Steve (specifically) or other CIS injection experts…(I hate to beg, but Pleeease….)

Well speaking of needing theoretical & experiential help…

I will cut this as short as possible. 1980 450, M117-985 engine, U.S. (California) specs. First year of Lambda controlled CIS injection. I am having problems with one cylinder (#5) running way too rich. All mechanical & ignition systems are O.K. after thorough & comprehensive checking. Fuel system pressure & control pressures are O.K. The injector itself is not the cause; I mickey-moused a test rig using a fuel pump & pressure gauges – all injectors are within specs for opening pressures, no leakage, and they have a good spray pattern. (I used mineral spirits as the test fluid – I don’t like the idea of flash fires ).

I have concluded (correctly, I hope) that the fuel distributor is failing. Seems odd, as the fuel filter has been changed religiously. Although I know that no one is supposed to touch the insides of these fuel distributors. (No one other than specially trained elves that live deep inside the black forest of Germany, that is ) I removed the center plunger to check for gunk (none present) and I cleaned the plunger & innards of the distributor with carb cleaner. Upon reinstallation, the same exact problem. No better, no worse.

Now for the tricky part… I can’t help but try to rectify the situation, hopefully without burning a valve in the process. From what I can see in the drawings in the “Bosch Fuel Injection & Engine Management” book (Great book, by the way!) underneath screw-on caps adjacent to each outlet port, there are adjustment screws. These screws bear down on the springs within the individual pressure regulators governing the pressure drop for each port. Theoretically, it looks that by backing off the setting, the spring pressure will be reduced, and therefore reduce fuel flow for that particular port. The setting for this regulator is a pressure drop of only 0.1 bar, and I’m sure that it is CRITICALLY important to get it right. I am tempted to try to see if I can lean out that cylinder. Am I playing with fire? I have nothing to lose if it does not work, as I would need to purchase a fuel distributor anyway. My plan of attack is to remove power from the lambda valve, reset the mixture (temporarily) to the highest & best idle. I would then adjust #5 until it stopped missing and reached the highest & best idle (again), then reattach the lambda valve and reset the mixture to specs.

Steve, please tell me your honest opinion on this - Am I TOTALLY insane? Or does this have a snowballs chance in h*** of working? Or better yet, do you know of a precise way to set the pressure drop, where I can reset all the ports to spec?

A MILLION THANKS!

[stevebfl]

I know you aren't supposed to mess with those adjustments and I have only done it a couple times. Both times the situation was as is yours; couldn't make it worse.

The one that I had the best results was on a Ferrari (someone had done as you are saying and really screwed it up). It has about the same injection. The biggest difference was that the car has individual exhaust taps so that exhaust gas readings can be had for each cylinder.

Your method for adjusting might not be sensitive enough and might need some refining. First I wouldn't disconnect the frequency valve (if that is what was meant by the "Lambda valve"). Doing this alters the differential pressures in a way that wouldn't happen in actual mixture control. What you should do is disconnect the Oxygen Sensor so that the system will quit trying to correct your mixture corrections.

Since the motor actually runs better with excessively rich mixtures. You are going to be left with a huge mixture window to shoot for. I would suggest a couple other techniques. Unfortunately they require exotic special tools; the best of which virtually no one has anymore.

The best method would be to find someone with a differential flow meter. This tool is pictured in the MB shop manual but I don't think any MB dealer has one. It was a required tool of all Bosch Service Centers so you might find one there. We of course have one and that is the way I set the other couple of distributors I have reset. This tool hooks to the injectors sort of like a milking machine works on a cow. The fuel pump is energized and the airflow meter plate is deflected by a special screw feature that holds the deflection at three different flow rates 10cc/minute, 35cc/min, and 60cc/min. These represent approx idle, midrange and full throttle conditions. Each injector and fuel distributor circuit is measured and must be within 10% different at each flow rate.

Doing this set-up using this tool indentifies the ability to correct. Often the differential between cylinder to cylinder is due to pressure losses across the stainless diagphram. In these cases adjustment won't work as the leaking affects the situation more at low flow rates. Thus if you alter idle, you may really screw the other settings. This will become apparent when testing the other rates.

If this tool isn't available, your scheme might work if you can use an exhaust gas analyser. Unfortunately you will not be able to test the results under load (without a Dyno - got one of those too) and as such you won't know how you have affected this condition till another sign comes up. Luckily that motor can't hardly be hurt by lean mixtures in the short range unless you race it. If that sparkplug gets burned up - lean, I would get that fuel dist. replaced.

Steve,

Much MUCH appreciated. I am really pleased that I am on the right track- since you have done this before, although in a much more controlled manner.

When I said that I would disconnect the frequency valve, I did mean electrically, not physically, so we are on the same wavelength on that; stopping active mixture control. I wish that I could check the CO of each cylinder individually, but…oh well.

About your ‘differential flow analyzer’ – Maybe I can obtain 8 graduated beakers and perform that test. I will not have the special screw to hold the air flow meter in place, but I can mickey-mouse something to function in the same manner. I would then time the flow with a stopwatch to obtain approximate flow rates of 10, 35 & 60 cc per minute. If I understand you correctly, if the 10% differential at ALL of these different flow rates cannot be maintained, then the fuel distributor is beyond adjusment?

And out of curiosity, in your experience, what may have caused the calibration to drift in the first place? Or is it just a matter of time & mileage? (about 250,000 miles so far, and still plenty more to go )

Thanks again!

[stevebfl]

When I suggested not disconnecting the frequency valve, I meant electrically! Do not disconnect it electrically. Instead disconnect the Oxygen sensor (under the pass side floor mat). The system needs the steady duty-cycle based fuel leak that the frequency valve gives to work properly. The fuel distributor will not work like a simple K-Jet distributor by electrically removing the freq valve. The simple K-jet dist has a preset differential pressure. The K-Jet w/Lambda dist has almost no differential pressure difference without an operating frequency valve. Try disconnecting and driving at any mixture adjustment. It just won't go!

By removing the oxygen sensor you achieve no electronic mixture control.

"If I understand you correctly, if the 10% differential at ALL of these different flow rates cannot be maintained, then the fuel distributor is beyond adjusment? "

Yes. If you measure the flow rates at three different rates (the numbers aren't as important as the relationship between each injector circuit) the relative differences from injector to injector should stay within 10% in each range. It is much more important at idle. Of the three numbers the idle spec of 10cc/minute is the most important.

Steve,

I hope you had a merry Christmas, and I once again, want to thank you for your help!

You are absolutely correct in that I was not cognizant of the need for the default cycling provided by the controller to the frequency valve when there is no output from the O2 sensor. Learning, learning, learning…. I’ll be off to purchase some calibrated flasks this week.

A THOUSAND AND ONE THANK YOUs - Have a happy & healthy New Year!