Fantastic job. I may show up at your front door with a couple of pumps and a cold twelve pack. I wonder how the airlines feel about pumps as a carry on item. Have friend that tried a Muncie 4 speed with his luggage. They were not happy about the leaking gear lube. Know one seems to have a sense of humor these days. Keep the pictures coming.

Fantastic information here and I love where this is going. Custom Space cams for engine configurations not contemplated by the factory would be too cool!

Finished up the graduated cylinder stand today. Funnels fit great.

http://forums.pelicanparts.com/uploa...1236299764.jpghttp://forums.pelicanparts.com/uploa...1236299837.jpg

Injector mounting plate and graduated cylinders look OK.

http://forums.pelicanparts.com/uploa...1236299933.jpghttp://forums.pelicanparts.com/uploa...1236299990.jpg

First, wonderful work.


It should not be that hard for someone that has the skills.

The space cam is just a simle three dimensional map carved out of metal that a stylist rides on to determine fuel quantities.

Along one dimension is throttle angle or load.

Along the other axis is RPM.

A stylist rests on the physical map and the lower it rides, the more fuel delivered per cycle.

The lowest point is likely at WOT and TQ peak.

Then this map is wrapped around a barrel that a stylist ride on. The deeper the indent the more the fuel. The map rotates under the stylist with throttle angle and is pulled from one end of the barrel to the other with changes in rpm by the centrifugal force of the rotating weights.

Measure fuel delivery at various heights of the stylist and you have the fuel delivery part.

Take a TQ curve from a similar cammed motor and your have your delivery curve.

We just need a program where we can enter enter the equivalent of duty cycle numbers into a the cells on a spread sheet and so the program can translate that into what a mill needs to carve it and we are set.

Would be cool.

:)

911st, Thats a great idea, especially because of the cost of a NLA space cam today. $750 for a new RS space cam. fits in the palm of your hand. I have no idea what the tooling would cost to get something like this going. I have a NOS RS space cam that I could let someone MAP out. Great ideas.

Copying a cam should be a bit easier.

The kind of person that can make this happen should also be able to set up a tool to scan in the different space cams profiles.

The RS cam I bought about 10 years ago cost me just over $400 through EBS I believe.

I hope one of our more technically gifted will step up so we can see how easy this all could to do.

Instead of burning EFI chips, we would be carving space cams. ;)

Aerospace machine shops with CNC grinding equipment could do it pretty easily, as long as you can provide them a program. They may even have a scanner that can capture it. Call around - I'd bet there are a few in Seattle since Boeing is here.

Mark, I have access to Pro E software to create a solid model of the space cam and a CMM to get some measurement to clone some RS space cams. Not sure about the material or surface hardness but you should be able to hard turn without finish grinding.

Gary, I don't know much about the hardness data, but I can tell you the material is HARD.

I remember them being hard but I had not problem taking a Dremel grinder to it or using a normal flat file on it. Was only on the 2.4, never did an earlier one.
:)

I could model it, as well (Pro E or SolidWorks). I have already done so with a 69E cam for the '69-71 pumps. The 2.4/2.7 pumps are a bit different from the '69-71 (roller stylus and the space cam is taller-same dia.); I will do a 72E space cam (which I have ) when I have some spare time (never). Then we could compare the differences of the 2.4E to the 2.7RS and figure out the appropriate amount of grinding to convert an 015 (2.4T) cam to RS or other similar hot rod spec.

I also have a real 2.8RSR Space cam and the lightweight set of flyweights and RSR springs that accompany it ;) I have thought about this project before; very interesting..let me know if you ever want to talk further about it Mark. I also have the fuel distribution graph and soon should have dyno sheets for a ~RSR spec. 2.8 engine I am building.

Also I know a guy that has a scanner (as opposed to modeling it). I don't have much experience in that area yet.

Gary; 911st; and Scott. All great ideas. I know nothing about the Cad or Software end of this. What I really need to know next is the step by step procedure for getting a pump ready before it goes on the test dyno. Then the adjustments and readings needed to do the job right. To be honest with you, I got the test stand ready, but really have no idea how to test it correctly. Just for fun today I put on a 72 T pump and run a series of tests at different throttle angles and RPM's. As per the factory spec sheet the pump was + - 2ml on all tests. Pictures below of the test tube stand.


http://forums.pelicanparts.com/uploa...1236389423.jpg

http://forums.pelicanparts.com/uploa...1236389461.jpg


http://forums.pelicanparts.com/uploa...1236389481.jpg

I am extremely impressed!

Are you saying you are getting the same fuel delivery as different rpm's? If so you rack may be stuck.

What is your goal for doing this. Is it to calibrate a pump back to stock settings?

:)

No, I'm getting pretty close to the factory specs. at each of the set throttle & RPM settings. There's about 18+ different throttle & RPM setting for the complete test. I'm at home now and do not have the specs I wrote down during the tests. All I can say is it's very easy to read the ml's after each test, and then I can lift off the test tube panel and dump all the Stoddard Solvent back into the drain and then run another test. The space cam ideas from everybody are exciting and I can't wait to get into that. I'm packing up tonight and going to catch a plane to LA for this weekend's Porsche festivities.
PS: 911st My goal is to be able to test and recalibrate customers MFI pumps. Also be able to modify the pump for larger engines and, or, replace there stock 72-73 T space cam with an RS space cam.

You need a few things:

These are the inputs to the pump. Know these inputs and what type of pump space cam you have, and you can determine its output.

1) Altitude calibration. The gold-plated barometer canister can on the top of the pump controls the little barometer adjustment rod inside the pump. The range of motion is not much; you could also substitute a bolt and then have an external barometer (like how they do it for the weather) to dial the pump in for that barometric pressure.

2) The warm-up thermostat (stack of disks) that is on the front of the pump. Sometimes, this thermostat is removed.

3) Throttle Position Protractor. There is a protractor that goes on the throttle rotation lever and controls the degree of rotation of the space cam inside the pump. The lever has a range of motion {0°<θ<90°}. Similarly the space cam as you've seen does not revolve, it simply rotates within a range of 0-90°.

4) RPM. Self-explanatory.

Once you know these things, you are totally in business. This is an awesome project, Mark. :cool:

Now to set up a digital flow meter, so everyting can be graphed! ;)

Scott, thank you very much. I'll be getting back to you in a PM today.
Kenikh, I have already looked into this idea. I found some very low flow, and high pressure digital flow meters, but out of my budget right now.

Custom Space Cam
 

You need a few transfer functions to achieve this...Ultimately you need to get to Throttle butterfly (or slide) position vs RPM vs cc fuel per stroke

so the base transfer functions could be.....

1 Throttle butterfly (or slide) position vs Pump lever angle (Correlation)
2 Pump lever angle vs Space cam rotation
3 RPM vs Space cam displacement (fore/aft)
4 Space cam "lift" vs Rack displacement (Translation mm vs mm)
5 Rack displacement vs Delivery volume (speed & viscosity effects)

and for compensation

6 Temperature vs Thermostat displacement (linear 18.25mm+0.091mm*T°C)
7 Thermostat displacement vs Rack movement
8 Barometric pressure vs Baro cell displacement (linear 0.58mm/100mmHg)
9 Baro cell displacement vs Rack movement

I wouldn't assume they are necessarily linear, or independent....

custom space cam
 

John,
Your so right. The space cam has many important map area's that have to be contoured into the design for each type of engine. Not easy for the custom ones, but maybe a little easier for stock reproduction ones.
Right now I'm working on the tools to get the pump dialed and set to factory spec before the flow tests begin.

I think the bottom line is that the stock ones give enough flexibility that any configuration can be supported if you can create the factory spec ones. Evidenced by Jeff Higgins being able to (just barely) get a stock 2.4T pump to work flawlessly with a "hotter than S" 3 liter, If you can get T, E, S and RSR space cams, you should be able to cover any config at any displacement.

It depends on what you mean by flexibility.

When you think space cams, think total Torque, not Total HP. As you know, HP is total power, TQ is measurement of how much power is made with each RPM or stroke. Make peak TQ at a higher rpm and you make more HP (TQ x RPM)/5252= HP.

My chipped 3.2 makes its total peak at about 5200rpm.

It needs the most fuel per stroke at TQ peak of 5200rpm.

At 2500rpm it is making 80% of its peak TQ and needs 80% of the fuel deliver per stroke at that point as it dose at peak TQ.

At 6200rpm it makes peak HP. At peak HP it is pulling 95% of peak TQ. Thus, it needs about 5% less fuel at peak HP as it dose at peak TQ.

Supprisingly not big differences in fuel delivery per stroke at different WOT RPM points. Most space cams can be adjusted to accommodate this w/o much issue. However, that dose not mean they will run very well at part throttle and idle.

Looking at it differently, say we had a pump that would support zero / idle to 200 ft lbs of TQ.

Lets say we put it with a motor that is has RV cams and architecture to make peak HP at 5500rpm at 95% of peak TQ. 190# TQ at 5500rpm is 199 hp.

Then we build another motor that has the cams and arcititecture to make peak hp at 7250 rpm. 190# at that rpm is 262hp.

Both motors pull 200# of peak HP but one makes 199HP and the other 262hp.

Could you use the same pump. Probably if you adjusted the space cam weights so they advance to deliver its peak fuel at each motors peak TQ points using different springs and weights adjusted to fit.

However, if you start to build a 3.6 mfi that will pull 250# of peak TQ. Will the above space cam work? You might be able to adjust it to fuel peak TQ but your idle is going to be very fat.

Thus, maybe we could build several basic space cams. 0 to 200# for typically a 3 liter motor, 0 to 225# for a 3.3 liter motor, and 0 to 250# for a 3.6 liter motor. Then adjust the space cam to put the peak delivery at each motors peak TQ by using differerent springs & weights adjusted to fit the motor.

With just a representitive TQ curve from a simmilar motor we should have enough information to set up a pump. Then the rack coud be field adjusted to get things near ideal.

Not an expert, just my thoughts. :)

How available are the RS space cams? Do you think there is a market for a clone of the Bosch cam? I would like to get a trashed one and run a material analysis and check the hardness and reverse engineer it.

Gary, Guss at Pacific Fuel Injection still has NOS one's @ $750.

Mark - if you don't mind me asking, could you tell us what you do for a living, or where you learned your fabrication skills (mad skillz says my kid!). I have to believe that your work as described here is exceptional.

Has Henry seen this thread???

daepp, Retired Fire Chief, but still run a Radio Communications Business that I started in 1988. I sell and service Public Safety radio communications equipment for the county. But 1/2 of the day I'm working on Porsche engines or transmissions and of course MFI pumps. Fabrication skills are "learn as I go".
Right now I'm trying to get all the tools made up for setting up the MFI pumps before they start the delivery tests.

Well, a few of you were right about my 1/3hp motor. Has a very hard time getting to 3000rpm at full load and I need to get up to 4000rpm on some of the tests. So I'm looking for a 1HP motor that will fit the same mounting holes. LOL on having it fit the same holes. Also, it's a little harder translating German from the Bosch PED 6 KL manual. I still don't really know if I have the "c" setting right, I think I do. Getting very familiar with setting these pumps up.

Can it be wired to 220v?

Hi Milt, Actually it is a 220 volt motor, but when connected up to the 110 volt variable speed control device it's now being controlled by changes of frequency and voltage to the windings of the 220 motor. This is how it's able to run from 0 rpm up to 4000 rpm and also able to program the standing start speed it milliseconds. I have located a 1 HP motor and will swap them out next week.
In the mean time; the #4 injector/piston stopped working on the pump I've been testing for the last few days. I mean the pump was working fine for all these tests then just like that, #4 is not flowing. I will be removing that assembly tonight and see what happened.

Mark,

I will be interested to hear of your findings on that #4 injector piston. I had the exact same thing happen to me (same piston even) during a track event a few years ago. It was difficult to diagnose on the spot, as the symptoms were a dead cylinder and none of the usual suspects indicated any problem. After verifying spark in all cylinders, I began pulling each fuel line in succession and putting them into a jar. The dry hole finally showed itself, and disassembly of the pump was required. Oddly, what I found, was a BROKEN piston in that #4 spot. A piece had broken off of the piston and the piston was "seized" in its cylinder, preventing any fuel from being delivered to that injector. Gus at Pacific ended up fixing mine, but I never did discover root cause for the broken piston.

Let us know what you find.

Thanks,

JA

I would think Herr Sattleburger, Bosch of Nurnenburg Germany, the gentleman that purchased all the MFI machines, testing machines, parts (the NOS RS space cams that Gus has...came from him) from Porsche back in the 1990's can help you on the test amd measure of the MFI pump. I stood next to his lead mechanic for 4 hours and watched him rebuild, test and dial in my 013 pump to factory specs.
Just google Bosch of Nurnenburg and click around until you see where they rebuild diesel and gas pumps. If you can't fine it...PM me and I'll dig out my receipt of work and pull off the telephone number. He speaks english...likes the US Army Soldier, thus giving me a great discount!
Just thought I could offer some help.
Vr,
Bavaria911

John, very interesting. Never heard of a broken piston happening before, but I guess anything can happen. I will let you know what I find.
Scott, Thanks for that information. I will pm you.

Mark - re #4

Not sure of the lubricating properties of Stoddard sovelnt vs fuel vs Shell calibration fluid....but I've heard many of the MB guys add ATF to their fuel systems to maintain lubrication and reduce gum buildup.

Hope you only encounter gum when you look into #4 !!!!!
Sent you a PM on "c"

Regards
John

This is a fascinating thread!

Mark, On the weekend, for no very good reason, (it was cheap), I picked up an Alfa Romeo Montreal engine-all four cams, dry sump, mechanically injected, eight cylinders if it. If you want to check an eight cylinder pump......:)

John, Very good point about the Stoddard solvent. I did not think about the lubricating properties. I was only looking at the Specific Gravity spec. of 0.70 to 0.80 so it would be the same as gasoline. This could be the reason #4 hung up for one of the tests. Lack of proper lubricates. I could not find a source for the Shell Test Solvent. Also, did not get your PM.
David, I'm only set up for 6 cylinders and would have no way of mounting it, but thanks.
Would like to see a picture of that pump.

Dimension "c"
 

Based on my interpretation of the test procedure (and my German is very limited !!) I think you're trying to achieve 2 ends with the compensation circuit.

Firstly, to set the correct RELATIONSHIP between the compensation components and their transmission to the main rack.

Secondly, to "LOCK OUT" the compensation mechanism so that it does not interfere with the rest of the calibration procedure. The target condition being "warmed up" 53°C (dimension "f" = 23mm) and 1 bar.

For the thermostat.... use the first 23mm tool in place of the thermostat, and then set the second tool (top screw and plate) to maintain the 23mm position when the main cover and thermostat (first tool) is removed. At this time, the "end point" for the thermostat compensation mechanism is also set by adjusting the hex nut at the bottom of the plunger (PED Section 4.2, Page 6, Figure 6, Item 10)

for the barometric cell.... I cannot find any reference to a fixed length at "1 bar" - only to its rate of expension = 0.58mm/100mmHg. This is where the RELATIONSHIP is important. My understanding is that with the thermostat at 23mm, and exactly 1 bar of pressure, the relationship between the components is established by dimension "c".

So at any KNOWN PRESSURE, a calc is performed to establish the necessary shimming of the baro cell (PES Page 12, Section 4.16,17, Figure 28,29).

So, in terms of locking out the influence of the baro cell, you have 2 options....always test at exactly 1 bar of ambient pressure (impractical), or replace the baro cell with a simple screw which can be locked to maintain the relationship (re dim "C") for the duration of the test/cal procedure.

Mark - this is my interpretation of the test procedure, and I'm happy to stand corrected !!

Let's know how you get on
John

Dimension "c"
 

John, Your interpretation is perfect. I just figured out why my measurements were a little off. Also not understanding German, you pointed out the "hex nut at the bottom of the plunger" that's the area were I can make the fine adjustment to get the "c" measurement right on. I couldn't really see that in my manual, thanks. It's a lot of work to make sure the pump is ready for the flow tests. I have 8 pumps and most of them I found only require the set up of the "f" and the barometric cell 1 bar adjustment, and the "c" is usually right on. I really need to come up with a better tool for the "c" measurement. That's a real pain to check for .8mm between those the two points.

Mark - I think you need to adjust "C" using the shims under the baro cell - the plunger hex nut locks the warm up end point.

To get the shimming right, you need to know the ambient air pressure and do the calcs & shim adjustment to deliver the correct "C"

Regards
John

Hey guys,
Bosch Koller&Schwemmer ( http://www.koller.de/index.en,195,51324660273781207206152433939315.html ) is very good but, they are known not only for Porsche MFI but all older inj systems and, as such, they are very busy which = not exactly cheap. I'd love an alternative as the cost has stopped me from doing more (I've had a complete 2.4 MFI system laying around for years...).

If you need any help w/translations, scan and post the sections here and I'll do it (small part of what I do professionally).

Then again, maybe I can exchange services w/Koller&Schwemmer - their English site is... not quite perfect :rolleyes:

Please, not 5 pages at a time, though, for the translation.

John, Thank you for that information. I figured that out real quick. I have spent a few days on my tool fabrications for checking the "c" measurement. Very easy and accurate to do now. All tests are done at 1 bar, and then after the testing is complete, the barometric pressure at the time is recorded and then appropriate shims are measured and installed under the baro cell. I did this yesterday on one pump when the pressure was at 29.92 (lucky me) for a few hours here in Seal Rock, OR which is at sea level. This made it real easy.
Steve, Thank you for the link to "Koller & Schwemmer". Right now I'm having the German Manual translated to English. First, just some of the important sections now, because it will take about a month or two to get the complete manual translated.