Question on MFI tuning on own build MFI test bench
 

Hi!
After finishing the full restoration of my ‘69 911T she is performing beautifuly ! Now a few years later I decided to build a more powerful engine and keep the matching number engine as it is.
Slowly I am collecting parts for this new engine. It should run with this fantastic MFI. I have restored the whole MFI setup. Bored throttle bodies removing the wear and added one size bigger valves. After that I restored the whole MFI pump, great piece of mechanical engineering.
With a lot of information from the forum
Jonny042 - Project heavy metal
Jeffc280sl - Open heart surgery
356RS [Mark Jung]
And many others!
I managed to restore the pump and build my calibration setup. Bringing it to Eisenbrand Renn & Motorsport is easy but only half the fun, right!?

So that as a short introduction, I ran into some issues that I do not really know how to resolve.
1)
In picture one [Bosch datasheet], rack movement at pre adjustment, I managed to get the following vallues for the rack.
Angle rpm rack
Max - 2000 - 7.02 (should be 7.00)
0 - 400 - 4.46 (should be 4.45)
8.3 - 800 - 4.25 (should be 4.25)
29.5 - 2000 - 4.73 (should be 4.75)

Now when I move forward to the next set of tests the Full load vallues,
Angle rpm rack
Max - 700 - 7.00 (should be 6.25)
Max - 900 - 6.83 (should be 6.75)
Max - 2000 - 7.01 (should be 7.00)
Max - 3000 - - - - (should be 7.3)

The full load vallues should be corrected via the rack head screw[as per Bosch data sheet], right?
Question 1: If so, I will alter also the value(s) obtained in the first table as the first line adjustment is through the rack head screw and that is the one I have to use to correct the second table. So feels like chasing my tail!

For now I am using a diesel calibration fluid that is odorless and not flammable (like petrol or white spirit) Only thing is that diesel (like) fluids have a slight higher viscosity.
Question 2: Does anyone has an idea on the effect of this difference on the quantities I am measuring/adjusting now?

Thanks a lot,
Bart http://forums.pelicanparts.com/uploa...1674231598.jpg
http://forums.pelicanparts.com/uploa...1674231598.jpg
http://forums.pelicanparts.com/uploa...1674231598.jpg
http://forums.pelicanparts.com/uploa...1674231598.jpg

Hereby a video of the working test setup!

https://youtu.be/o7XkqSn0jDk

Nice setup! I wish I could help you but it’s over my head. I would like to see what my pumps would do as I’m certain after 50 some years they could use some fine tuning.

I plan on doing similar in the next couple years. I’ll be following along!

I have built a setup as yours. Yours much more clinical appearing than mine. To quote a friend who stopped in for a viewing, ''looks like a collaboration between Juels Verne and Buckwheat''. I am focusing on race pumps primarily, so am not worried too much about any thing under 4500 rpm. Looking at your chart, the only piece that appears significantly out of range is Max/700. I assume you have checked the delivery amounts, and that is what I would want to see. You are going to have to spend a bunch of hours doing repeats and adjustments with detailed notes. Eventually you will have your reward, but it is not just monkey see monkey do. A lot of understanding will need developing. Have fun and good luck. Bob

@rudedude
Yeah, I guess they do run out of tune after 50 years….. mine was all over the place wen I sent it to Eisenbrandt in Germany. That is when I decide the pull the whole pump apart and do it all myself….Big fun and getting closer. New engine build will follow after this!

@Stownsen914
When I am done, as I probably will not start a business on this you could be interested in my setup ;-) By the way there are very interesting threads already on this topic. Going way more deep than mine, even that I just started.

@Bob
Thanks for your comments, the last line is where it is about, Have fun!
Cheers,
Bart

P.s. Will be continued probably this weekend and will be posted!

I hope the data sheet I sent you will help out.

Mark,

I don't see how the flow rate at rack movement of the 8.3 degree/800 RPM test can have rack movement of 4.25mm gives a 14.0-15.0ml flow rate, yet the 0 degree/400 RPM test has a rack movement of 4.45mm and the lower flow rate of 13.5-14.5?

There has to have been some error in that sheet originally from Bosch? As long as you ignore the rack movement for those tests and tune for flow rate (which is the only sensible thing to do) it doesn't really matter... but curious nonetheless.

Jonny, You're right about some error's on the data sheet. I sent Bart a 2.4S original data sheet yesterday. As for ignoring rack movement measurements on some pumps you are correct. Reason for this is because there are a lot of pumps out there that have been rebuilt, re-calibrated, etc; etc; and when the rebuilder adjusts the equal flow rates by loosening the plunger rack gear collars then that can change the real rack movement measurements while you are adjusting for equal flow rates. It's more important to have the correct flow rates than rack movement measurements on some of these old pumps.

Gents,

Thanks again as that is exactly what I am going to do, measuring flow rates also suggested by Jonny at my thread at early911s!

Cheers,
Bart

OK I'm glad it's not just me that couldn't make sense of that!

As for measuring flow instead of rack movement that makes total sense, at least once you are in the ballpark.

Part 3

Finally found some time this weekend to work on my pump.

Taking Jonny’a advise I started with
2000 rpm full throttle position
After some tweeking the clamps/collars on the main rack the results were:

Rack movement 7.02 mm.
Cylinder 1 2 3 4 5 6
CC. 48.5 48.5 48.5 48 48.5 48.5 [47 -48.5]
So check, that one is ok!

Now moving on to the next 4 tests
400 rpm; 0 load; run time 2:30
Cylinder 1 2 3 4 5 6
CC. 14 13.5 13.5 14 14.5 14 [13.5 - 14.5]
OK too!

800 rpm; 8.3 load; run time 1:15
Cylinder 1 2 3 4 5 6
CC. 15 14.5 14.5 15 15 15 [14 - 15]
OK too!

2000 rpm; 29.5 load; run time 30
Cylinder 1 2 3 4 5 6
CC. 25 24 24.5 24 24.5 24.5 [24 - 25]
OK too!


3000 rpm; full load; run time 0:20
Cylinder 1 2 3 4 5 6
CC. 54 54 54.5 54.5 55 55 [48.5 - 50]
NOK. I had to turn the rack 10 clicks CW (Less rack travel) but seem s to react slowly and only moved +/- 3CC by 10 clicks. I can go further of course to get to the 48.5 - 50 range, however I was afraid that I ruined the earlier settings…..
I did a check after this 3000 rpm full load exercise by redo the 2000rpm and 29.5 load. this resulted not in the earlier tuned 24.5 ‘ish CC but in 20’ish. This sounds logic as I turned the rack to achieve the 3000 rpm and full load volumes…..
So I am running in circles somehow!

I was pretty happy till the last bit. Is this me not understanding the system or drill!?
Thanks,
Bart

Bart, You have all the Low to Mid settings about perfect so why are you adjusting the main rack for the 3000 WOT. Adjusting the main rack now will alter all the other settings. You should be adjusting the White head screws on the spring capsule for the 2000 to 3500 adjustment. Back out CCW the White heads 1/2 turn each and see if that lowers the numbers.

Hi Mark,
I received the advice to use the rack for 3000 rpm and full load. I am a novice enjoying the learning curve!
Iread read somewhere (maybe from you) that each spring will generate a part of the total spring curve and the stiffness can be adjusted separately, however the whole curve can be shifted over the y-axes by using the rack……
So going back where I was before adjusting the rack at 3000 and 29.5. Checking earlier quantities again and than use the white screws?
So both 2000/29.5 & 3000/full should be adjusted through the white screws (stiffest spring)?
Issue I see that if I get 2000/29.5 ok through the white screws and 3000/full is not ok, by making 3000 ok I will alter the 2000 condition or do I miss something?
I wrote to Jonny on early911s. Not handy choice from me to be on both forums….
Thanks in advance,
Cheers,
Bart

Bart, First off, the only Bosch data sheets the state 3000 rpm @ 29.5 is for the 2.7RS. I thought you had a 2.4 S. For a 2.4S, the White head screws will get the main rack set for 7.00mm @ 2000 rpm and 7.30mm at 3000 rpm or as close as you can get. Then check the flows for 47 to 48.5 mL/1000 @ 2000 rpm and 52 to 54 mL/1000 @ 3000 rpm both at WOT. A little richer will be better because of todays gasoline specs. Remember these data sheets were made for real gasoline octane in the early 70's not like todays. Buy the way, you are really close to perfect specs on your testing. Great job Bart.

Sorry to be late to the party, Bart. It wasn't clear that you made adjustments between the tests.

Yes, as Mark says you'll have to bounce back and forth a bit between the main rack and part load screws to get everything dialed in. It would be good to double check that your test screw has fully seated the warm-up linkage in the "hot" position as this could cause issues, and also double check the other measurements ("f" and "c") to make sure they are in the ballpark.

There are other adjustments you can make but let's see if we can get you there without getting into those!

PS - Subscribed to the thread :)

Hi Mark & Jonny,

Yeah sorry for confusion….
@Mark: yep it is 2.4S spec and it was a typo the 3000 rpm and 29.5. Should have been 3000 rpm and full load. Also the CC numbers for the 3000 rpm full load check where wrong should be 52-54 CC.

@Jonny: yes I put the test screw in place in stead of thermostat. (Will buy eventually the lever type from Henry@Supertec). I choose a value in the given range that I have full bottoming out. Will measure the c-value to be sure.

Will do the measurements again when I can focus as I hate it when typos and misinterpretation of al your info happens. As said, I am close but not there yet……

So my test plan for the weekend will be:

Rpm Load Rack CC Adjustment Action
(1) 2000 Full 7.0 47-48.5 Rack Get rack displ, equal and correct quantities

(2). 400. 0. 4.45. 13.5-14.5. Idle.
800. 8.3. 4.25. 14 -15. Black
2000. 29.5. 4.75. 24-25. White
3000. Full. 7.3. 52-54. Rack

So is this test scheme correct?
(As both 2000/full/7.0 and 3000/full/7.3 are to be controlled by the rack? If so they are dependent and if one is correct, automatically the other has to be correct too….right!.)
Could you say looking at the Bosch test sheet, all full load conditions are corrected by the rack screw and all other load situations are by the ‘capsule screws’?

Will start fresh on Saturday and see what that will bring ;-)
Thanks again,
Bart

p.s. this 3000 rpm stuff is impressive. If you see the technical effort, all mechanical, of the MFI compared to a Weber carb where some simple physics will do the job. That is what MFI so much fun and I am really curious on the driving experience and throttle response!

Bart,
That test screw in place of the thermostat is you "f" setting basically. So next would be the ""c" setting with correct shims under the Barometric cell, .8mm is what your adjusting for.
As far as the spring capsule screws, the white pair adjustments are for 2000 rpm to 3000 rpm 80 degrees throttle. The main rack is NOT for adjusting the 3000 rpm full throttle. The main rack affects all the settings from Idle to WOT.

Keep at it, Bart! You're on the right track. You are best to study what's in front of you and come to your own conclusion about what's going to work, and why. Not saying that to be discouraging, rather the opposite - once you fully understand it you'll have that as your own.

The white screws AND the main rack will both affect the 3000RPM full throttle settings. The settings interact and intersect to some degree so making blanket statements about what screw is for what test result is asking for trouble IMO.

Don't forget that you can correlate rack movement to fuel delivery - if you keep an eye on rack movement you can make a change to an adjustment and instantly see the result without going through a full volumetric test each time. This one tip can save you quite a bit of time and frustration.

Also, I don't know if it's just me but at first I would get confused about which way to turn the screws and/or main rack adjuster to add or subtract fuel. I always try to think of the main rack like a tap, and you are turning it CCW to open it and add more fuel. The lower screws are the opposite and the phrase "Lower > Left > Lean" is a good memory aid.

Gents,

Thanks again for the motivational words and advice!
I know I am close and next Saturday/Sunday I will have some dedicated quality time to be invest in my MFI ;-)

More to follow!
Cheers,
Bart

Hi Mark, Jonny and others,

I dare to say that it was a successful Sunday as I believe I cracked it ;-)
It is late now and I will write my findings and kind of protocol down coming days.
Happy and also a bit sad as it was big fun getting closer and starting to understand how it works.

Now I still have to build the engine I plan and will write down the spec list down in another thread.
That’s all for now, more later!

Cheers,
Bart

Congrats, Bart!

Look forward to hearing about it. Thanks for sharing your journey!

Part 4

Hi!
As already mentioned in the last mail, I finally calibrated my MFI pump ;-)
I will try to write down my experiences and learnings.

First I would like to thank the forum and in particular Mark Jung from MFIwerks and Jonny for there support and input they were a great help in this jungle of springs, screws and CC’s…..
Below is just my experience based on the very first steps in this field and everyone is more than welcome to comment!

Anyway I focussed to get the following settings correct.

Rpm load rack CC’s time/1000cycles adjust
1) 2000 full 7.0 47.0-48.5 0:30 rack
2) 400 0 4.45 13.5-14.5 2:30 idle
3) 800 8.3 4.25 14.0-15.0 1:15 black
4) 2000 29.5 4.75 24.0-25.0 0:30 white
5) 3000 full 7.3 52.0-54.0 0:20 white

For 1) focus on quantity and equal over cilinders

First I put a lot of energy to get the rack displacement correct, however later on, you realize that CC’s are the dominating factor to get right. Yes they are connected but the focus and attention should go in my opinion to the fuel quantities.

When I, after a lot of runs, got everything correct from test 1-4 test5 (3000/full) gave to much CC. Changing the white screws would automatically influence the 2000/29.5 test. When I started adjusting the white screws for test 5, I felt that they where way less (not) responsive (influencing the rack movement) as in test 4. One should realize that all the springs behind the capsule are pretensioned by the different screws and all interact. So it can be seen as a graph where 3 spring stiffnesses who are connected/interacting. Each spring has a certain stiffness range being controlled by pretensioning the screws (idle = idle, mid rand = black and high range = white). This all will influence the movement of the rack through the flyweight assembly that is influenced/restricted by these different springs. (Maybe I should draw this graph (has been done before in the forum….) the rack screw will move this 3-spring graph up or down in total!
Found it, a graph explains it better [Graph from JCGE] Interesting read anyway:

http://forums.pelicanparts.com/porsche-911-technical-forum/386414-mfi-pump-open-heart-surgery-7.html

So back to my issue on the high volumes in test 5. The only solution I saw was moving the graph down by adjusting the rack and consequently using the different spring screws to correct quantities again after the rack adjustment, remember the rack adjustment influences the whole rev range.
.
Something else I learned, at least it worked for me is starting with test 5 and than working your way down. Reason seems that the spring differences are big, so when the highest stiffness spring is adjusted correctly(white screws) it will have little influence on the less stiffer spring (black screws) etc. The other way around that if you adjust the black ones correct for 800/8.3 and you have to adjust the white screw later these will have an impact one the earlier setting (800/8.3)

In the end it will be a bit back and forth and double check but as Jonny already said if you get it right because you understand the internals the satisfaction is sweet!

Another observation is that if one cilinder is on the low side of the range in one test, in another test it can be another cilinder. (By the way I have 6 new injectors tested with the injector tester from: https://mercedessource.com/store/mfi-fuel-injector-tester-assembly-testing-and-cleaning-operations-demand-video

So these are my vallues

Rpm load rack 1 2 3 4 5 6 must
400 0 4.23 14.0 14.0 13.5 14.0 14.5 14.5 13.5-14.5
800 8.3 4.04 15.0 14.5 14.5 15.0 15.0 15.0 14-15
2000 29.5 4.51 25.5 25.0 25.0 25.0 25.0 25.0 24-25
3000 full 7.09 52.5 52.5 52.5 52.5 52.0 52.5 52-54

Here a run at 3000 pump rpms and full load……pretty impressive if you stand that close! https://youtu.be/VIkHr7-ULH0

Some fine tuning can be done but I guess when I finished my new engine build (will be another thread) I will mount a LM-1 and see what is really going on!

I also added a picture of a few of the notes I made to get it all right.

That’s all for now!
Cheers,
Bart

P.s. I have no idea how to get the tables well alligned (They look good in the editing box but not in the preview)
http://forums.pelicanparts.com/uploa...1675724211.jpg

Excellent write up Bart. Im fairly sure you have just about mastered the MFI pump adjustments. Using the LM-1 when you get your engine done will open up a few more adjustments may be needed. Great work.