CIS System - Testing and Adjusting

by Peter Kurzenhauser

Date: Wed, 17 Jan 2001 17:40:40 -0500
Subject: [MB] Info on testing and adjusting CIS systems, save for reference

Listers, Lurkers, and other denizens: Over the past couple months I've been experimenting with the CIS system in my '87 190E-16V. It was running unevenly from time to time, and using Techron, or several tanks of Chevron fuel (which has Techron additive), generally resulted in improvements in drivability. After checking other systems and sensors, I narrowed down to the fuel system, particularly the fuel distributor and injectors.

The results of my tinkering and testing are applicable to all CIS systems (most gas cars before 1993)

Earlier last year I had kludged together a fuel delivery test apparatus which I used to check the fuel delivery rates and injector patterns. For those of you familiar with the Bosch fuel delivery tester, it is an electro-hydraulic device that measures fuel flow in cc's per minute or some similar units. However, it is expensive and not readily available. I went to my local MB shop and 16V racer/expert (Rick Ellinger's RC Imports) and wheedled some leftover steel fuel lines from him (the ones that go from the fuel distributor to the injectors). I straightened and bent them to get a set of four that screwed onto the top of the fuel distributor, went up about 6 inches, then hooked down, in a line of four, so I could screw the injectors onto the ends pointed down. I then got 4 test tubes of 15 cc capacity and drilled 4 holes in a block of wood to hold them aligned and spaced with the fuel injectors. This way, I could collect the output of the injectors into the test tubes and measure how much was delivered in a period of time, or (what's really important) how much fuel was delivered through each injector relative to the others.

To use the setup, I put a jumper into the fuel pump relay socket to run the pump, and pressed down on the air flow plate to replicate idle, part-throttle, and WOT fuel flows. By swapping injectors to different fuel lines, and testing at different flow rates, I determined there was almost 10% variation in the fuel delivery to the cylinders. Swapping the injectors did not make an appreciable difference in fuel flow, which is to be expected, because the differential pressure valves should minimize the effects of variations in injector opening pressures. Therefore, variations in fuel flow were isolated to differences in the 4 circuits of the fuel distributor. By testing different air sensor openings, I determined that the differences in fuel flow were significantly the same for both low, medium, and high flows (idle and WOT and in between). That is to say, if one cylinder was rich at idle, it was also rich throughout its range.

The Mercedes Manual for the 16V and the regular 190E say that up to 10% variation in fuel delivery is within tolerance. I have a hard time with this spec, because a 10% variation in the resulting mixture makes a significant difference in power and emissions. For example, if the leanest cylinder is at 15.4:1 (lambda=1.05), it may miss or have an incomplete combustion occasionally, especially when cold. It will also make less power than stoichiometric (14.7:1) and have higher NOX. If the richest cylinder is at 13.9:1 (lambda = .95) then it will run smoother, and make more power and be more fuel efficient, although it will have higher CO and HC. This info, by the way, comes from Bosch's own technical books and charts, in addition to other sources.

I figured that less than 5% variation was a much better figure to shoot for.

One of the other things I borrowed from Rick was a spare fuel distributor, which was gummed up (he was saving it as a core for a rebuild). I took it mostly apart (I did not split the two halves, but removed everything else) and cleaned it out well with Gumout spray and soaking, and got it working again. I carefully measured the depth of the gland nut around the plunger and the depths of the 4 adjusting screws (see below) in order to put it back together with the same adjustments. I put it on my car and drove it for a week to be sure it was working. Then I started fiddling with it.

The manual does not indicate how the fuel distributors are adjusted, it only says to replace them if the flow rates are out of spec, but it's actually pretty simple. ON the bottom of the distributor are four 3mm Allen screw caps. Under the caps are 2.5mm adjustment screws. These screws adjust the spring pressure against the differential pressure valve diaphragms, and thus the differential pressures, and thus the fuel flows. I found by testing with the delivery apparatus that a 1/4 turn in leaned the mixture on that circuit about 5%. By testing and adjusting a half-dozen times, I got the flow rates within 5%. I drove the car again with the adjustments, and it ran a little better.

Having proved out the procedure with the spare fuel distributor, I applied it to my original distributor and tweaked it to within 5%. Car ran a little better afterwards, specifically, smoother power on very light throttle and low speed operation. It also passed emissions inspection for the first time!!!

This apparatus and procedure will apply to any CIS car. Therefore, as I said in the subject line, save this for reference, unless you have something better, in which case share the wealth with the rest of us!

Pete K.