MFI Pump - Open Heart Surgery

[dicklague]

Here is a one minute slice of AFR with the engine fully warm. On left is idle AFR and RPM. RPM red AFR green. Seconds along the bottom.



The idle mixture is good, maybe a little lean. The pump responded well to that 6700 RPM spike. It did not go excessively lean, which sometimes happens as the throttle positionopens and the pump gets ahead of RPM curve.

I will post a longer run later.

[Jeff Higgins]

Wow, I can't believe how this thread has taken off. I haven't checked in for awhile, so I just took the time to read through it from where I left off. Amazing... now that the real MFI experts have chimed in, the information you guys are revealing is just incredible. I'm just guessing, and hacking away at it, compared to you guys who know what you are doing.

Anyway, I thought I would update you folks with my latest efforts. My biggest "breakthrough" to date came as a result of studying Mark's space cam pictures and explanations on just what the stylus is supposed to do when tracking on the space cam. You see, initially (when I started this thread) I was assuming "left for lean, right for rich" on the idle spring screw held true for the mid and high range screws as well, and that they had unlimited range of adjustment from rich to lean, from bottomed to fully backed out.

Well, talk about an "ah hah!!" moment... Mark's photos and explanations blew the bottom out of that assumption. I now understand that simply screwing in these screws will not richen the mix over their whole adjustment range; they are simply meant to adjust stylus position on the cam, making up for manufacting tolerances in pump assembly to achieve a specific delivery at given rpm's and throttle positions. So, with a better understanding of what the stylus is supposed to do and the role of those adjustments in making it do that, I broke out the LM 1 and started over.

I went back to square one and re-centered all of the adjustment screws halfway between bottomed out and flush with the tops of the detant springs. The goal was to now see if I could identify when the stylus was actually following the ridges and groove in the space cam properly, by analyzing the A/F ratios under different operating parameters. I was hoping, as I was initially, that I could do this "at home in my spare time" without the benefit of a pump dyno at my disposal. To see, once and for all, if the "home tuner" had a prayer of pulling this off. Yes, I already knew I could make the car run, and make it run well, but it had been nagging at me that it could be better. As Kenik has noted, the car had always belched a pretty good cloud under initial acceleration, with the pump going quite rich at the first stab of the gas at lower rpm's. I think I've taken a few years off poor Kenik's life, what with all the fumes he has ingested following me around...

I'm happy to report, however, that those days are behind us. While I'm now convinced that the 015 "T" pump will never be perfect at all rpm's and throttle positions on such a dramatically larger, more powerful engine, it can be made to run very, very well. I now have idle at about 12.8-13:1 @ 1,000 rpm, steady state freeway cruising (3,000 rpm @ 60 mph in fifth gear) at about the same, with it richening up to about 12:1 on slight grades and leaning out to 14.5-15:1 coasting downhill at part throttle.

WOT, full load acceleration still shows the greatest delta in A/F, as one would expect. In running through the gears from 2,000 rpm to 7,000 rpm, it starts out at a rich 11.2:1 or so and leans out to 13.5:1 @ 7,000. After "climbing out of the basement" at 2,000 rpm, subsequent shifts never drop rpm below the high 4,000's, and A/F remains between 12.5:1 and 13.5:1. The A/F ratio doesn't go leaner than 13:1 until rpm's top 6,500, and this motor will never stay there for long at WOT anyway. "Perfect"? Hardly, but I can certainly live with that.

The key has been in making very subtle adjustments, and (just like my first efforts at this) keeping very good notes, along with charting every test run in my LM 1 software. The adjustments are much more sensitive than I first thought, especially after I found the "sweet spot", where I believe the stylus is actually following the ridges and groove. In the end, I was able to achieve a much greater delta in fuel delivery with this approach than I was when I thought I could just keep screwing in the adjustments to make it richer and richer.

The motor now dynos solidly in the 250+ hp range, and 235+ ft lbs of torque at the crank (corrected @ + 15% from rear wheel figures). My wife and I took it over our Cascade range on Sunday with fellow Pelican Scooter and his wife in their car, and we averaged 17.4 mpg for the trip, over mixed two lane highway cruising and somewhat spirited backroad bombing. The pipes were a nice light ash color, and best of all, Scooter and Lori didn't smell like gas after following us around all day...

Anyway, it's certainly been a learning experience, and one heck of a lot of fun. I think the answer is "yes", the home tuner, with a bit of care and attention, coupled with the knowledge offered up by our real experts, can in fact make an oddball application like mine work very well.

[911st]

Wow, never seen anything like that. Nice to have good information.

What is your build and what space cam are your running?

I am guessing you did a run up on the street. We can see the shifts and the 2 second gear run is the cleanest to judge.

Looks like idle is about 13/1.

There is a small lean spike just off idle.

With the 1-2 shifts it goes lean at first to about 3000rpm (first spring stage?).

Richens from about 3000 to about 4500.

From about 5000 to 7000 it progressively lean's.

At 4000rpm cruse it is to fat.

I might experiment with some of the following:

By shortening the linkage to the MFI pump it might reduce the small first lean spot as it gets the stylist down the acceleration ramp on the space cam ram quicker.

4000rpm steady state is not where we plan to cruse so I would disregard that. More interested in what AFR is at around 3k rpm in 5th.

Richening from 3k to 4.5k indicates the stylist is progressing to fast.

Leaning from 5k to 7 k may indicate the stylist is progressing to fast.

Thus, might try tightening the second stage spring at first and see how these two areas are effected. If it improves the 5 to 7k only, I would then try tightening the first stage, lighter springs.

Could be wrong.

My goals being 14/1 at idle (air bypass screws), 13/1 with first accel (rod length), 14/1 at cruse (tighter springs), and at WOT 13/1 with a flatter AFR curve (the deepest part of the space cam where it delivers the most fuel is at TQ peak).

[dicklague]

911st. Great information from that brief graph. I bit of it over my head, so I will go back to all the data on here and try to figure it out. I have the linkage to the pump set at exactly 114mm which I think is a good starting point.

So far I have been fiddling with idle mixture screw, and the center rack adjustment. Also just took off the warm up sensor and put in a screw [see above]. That is pretty much set to full warm setting.

Maybe my next step will be shortening the linkage from 114mm. What do you suggest I try?

And Jeff, thanks for starting this great thread and all your work.

[356RS]

Shortening the MFI pump control rod from the 114mm setting will cause the TB's to open up some off the idle stop. Lengthening the rod will cause the MFI pump to start with an added degree of more fuel.

[911st]

Mark is right. you would add length to the rod to get it more fuel faster. Sorry have not touched a pump in years.

Adjusting the rack will move the whole AFR line up or down. The springs change the rate the stylist progressed to usually a higher deliver rate with increases in rpm and then after TQ peak the rate at which fuel is pulled back.

I would do one thing at a time. I would probably start with idle and off idle. The of pump rod adjustment is easy to experiment with and it should not effect the progression much but it might make the cruse AFR richer so I would watch that with your experiment.

Then, if up to messing with the internal springs do that one step at a time.

When I was really concerned about a given point of operation, I would mark the space cam with a felt pen. Then repeatedly run the motor up to that point so I would better identify where on the space cam I was. This is a lot of effort.

If it is a stock MFI build like a 2.7RS it is often just better to have it set on a test bench back to spec.

Disclaimer. I am thinking this stuff through and making it up based on what I know about how the pump works. If you do want to try this, keep good notes so you can put things back to where you started. I spent a lot of time figuring this things out but still learn something new about them from time to time.

I assunm one has a close to stock starting point, things work well, are balanced, and you understand other basics like effecting idle AFR with the idle bypass screws and butter fly angle and have already gone there or will do so to further balance things.

Only trying to be helpful.

[dicklague]

I am not sure what cam is in the pump. Mark Kinison built the motor about 12 years ago. I assume it is 2.7RS specs.

He told me the engine has a 911e cam. He likes that better than the RS cam.

The pump rod was 111.7 up until a couple of weeks ago, and I moved to 114mm as a starting point, and that made a big difference is staying rich on throttle opening. I think I will leave it this way for a bit.

I am thinking I may lean the main rack a couple of clicks next and see if that will lean the part throttle 3000 and 4000 running.

And thanks for the reminder. I am keeping good notes on what I do.

I have to put in a plug for the Daytona Sensors Wego IV AFR instrument. Would be lost without that. It has opened that door to see what ANY adjustments are doing for the AFR. Our hosts now carry it.

[911st]

Cool, so you already increased the rod length once and found some improvement off idle.

E cams on a 2.7 is a custom set up which to me means you have license to monkey around with things to try to make improvements.

A 2.7E motor is a great motor.

If you have Bruce Andersons book he has a dyno curve of a 2.7E w sport muffler and carbs overlayed on a 2.7RS motor where you can see the probably differances in HP by rpm.

They are not to far off but the E cams need more fuel up to 4500rpm and the RS mostly more past that.

[al lkosmal]

Wow,
I am starting to learn something.....I will be back out in the garage tonight with LM-1 fired up.

Thanks for the info.

Al

[356RS]

As for the 114mm MFI pump control rod length, I would not alter it. My earlier post was just to describe what happens when you alter the length. If you need to advance the space cam, that's another critical adjustment inside the pump.

[911st]

Mark,

I mostly agree. Especially with a stock style MFI motor build.

Along those lines I guess I would also not recommend playing with the internal spring tensions to.

My hope is that you or some one could take just take a TQ chart with overlayed AFR plot and find the best space cam and calibration to fit odd combinations as a more one shot thing.

However, with a bastard system where there is no proven spec from Bosch or Porsche it is just another unorthodox option if one needs to work a flat spot off idle.

[356RS]

You might help the flat spot when you altar the control rod but your also limiting the WOT of the throttle bodies. Reason is the pump control lever, because it was lengthened, will bottom out now at WOT (80 degrees) before the throttle bodies can. So you end up with the butterfly's at maybe 70 degrees instead of 80 degrees at WOT.
What pump do you have that has a flat spot off idle?

[dicklague]

Good point in the butterflies. I was wondering about that. Makes sense. If they are set up to be fully open at 114mm of rod length and then you make the rod longer they will not fully open any more.

[BigD9146gt]

Mark and everyone else...

its taken me quite a few hours to work through this thread... and awesome stuff all the way!!! Theres a reason why that company in deutchland (first couple of posts) didn't give up any info... cause you can charge a mint for this info! Oh, and as a machinist (with a good friend as an EFI expert) if there is any two that can build a unit to attach to these and adjust like a piggy back unit, Matty and I can!

On a separate note, I've been using an LM-1 to dial in my 2.8 MFI engine, and having most of my tuning issues at part throttle sub 3K rpms, violent engine jerking. Do I need to go alter the white screws? Leaning the idle doesn't seam to make much of a difference. Before I get too far into tuning it, I talked to the two guys who, before me built the engine I currently am running, a 72E with original Mahle 92mm p&c's and RSR sprint cams. They said the MFI pump has been rebuilt by Gus back in the day with the RSR space cam... do I dare open it up and take photos for the cause???

[dicklague]

do you have a graph from the LM1 you can post. Might help these experts [not me!] figure ouit what might be going on.

[jeffc280sl]

dicklague,

Have a look at the afr chart on my 3/26 post. Prior to tuning I had the same sort of profile that you have. It would be nice if you had a map sensor input but I can see a lean condition occuring as you begin accelerating. Throttle opens wide and the mixture goes lean for a period of time. That can be tuned out so that afr is steady over the entire acceleration sessiuon.

[dicklague]

I do have MAP. I just did not have it hooked up for this run as I am messing with a temperature sensor.

I will do some with the MAP sensor.

[911st]

Originally Posted by Jeff Higgins

...The motor now dynos solidly in the 250+ hp range, and 235+ ft lbs of torque at the crank (corrected @ + 15% from rear wheel figures)...

Jeff,

I almost missed this with all the fun.

I am very, very interested in the thought of an SC long block MFI conversion.

Am I recalling correctly that you are running stock SC cams & pistons? Is that correct?

Very good number by the way.

If so, how did you time the cams?
Which CR pistions?
Did you mesure you CR or do any thing to increase it?
What exhaust / muffler?

Link to a Dyno plot?

Thanks if you can help.

[kenikh]

Originally Posted by 911st

Jeff,

I almost missed this with all the fun.

I am very, very interested in the thought of an SC long block MFI conversion.

Am I recalling correctly that you are running stock SC cams & pistons? Is that correct?

Very good number by the way.

If so, how did you time the cams?
Which CR pistions?
Did you mesure you CR or do any thing to increase it?
What exhaust / muffler?

Link to a Dyno plot?

Thanks if you can help.

Jeff's cams: Dougherty; .485"/.470" lift and 254/238 degrees duration at 1mm lift, timed to 4.5mm
10.5:1 JEs in Nikasil barrels
Electromotive twin plug ignition
Modified factory muffler to "Upfixin" sport spec

[Jeff Higgins]

Wow, Kenik - I'm impressed. You remembered better than I could have.

There are a few of us running this particular cam, with my friend Kent Olsen down in Florida with what I believe are the second set. It's a grind John Doughery came up with for me after a few phone conversations and emails. It's essentially just his GT2 cam with the lobe centers narrowed to 102 degrees from its standard 113 degrees. He calls it his "GT2-102" grind.

The whole idea was to provide substantial lift to allow the 3.0 to breathe, but short duration to drive the power band down into a more useable "street" range. My peak horse power comes in at 6,000 rpm, and peak torque somewhere in the low 5,000's. I'll have to see if I can find the latest dyno chart from last year. One of my early ones is back on the first couple of pages of this thread, but I was still running it single plug back then, and had detonation issues. Here is one from two years ago, after twin plugging it and initially sorting through the MFI:



It still wasn't in as good of tune with regards to fuel delivery and timing as I would have liked back then, but I do have a newer, "improved" chart somewhere.

Exhaust consists of SSI's, and now has one of Ben's (M&K) dual out sport mufflers. The above chart was with the old Leisritz sport muffler. We are heading back to the dyno in a couple of weeks, armed with European Racing 1 3/4" headers and megaphones. We'll see how it runs on those.