Finally drove MFI

[Jeff Higgins]

Originally Posted by winders

When you watch videos like this, I see why!!!

https://youtu.be/u_E14RKskbM

Cool video, Scott. I have to say, I kind of "geek out" on that kind of stuff. Some of my favorites are the Plexiglas cylinder heads wherein we can watch the whole episode, from induction through combustion to exhaust. I love the fact that we are able to film these things today. I have actually seen extremely high speed photography from inside a "real" combustion chamber. The learning opportunities are truly exciting.

It's surprising how much fuel we see clinging to, and condensing on, those throttle bores. And yes, could you imagine a backfire into one of those? A "carb fart" of epic proportions - Fat Bastard himself in the bathroom stall. Not pretty...

[winders]

What surprised me the most was how much fuel was kicked back out due to reversion.

[Jeff Higgins]

Yup. It cleared right up as RPM's increased. I would love to see the camshaft profile and timing on that thing. Obviously lots of overlap, very narrow lobe centers, and very long duration. Glorious up on top, but down low, not so great. This video was a wonderful illustration of that.

[winders]

That illustrates how 996.2 GT3 intakes can help with mid-range torque....that reversion is used to push air into the cylinders sucking in air at the same time. That would improve throttle response!

[Jeff Higgins]

Originally Posted by winders

That illustrates how 996.2 GT3 intakes can help with mid-range torque....that reversion is used to push air into the cylinders sucking in air at the same time. That would improve throttle response!

Absolutely. It's amazing what we have accomplished with regards to that kind of resonance tuning. On both ends, really - on the intake and exhaust. Our ability to actually see, measure, and analyze this kind of thing has really opened our eyes. The level of reliable power we see today is a testimony to all of that.

[panzerfaust]

High injectors location and or shower heads (for +8000 rpm ) = better atomization at high rpm’s, cooler charge for more peak power.

Low injectors location = better low end response, typically placed at a higher velocity location since the ports are necking down near the injector nozzles, less sensitive to injector inclination angle, less wall wetting , better emission

There seems to be a big misconception here of what Direct Injection actually is to compare it with a 45yr MFI ��. The fuel delivery systems and attributes aren’t even remotely close to each other

[QueWhy]

Originally Posted by Jonny042

I've made a career of working on computer controlled process systems and instrumentation - EFI always feels a bit like work to me, so one of the attractions of MFI was that at least some of what I need to do to get it to work involves using my hands and tools, not the laptop, and trying to get it to connect via a serial connection over a USB cable and hoping the software, firmware, blah blah blah all match.

If someone wants to pay me my full hourly rate to do that, I gladly will, but it's not something I do for fun if I can avoid it.

Yes I understand that I'm using datalogging and a laptop to do the tuning.....

This is why I'm 100% dead set against EFI, regardless of its objective benefits.

[Jonny042]

I'm really impressed with the passion we're seeing on this forum!!! That's what makes it so great. Lots of interest in MFI, too, and dragging it from the dark ages into the light!!!!

Jeff Higgins gets credit for starting this ball rolling and I like to think I've been able to prove he was right!!! Here's a post I just put up on my build thread with some datalogging results on my 3.2SS MFI motor. If I didn't know myself better I would probably call it "done".... but I'm not going to be that cheeky.

I'm using a TPS, RPM, and AFR into a datalogger to record the behavior of the motor, and using the data to re-shape the space cam in a few strategic areas. So far, so good:

Originally Posted by Jonny042

Ok so that was a success..... apparently MegalogViewer agrees since it's re-scaled it's colour representation on the scatter plots:



I still need to adjust the idle and perhaps lean the entire map but first I wanted to check the data and make sure I'm not seeing anything dangerously lean. I'm not, although the WOT traces are getting there. At this point I'm not inclined to think I will be able to improve on the following:



It is, after all, tractor technology

The histogram view is also much improved - there is still a small lean spot just above and to the right of where I did my last adjustment, which I might chase down, particularly if I do lean the whole works out a smidge:



The car is running like a CHAMP. It's amazing. Runs, idles, pulls, coasts, any rpm, any throttle opening, not a fart, pop, bang, or miss. And it just SCREAMS. It does like those AFR's in the high 12's and low 13's.....

[Jeff Higgins]

My hat's off to you, Jonny, you have done a magnificent job of demonstrating what can be accomplished with this "archaic" system when utilizing modern technologies. Your hard work and, dare I say - OCD - (meant in the very best way) have so vastly eclipsed my "rude and crude" efforts. Congratulations, Sir - well done. Very well done.

[PorterF]

Bravo Jonny042! The scatter plot to space cam relationship is brilliant.

Found a little article on MFI that is light on details but is entertaining nonetheless.

https://www.roadandtrack.com/car-culture/a26665/mechanical-fuel-injection/

[r lane]

Originally Posted by Jonny042

I'm really impressed with the passion we're seeing on this forum!!! That's what makes it so great. Lots of interest in MFI, too, and dragging it from the dark ages into the light!!!!

Jeff Higgins gets credit for starting this ball rolling and I like to think I've been able to prove he was right!!! Here's a post I just put up on my build thread with some datalogging results on my 3.2SS MFI motor. If I didn't know myself better I would probably call it "done".... but I'm not going to be that cheeky.

I'm using a TPS, RPM, and AFR into a datalogger to record the behavior of the motor, and using the data to re-shape the space cam in a few strategic areas. So far, so good:

I guess it is possible that I am the only viewer that doesn't understand the graphs, so could I get an interpretation of them. Not clear on your horizontal or vertical values. What am I looking for, AFR, throttle angle, RPM, HP? A users manual for each one would be wonderful. Thanks, Bob

[Jonny042]

Originally Posted by r lane

I guess it is possible that I am the only viewer that doesn't understand the graphs, so could I get an interpretation of them. Not clear on your horizontal or vertical values. What am I looking for, AFR, throttle angle, RPM, HP? A users manual for each one would be wonderful. Thanks, Bob

I doubt that you're the only one! I did receive some good natured ribbing on my build thread for posting graphs without properly labeling the axis:



So if any of this is hard to understand, I might be partially to blame. I showed the above graph because it's a further graphical representation of the fuel delivery of the pump. I'll get back to it.

All of the graphs/charts/plots you see related to MFI will have an RPM and Throttle Position component. These are the only two input variables to the pump fuel delivery rate, which needs to vary for different combinations of RPM and throttle opening (Lets ignore cold start enrichment and barometric pressure). Of course the pump inherently delivers more fuel as the motor RPM increases, as it is timed to the engine - the motor speeds up, so does the pump, and you get more fuel. But the fuel delivery rate will change even at different RPMs as the engine's needs are not linear.

So, back to the first graph to interpret, the fuel delivery rate, seen above, is a result of actually doing a volumetric test on the test rig, measuring fuel delivery over time. The results are plotted in the units "ML/1000 Strokes" and that axis is the Y axis on the left. The X axis on the bottom is RPM.

That graph has fives lines, each of which is the result of volume measurements at 7 different pump RPMs = 500 through 3500 (which is 1000 through 7000 engine RPM). Each of the five lines represents the test results with the throttle angle clamped at a different position (TA = throttle angle) 0 (closed), 16, 30,55,and 82 (full throttle)

So yes that is 5 throttle openings at 7 different openings, for a total of 35 different tests..... here's a video of me doing a test at 2000 Pump RPM (not sure the throttle angle but based on the fuel in the test tubes I'm going to say WOT:



Here's another video showing the max fuel delivery at full RPM. It's actually pretty astounding how much fuel gets pumped into the motor at "full chat":



Once the pump was set up on the test rig, as best as possible, you can see that my best guess of the fuel requirement at 6000RPM was about 73.5 ML/1000 strokes which has turned out to be a very good guess based on actual AFR measurements.

Less correct was the 18 ML/1000 at 0 throttle/500 RPM (idle) which is pretty damn rich.

(To be continued)....

[jluetjen]

That's some great experiments Jonny0412. What year model did that pump come out of? I'm curious how your findings stack up against the factory settings. Porsche published the fuel consumption at different revs and WOT as part of their dyno graphs for the different cars. For example...

[356RS]

Originally Posted by Jonny042

That graph has fives lines, each of which is the result of volume measurements at 7 different pump RPMs = 500 through 3500 (which is 1000 through 7000 engine RPM). Each of the five lines represents the test results with the throttle angle clamped at a different position (TA = throttle angle) 0 (closed), 16, 30,55,and 82 (full throttle)

So yes that is 5 throttle openings at 7 different openings, for a total of 35 different tests.

Great work Jonny. I'm impressed with your data. Fuel flow data recording is very time consuming. About 2 hours between dumps and fills for a flow chart that goes from Idle to WOT @ every 10 degrees and from 500 rpm to 4000 rpm. It's the only real data of the Space Cam's complete flow profile.
Modified MFI engines require extra time and engine dyno time to make adjustments on the space cam re-mapping. If you do enough of these re-map's and collect the data you end up with lots of space cam re-mapping data. A #55 T space cam can be re-mapped for a 3.5 to 3.8 MFI engine. Lots of work. Some flow data below:

[Jonny042]

Originally Posted by 356RS

Great work Jonny. I'm impressed with your data. Fuel flow data recording is very time consuming. About 2 hours between dumps and fills for a flow chart that goes from Idle to WOT @ every 10 degrees and from 500 rpm to 4000 rpm. It's the only real data of the Space Cam's complete flow profile.
Modified MFI engines require extra time and engine dyno time to make adjustments on the space cam re-mapping. If you do enough of these re-map's and collect the data you end up with lots of space cam re-mapping data. A #55 T space cam can be re-mapped for a 3.5 to 3.8 MFI engine. Lots of work. Some flow data below:

I can see how if you used a "T" space cam as a starting point (like I have) you'd get good at making the necessary adjustments for the bigger displacement motors. Time consuming (and therefore expensive!!!!) but worth every penny IMO!!!! It's great that you are able to provide this service for the hot rodders and racers.

I'm looking with great interest at the flow rates for that 3.2SS and comparing them to mine - I'm seeing that 80ml/1000 rate and comparing it to the 74 or so for mine, that must be one hot motor!!

Thanks for sharing!