rs@peak.org
He's our west coast MFI expert

[bkreigsr]

Just checking...do you have the thermostat hoses connected to the left heat exchanger?
Your symptoms sound like mine when I thought I could run without them (inner+outer)
Bill K

[motomiata]

Thanks for the replies and suggestions.

It turns out I do have ignition problems, even though I'm on the 3rd dizzy. A thorough check with the timing light revealed consistent lack of spark on no. 5. The unburnt oxygen dumped in the exhaust makes the sensor report a lean situation.

The problematic rev range is exactly where the ignition must advance the most in the shortest rpm interval.

I will get the dizzy rebuild and bench tested, re-tune the MFI pump an report back.

Lesson learnt: never EVER deviate from CMA!!

[ed mayo]

I doubt if three distributors would all create a problem on the same cylinder. Check the resistance of the entire spark plug lead from cap tower to spark plug connector.

[motomiata]

Coming back to this thread after exactly one year. Work, moving house and life just didn't allow enough time to crack my problems with the 2.4T. A defect CDI box soaked up a lot of time.

Believing I was chasing an ignition related issue I have ended up buying a Classic Retrofit CDI replacement box and a 123ignition electronic dizzy. This combo seems to work just fine.

So to recap:
MFI pump has been rebuild and calibrated, injectors cleaned and tested with the pump.
Throttle bodies checked, flow adjusted, no signs of leaks
Compression good, but no leak down test yet
All ignition components new from dizzy to spark plugs
Warm up regulator bypassed with set screw
(please see the 1st post for further details)

Problem remains

It starts fine and idles smooth and rock steady at 700 rpm, but as soon as I increase the throttle it stumbles and misfires both up the stacks and out the exhaust. Revving to above 2000 rpms clears up things and it revs happily above this.

I can tune up he main rack to an 13 AFR at idle, and it shows lean when the stumble appears ( this may be due to the misfires). It goes very rich above 2000 rpm around 10 AFR. Leaning out on the main rack further kills the idle. And enriching idle does not change the stumble.

I'm lost as I have tried all the obvious chasing down my problem. I'm pretty convinced it's not ignition related. Speed switch and stop solenoid has been bypassed as well.

Any suggestions for how to proceed is highly appreciated, thanks

[tobluforu]

You have done more than most shade tree mechanics would have done. I'm guessing that since you are really thorough that you have timed the pump? What happens when you plug the dizzy vacuum hose? Have you deviated in regards to timing, in other words, have advanced/retarded timing to see how it reacts? Is the fuel pump new/rebuilt? Is there a reason why you went with the set screw method? Are all of the fuel lines new?
By the way, what cams are in it?

[motomiata]

Thanks for your reply.

Yes, I have really put a lot of efforts and money in to cracking this nut.

Pump is timed, ignition timing checked and looking correct, I have even double checked the timing mark on the pulley the is at TDC.

I haven't checked pulling the vacuum line or tried fiddling with the ignition timing.

Fuel pressure to the MFI pump has been checked, all fuel lines except the ones in the tunnel has been replaced.

I assume the cams are original but don't know for sure.

I have a Megasquirt lying around and will try hooking this up to log rpms, MAP, AFR and the position of the MFI lever. Having some logs showing the correlation between those data might be valuable.

[stownsen914]

Sorry to beat a dead horse, but re: fuel pressure, just to confirm ... you checked pressure inline and got a 1 atm (~15 psi)? Ideally checked with engine running to ensure no pressure drop. There's a spring check valve in the filter housing that is designed to maintain pressure, and the spring gets weak over time. The MFI system doesn't like low pressure ...

[motomiata]

@stowsen914

Yes, fuel pressure has been tested good with engine running.

The car is drivable once I get passed the troublesome rpm range. I assume a too low fuel pressure to the MFI pump will show up increasingly with speed and load?

[manbridge 74]

I would start with leak down to be sure that all cylinders are contributing equally.

[stownsen914]

Originally Posted by motomiata

@stowsen914
I assume a too low fuel pressure to the MFI pump will show up increasingly with speed and load?

In my experience, low fuel pressure just made the car run like crap at all RPMs and different loads. If you got 1 atm with engine running, it's probably fine. You can separately test the electric fuel pump for proper volume, which you may have done already (I don't recall if you mentioned it above).

Scott

[kent olsen]

Contact Mark Jung at 356rs@peak.org

He is the US west coast expert on MFI systems.

Kent Olsen, my family came from Denmark, cool.

[ed mayo]

I didn't see if you ever mentioned adjusting the pump control rod to 114 mm center to center. Since all linkage was done I suspect its off.

[matthewb0051]

I had the exact same symptoms and they cleared when I got new injectors. I cleaned the injectors and thought they were good, but at 40+ years of age they were past it.

Do yourself a favor and get 6 new injectors.

.