MFI Pump - Open Heart Surgery

[jeffc280sl]

I own a 1970 Mercedes 280SL with a similar Bosch fuel injection pump. I was inspired by Jeff Higgins to buy an LM-2 and tune my fip. It was tedious but my car now meets the MB on road afr specs. The on road tests require the engine to be under load so I added a MAP sensor to the intake manifold and used one of the analog inputs to the LM-2. I also added a throttle position sensor and accelerometer to the mix. With the accelerometer I can calculate rear wheel horse power using a formula in the Data Logger software that came with the LM-2.

Since Jeff was the pioneer in this area and his work was invaluable to me I thought I would share some information related to the barometric compensator (BC). Inside the BC canister there is a brass bellows and a pin with a spring. Inside the bellows there is a stiff spring.







The bellows is a sealed chamber under partial vacuum. The spring inside the bellows and vacuum contend with each other. As you increase altitude the vacuum force inside the bellows is reduced and the spring force expands the bellows. The expanded bellows forces the pin outward and this change leans the rack and afr.

The Bosch service manual for the R 11 fuel injection pump tells us the following:

4.16 Checking the altitude compensator

100 mm HG changing (vacuum) = 0,58 mm stroke of plunger

First you have to mount the altitude compensator on the Bosch testing Equipment number EFEP 418 with gauge EFAW 7 and o ring seals.
Adjust 20 mm HG vacuum.
Adjust the scale of the gauge to 0.
Increase the vacuum to 200 mm HG.
Dilation of the altitude compensator plunger now should be 1,00 - 1,25 mm. If not, it does not work correctly.

4.17 Mounting the altitude compensator

Calculate the difference between average atmospheric pressure of 737 mm HG and the real atmospheric pressure you have at your place.
For instance:
Average value for atmospheric pressure is 737 mm HG
Real pressure at your garage is 600 mm HG

Difference therefore 137 mm HG

The formula says : 0,58 x 137 / 100 = 0,79 mm stroke of plunger

Now mount the bosch testing equipment Nr. EFEP 417 with the gauge EFAW 7 on the injection pump altitude compensator area
Adjust 0 on the scale of the gauge, when you have about 3 mm pretension.
Alter the knurled screw on the injection pump for 0,79 mm. (you have calculated)
Now you have to notice how many mm is the movement of the control rack and notice the fuel delivery rate at full throttle.
(for this procedure the injection pump has to be mounted in a bosch injection pump test stand)

Now you have to mount the altitude compensator instead of the testing equipment Nr. EFEP 417 on the injection pump.
Now add shims to get the same mm movement of the control rack and the same fuel delivery rate.

I was able to build a vacuum chamber in a Ball jar to test my BC. Here is what the setup looks like.




Here are the results of the Bosch vacuum chamber tests described in 4.16 above.




I think it’s safe to say the failure mode of the BC is a loss of vacuum seal in the bellows. When this happens the bellows is expanded by the spring inside and the pin extends to max length. Here is a side by side pic of a good and bad BC.



There are at least 2 versions of the BC. One with a collar at the end of the threaded section and one without. Here is a pic of a BC without a collar.




Because of the version differences I will provide some measurements from the top of flat section of the hex head bolt to the end of the pin. A good BC will be about 21mm in length and a bad almost 25mm.


Thanks again Jeff

[356RS]

Great information and thanks for sharing this. I was also trying to figure out the best way to test the BC. I built (over built) this vacuum chamber for testing the BC plunger movement at different simulated altitude changes.
And Yes, a big thank you to Jeff for getting the Open Heart Surgery started. All kinds of great information.

[304065]

Mark, you need an aircraft altimeter mounted flush with the plexiglass with a gasket!

[356RS]

Exactly whats on order John. Altimeter from a friend with a small airplane and spare parts. Plexiglass seal is a square o-ring. Vacuum pump is another over-kill but was cheap.

[jeffc280sl]

You're welcome.

I really like your setup. The data you can capture may help people shim their BC properly. Maybe I can shed some light on the expected movement of the pin.

At rest your vacuum chamber pressure is equal to the atmospheric pressure at your altitude. Here is a website that calculates altitude based on pressure readings.

Pressure Altitude Calculator

One needs to determine the pressure at their altitude. That pressure becomes zero on your vacuum guage. Let's use sea level for this example. Sea level is 29.92 inHg. As you pull a vacuum, let's use 4.08 inHg, you have reduced your starting pressure by that amount. So at 25.84 inHg inside your chamber you are simulating an altitude of 4000ft.

The Bosch test tells us the pin moves .58 mm per 100mmHg change in pressure. Let's convert these numbers to inches and inHg. This site has a pressure conversion calculator.

Convert in Hg to mm Hg - Conversion of Measurement Units

After the conversion we know the pin moves .0228 inches per 3.94 inHg.
The Bosch test tells us to establish a .79inHg vacuum. At this point zero your dial indicator. Take the vacuum chamber up to 8.66 inHg. The pin should extend .039 to .049 inches if you have a good BC.

BTW 29.92 inHg less .79 inHg is equivalent to the pressure at 834 feet and
29.92inHg less 8.66 inHg is equivalent to the pressure at 9152 feet.

According to my calculations the pin should move .003 inches per 500 ft elevation change.

[jcge]

Nice addition to this thread Jeff - thanks for posting

Particularly useful is your assessment of the failure mode for the Baro cell - a simple visual assesment that anyone can make with the reference pics you posted. Excellent work.

My local Bureau of Meteorology post the ambient pressure on their website every 15 minutes, measured using better equipment than I'll ever have...I take their pressure/elevation and correct for my local elevation.

Regards
John

[jeffc280sl]

John,

You're welcome. I've found a MB fip with a failed BC can be tuned to meet afr specs. I tuned my fip before I understood the BC failure mode. If your BC has failed I recommend you install a bolt and lock washer in its place and over time adjust the bolt up or down for best performance. Should you operate your car at other altitudes a manual adjustment can be made.

If you plan to drive at varying altitudes a replacement BC is a better way to go. At this time the shim selection process used in the factory setup is still a mystery. This means that you will have to add or remove shims for best performance.

If the trial and error tuning method is not to your liking you can follow Jeff Higgins tuning process with a afr meter.

[dicklague]

Thanks for this great thread....I finally go to studying it. I am working on my MFI now that I have a great data recording A/F instrument. MFI has always been a bit of a mystery and i have never been able to get my hands on enough information. I have to continue digesting this whole thread.

Dick

[jeffc280sl]

The fip is so complex I think it will always be mysterious. Does your test spec include an engine load. Are you going to install a map sensor?

JeffC

[dicklague]

Yes, I am installing a MAP sensor. I just got the WeatherPak connector today and I need to wire it up. Currently I have a wideband sensor with RPM being fed to it, so the data charts have RPM and AFR. Of course adding MAP will give me load as well.

I will post some photos and charts. I have them on my other computer. So far I have managed to get the idle mixture a lot better and the car idles a lot better.

I have a 1973 2.7 engine RS clone. Here it is:


I am using a Daytona Sensors Wego IV AFR instrument. It gathers 1.5 hours of data, and AFR+RPM and one more input.....MAP, or TPI anything 0 to 5 vdc. I like it so far because the complete unit with data logging is $440.00. You don't have to buy a lot of extra stuff.





I have not used the LM1 or LM2 so all I know about them is what I read. I am happy with the Daytona in that it is a pretty good rolling dyno, and data logging is absolutely necessary to see where you have been, and study what is going on at varying RPM and loads.

Will keep you posted.

[jeffc280sl]

As you have read the Bosch fip has several adjustments. MB provides an on road dyno test spec with factory afr's. There are two partial load tests and a full load spec. Measurements are taken in 3rd gear with a 300mmHg load at 1500 and 3000 rpms.

Full load test data is taken in third gear starting at 3000 to x rpms.

There are co specs for each of these tests and this is what i tuned my fip to meet.

Do you have a similar on road dyno type test spec you will try and meet with your afr meter?

Nice car!

[dicklague]

No I don't have a similar test. Do you have one? Is it available?

So far I am just setting a base line to see what I have. I have leaned out the idle a couple of clicks and that has been good.

I would be very interested in have a test routine for on road testing. Any help would be greatly appreciated. I should have the MAP sensor running tomorrow. I had to build a power supply to drop 12vdc to 5vdc to feed the MAP sensor. I then needed to get a 3 pin weather pak connector to fit the GM 1 bar map sensor which I bought used. I just have to finish the wiring and test it out.

[jeffc280sl]

Here is the dyno test spec and its on road test companion. This is for a similar Bosch fip. Porsche must have a similar test spec. Someone on this site must know of one.



As you know the white, black and rack screws afr settings interact with each other. Balaning a change here with another there can bring you inline with this spec. Its a bit of trial and error with lots of data taking and result comparisons. Make sure everything else is in good shape before adjusting the pump. Linkage, fuel pressure, ignition, etc

[jeffc280sl]

You can find tables or calculators on the web to convert CO to AFR and mmHg to inHg.

[dicklague]

Thanks for the information. I did download a co to AFR table so I can convert that.

thanks for the MB table. I may have the Porsche information in the MFI setup brochure I have and I think it may be on this forum somewhere. I will look, but if anyone knows where it is......

[jeffc280sl]

I've not read anything in this thread about an on road dyno test. Something similar must exist. The dyno test stand spec has hp load factors for the lower and mid rpm ranges.

[dicklague]

I will have to look for my MFI brochure. Will let you know what I find.

Dick

[jeffc280sl]

I found emission specs for part load tests at 2500 rpms in second gear. It seems you set the throttle linkage at a specific stop. I suppose the engine load is determined by the drivers ability to maintain 2500 rpms.

[dicklague]

I found some data in the very complete MFI thread....for example:

[jeffc280sl]

Good information and it answers the idle and partial load spec questions. I think we are still missing a full load test specification.