RobzWorkz #8f: Improving the 328 Fuel System: CIS Rebuild

Greetings FChatters!

Welcome to the 6th installment of RobzWorkz #8: Fuel System Refurb.

This thread will describe my effort at rebuilding the 328 CIS fuel distributor.

Again, I would like to thank Geert Jan Schreurs for providing the cast iron CIS distributor rebuild kit:

http://www.ferrari400parts.com/

The kit:

http://www.ferrari400parts.com/boschshop.php?view=productPage&product=40&category=12


Also available there downloadable pdf's for rebuilding the distributor; the 328 has the "adjustable" type. I also found a Porsche 928 distrubutor rebuild download:

http://www.porsche928forums.com/download/manuals/CISRebuild.pdf

These downloads, plus Charles Probst CIS book (http://www.amazon.com/Bosch-Fuel-Injection-Engine-Management/dp/0837603005/ref=sr_1_1?ie=UTF8&qid=1344094834&sr=8-1&keywords=charles+probst) were very helpful.



Now for rebuilding the 328 fuel distributor.

After removing the CIS distributor from the air flow sensor, remove all the attached stuff; this includes the supply and return fuel hoses, the frequency valve and damper for the frequency valve. Recall, I removed the fuel injector and WUR hoses earlier (remove these if not done). Store all removed components in boxes organized enough for you to recall the disassembly when rebuild time comes around.
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

Disassembly is fairly straight forward with a few detail notes:

1) before disassembly, identify with numbers and symbols, everything on top so as to be able to keep track of the parts. Take a picture after numbering so as to be able to reference when reassembly time comes;

2) on the bottom, mark the indent location on the plunger housing; take a picture of this for reference;

3) during disassembly, keep parts in bags identified with their numbers and symbols; I used small storage bags and made notes right on the bag and took a picture of each as well

4) disassemble without haste and look for small parts to dislodge.

Below are some pics showing the my markings on the distributor as well as examples of markings on the parts bags (details on the bags will be understood as the process is described).
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

Important notes:

1) hopefully the plunger did not fall out when you removed the distributor from the air sensor. If you did, you will need to clean and inspect for dents and nicks. Although replacement is ideal here, you may be able to dress clean any tiny distortions and reuse the plunger;

2) it is essential that the plunger move freely within its chamber...absolutely no snags;

3) also essential is that the plunger surfaces that reside inside the bore of the plunger housing be untouched; nicks here will change operation and warrent replacement.


To remove the plunger, bend back the retainer and simply slide the plunger out, taking care not to snag. On top of the plunger is an ESSENTIAL spring that needs to be kept as well.

While there, bend back the retainer tab far enough to unscrew the hex ring nut. Remove these.

Place the plunger/spring in one container for storage and place the hex ring and retainer in a different container (the hex ring and retainer are usually worn and rusted and need restoration; don't mix that with the precision-made plunger).



Pics are below.
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

After removing the plunger, the retainer and hex ring, the distributor can be placed bottom down an a clean surface so as to allow removing the stuff on top. One thing I did was to use a cardboard tube section on which to place the distributor so that I didn't set the bottom of the plunger housing onto my surface. You can do without the tube to support, just be aware and gentle (removal of top parts doesn't require a lot of force and you could hold the distributor with your other hand vice setting on a surface).

Get the following:

1. your individual containers ready with a marker to make notes;
2. a 3.5 mm allen wrence (9/64" should work);
3. a modest sized sheet metal screw;
4. a camera to take pictures of the organized parts (you'll want this in case the notes get smudged or something).

The process:
1. the banjo fittings going to each fuel injector and to the WUR should be loose from CIS hose removal; remove and place into a container marked with the number given before (I kept my copper washers for reference and test fitting...I put into respective bags as well...I replaced them with new on final assembly).

2. under each of the injector fittings should be a tiny strainer/filter; it can be removed by gently reverseing the sheet metal screw and lifting out; retain these for reassembly (I had two fall out when I first removed the banjo fittings and turned the distributor over...so watch out); as a note, replacement filters are not part if the rebuild kit (Geert Jan did have some but they were the incorrect shape, having "ears" that got in the way...definitely NLA); BTW, don't worry that the pics show other states of dissassembly...when I present these threads, I use my experience to assess the best disassembly routine for the description;

3. remove the adjuster covers with the 3.5 mm allen wrench; place in proper containers;

4. FOR EACH IDENTIFIED PORT, unscrew the adjuster with the Allen wrench, while COUNTING THE NUMBER OF CCW TURNS IT TAKES TO REMOVE COMPLETELY; when the adjuster is removed, label the container with the # of turns; this will be essential for reassembly when these are reinstalled with new o-rings; place cap, adjuster and washer in the properly numbered container/bag; take a picture of each port removal and its bag with notes for reference; all ports for my distributor required 10 to 10.5 turns to remove; BTW, in Geert Jan's download description, his procedure is to leave the adjusters alone until after splitting the body and removing the diaphram, measure the spring deflection for each, and record the length, and adjust to same displacement when reassembled.

A brief note on the adjusters: the adjustability can be seen by observing from inside the distributor. The adjuster, when turned, moves down onto a "top hat" piece that compresses a spring which pushes on a seat which pushes on the diaphram. Read about the theory in Probst's book. The last two pics here show: 1) the static set of emplaced seats (with springs and top hats in place) without installed adjustors all around, and; 2) the apparent pressure applied on the diaphram (as observed by seat being displaced) when one of the adjusters are installed to its recorded # of turns. BTW, I could not find a description of the proper adjustment of these in any of the test procedures or theory descriptions encountered: thus the only remaining method at reassembly would be to set each seat as it was when disassembled.
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

Remove everything else that can be removed from outside the distributor body.

This includes removing remaining fittings, the fuel pressure regulator and the 13 TORX bolts from the bottom. Place all fuel fittings in a marked bag. Place the regulator as an assembly in its own bag. Place the 13 TORX bolts with the hex ring and retainer removed earlier in its container.

As a special note: when the fuel pressure line was removed from the body, there was a joining connecter at the body which has a filter in it (it is referenced explicitly in the WSM torque specs). I removed this fitting when I removed the pressure hose itself and reinstalled when I installed the refurbished hose. All other fittings comprise the banjo type.

There is also a single hex screw to remove.

The fuel pressure regulator has a hex bolt/copper washer that can be removed from the outside and is shown in the pics; removing this is not necessary when removing the regulator from the distributor.

Also, it's a good time to remove the large external o-ring from the bottom of the body.

Also, please don't worry about the different states of disassembly here (you don't need to have the body split at this stage).

Pics below.
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

And precisely why this is not a DIY repair! All of the information, including the test jig/flow bench procedures are available through the Bosch inner circle. Without knowing and verifying flow/output, you're not rebuilding anything...but rather cleaning and replacing gaskets,etc. For the pittance that Larry Fletcher charges at FlowTech, why bother? And if your answer is for the fun of it; think about the ramifications of having a significant imbalance of fuel volume, cylinder to cylinder. The fact that "turns were counted" doesn't mean anything, other than that whatever error was there before has, at the very least, been repeated.

Fun to watch on the Internet, but not on my car....or the ones I service. Thank you.

 

Splitting the distributor body is not as easy as it may seem; you dont just tap an let it come apart after removing all the TORX bolts. The plunger carrier inside will have a lot of retaining pressure within since it has several o-rings distributed all around inside. Take this process fairy slowly with deliberate, patient effort.

BTW, since all o-rings will be replaced, you dont need to worry about them being deformed during disassembly (and you will...you must...replace all of the o-rings since you got this far anyway). It is possible that one or more o-rings will get sliced by the sharp edges of the diaphragm.

To split the body, be sure all the TORX bolts are removed, the 3 air senser bolts removed, and the hex nut/retainer removed. Then, using a flat bladed device (plastic or metal putty knife is good; a screwdriver is ok since this is the cast iron body; this is a NO-NO for aluminum alloy bodies: be careful there and minimize using metal), split the body at every corner and loosen until there is apparent freedom (can be moved a little by rotating like an OREO cookie...with about as much resistance, too). Now, pull and rotate so the bottom part of the body comes off; the plunger carrier will most likely stay in the top half since there are more o-rings holding things together there. Pull straight and rotate only a little and while keeping straight (don't want to damage surfaces and threads on the carrier.

You should notice that the thin sheet metal diaphragm chose to remain attached to one of the halves. If you will be reusing, try to keep any bending from taking place on the diaphragm metal as the body is being separated. Remove the diaphragm metal and store in a padded envelope with a piece of cardboard to keep from bending; the diaphragm should not be siezed onto any half; it should easilty peel away by hand with no bending (if it's siezed, then somebody glued it on with something "sticky", which is not good; the only previous assembly should have been with fuel-safe varnish or, perhaps, nothing at all).

Before taking out all the springs seats and top hats, take a few pictures so as to recall the parts arrangement when reassembling. Then take all these and place into a bag. It is not necessary to place each seat/spring/top hat into a container according to the numbering of the parts on top; these parts are very hard, the seats precisely made, and are pressed onto the diaphram, which gives into the pressure (even if you are reusing the diaphragm, identifying seats,etc. should not be an issue since: 1) the diaphragm has a unique orientation and thus can't be mixed up, and; 2) the seat springs will allow some form of realignment into the diaphragm's original seat indents; however, this concern should indicate that it would be better to replace the diaphragm when possible since there will be a placement error when reinstalling).

Two of the disassembly pictures show a close up of the tiny control pressure hole in the diaphragm plate; this must be oriented with the (slightly larger) holes in both body halves for proper operation to occur.

Removal of the plunger carrier can take some patience. My guess is that some carriers could be easily removed by modest pulling, but mine had significantly more resistance. Therefore, below is the process I used:

1) remove the fairly-large lower o-rings from the carrier and the (roughly) same-sized, lighter-colored o-ring from within the bottom body half;

2) place the bottom half back on and reinstall the hex ring to the carrier (provided the counter force);

3) using a well-taped screwdriver, lever the halves apart at various points about the body perimeter and lift out the carrier when resistance is low enough.

Pics are below.

BE VERY CAREFUL NOT TO DAMAGE THE SCREEN THAT SURROUNDS THE CARRIER'S BELT LINE! (I could not find a replacement and I didn't want to invent one)
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

I agree with that.

Thanks, David.

Below is the only procedure I could find on making these settings. I may address this in the testing thread later. I'm certain more precision would be made available if send the distributor in for adjustment.

Basically, recommendations I've read thus far for issues with the distributor comprise replacement only.

For now, I'm satisfied with how my car's running.
Image Unavailable, Please Login

 

Please recall that I am restoring a cast iron CIS distributor. Iron is quite forgiving when using metal edges (razor blades, #11 XACTO blades, etc) to clean the body. For anyone having a different metal, I recommend using non-metal edges (wood is great) and more chemicals (parts cleaner,carb cleaner, etc.). Anyway, I used an XACTO knife #11 to scratch off old paint and residue from non-machined surfaces and a flat razor blade to clean the external machined surfaces only. I also dressed the cast-textured surfaces with a dremel. For internal machined surfaces, all was in excellent shape, so I just used paper towels, Q-tips and rubbing alcohol (91%). After mechanical cleaning, I used carb cleaner to finalize and brake cleaner to dry out. All spruced-up quite well.

During "scratch-cleaning" I used a brush to sweep off debris.

Before any assembly, I took some long thin carbon fiber rods (I use these for other hobby work) and inspected all channel orifices for cleanliness and for being obstruction free.

Below are some pics.

Last pic shows a comparison of external surfaces. Obvious paint was removed (repainting to come).

I also cleaned the plunger carrier with carb cleaner only...no abrasive stuff here.
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

That setting and what your counting turns on are not the same thing; actually, not even close. The picture that you show depicts the process to set the initial opening point in which, ideally, all of the injectors will open/spray. This is immensely helpful when you've lost your (mixture) baseline due to either new components or someone who didn't know what they were doing was screwing with the mixture adjuster. This particular adjuster, in Larry's picture, has absolutely nothing to do with what I was talking about early; specifically matching the volume output, port to port.

 

True, and this is all I've seen in the literature available to the DIY'ers.

That's why I chose to not "alter" my mixture settings. The only abridgement should therefore occur if the new diaphragm presented a significantly different opposition to seating force. Otherwise, I would have started with a new baseline.

Perhaps you will be willing to share a source for describing theory-to-practice aspects of flow differentials cylinder-to-cylinder, the associated effects on engine performance and methods for adjustment and measurement to achieve desired performance.

I would appreciate your input. This would be fun and enlightening.

 

I just disassembled mine to replace the oring's. In addition several of the small filters at each injector output are more or less clogged. The problem with trying to adjust those small lift screws at each port is that they only indirectly change fuel flow by altering the pressure drop at the slit (and so fuel flow) by changing the deflection of the diaphragm between the upper and lower chambers. Those little springs are designed to provide .1 bar pressure drop between the upper and lower chambers without any additional input from the lambda valve or pressure control regulator (WUR). Adjusting the height of the spring via the lift screws looks to be a tactic that requires measuring the pressure at the injector port compared to the system pressure. I will probably send mine to Larry after I put it back together and let him calibrate it.

David I believe Rob does not post anything until he is finished with the project so I will be interested to see his solution.

Rock

 

After the time I have spent looking at this thing I am wondering exactly how accurate the flow rates are between cylinders. I would guess that a perfectly adjusted system MIGHT hit 5%. I would not be surprised if it was more like 10%. Just too many places where a little dirt or slightly out of spec component could make a difference.

Rob if you want to replace each lift screw exactly where it was before you remove it leave them in until you seperate the halves of the fuel distributor then depth mic to each top hat from the inside flat face before removing them. Much more accurate than just counting turns. Also I do not see any reason to worry about replacing the banjo fitting bolts in the same port from whence they came. They will have no effect on pressure. This whole system.is about pressure control instead of volume control like modern EFI.

 

The distributor body reassembly is straight forward:

a) prep: acquire rebuild kit
b) acquire replacement copper crush washers

1) replace all o-rings on the plunger carrier; place top body half upside down for assembly;

2) reinsert the plunger carrier (with the new o-rings on) into the top body half; be sure to set in same position as removed (indentation aligned as seen during disassembly; you will need to check this by looking at the bottom half with the pics taken of your disassembly; be gentile on insertion so as not to overstress the new o-rings of flip the out of position; I found that mine was fairly easy

3) place top hats, springs and seats in the injector line holes from;

4) check fit the diaphragm plate, noting the tiny control pressure hole is aligned with the same hole in the body; make sure plate lays flat; be gentile and don't scar the carrier body o-rings (remove these if necessary; I actually sliced an o-ring here and had to replace with new from Ace Hardware...ooops);

5) gently remove the diaphragm plate, apply Indian Head varnish onto flat surfaces where the diaphram will lay; reinstall diaphragm in same position as before; I used a smaller sponge brush to apply the Indian Head varnish (the larger one in the bottle was too untidy);

6) apply Indian Head varnish on lower body flat surfaces that will lie on the diaphragm plate;

7) install new lighter-colored o-ring inside lower body; if removed, reinstall carrier body o-ring(s);

8) place lower body onto upper body assembly, being sure to align control pressure holes; use even pressure;

9) install the 13 TORX bolts screwdriver-tight; I placed some varnish on the threads;

10) reinstall plunger with spring;

11) (restore the retainer and hex ring by cleaning and polishing) place retainer and install hex ring nut (smear some Indian Head onto the nut threads; use large gapped adjustable wrench and tighten to "mild wrist" tightness); fold down retainer to hold plunger while allowing ample, non-sticking plunger movement.

Pics are below. You will notice some differences in procedure order because I derived this order after the operation as being better.

Also are shown more pics of cleaning the plunger carrier as well as some showing the very fine slits for fuel metering.
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

The F specs vary a little depending on whether it's at idle, partial load, or full load, but it's roughly no greater than a ~10% difference between the minimum to maximum injector positions. However, whenever I've measured the fuel delivery from FDs received back from Larry F. (and it's been several), they've been no more than 1~2% different between injector positions (and the accuracy of my measurement equipment/method is no better than that -- for example, with the airflow metering plate held in the approximate idle position, the amount of fuel delivered in 3 minutes into 100 mL graduated cylinders would be: 21, 21, 21, 20.5, 21, 21 mL):
Image Unavailable, Please Login

 

Rock, I agree to a point. You describe the method invoked on Geert Jan's site. Please see the below pic, which is a screen capture of an exerpt from his downloaded procedure (not for 328, but for 6 cyl). I believe the measurement tolerances for this type of length measurement are about the same as with the #turns method. Also, recall that the "top hats" can have error (if distorted at all and not positioned under the adjuster in exact same orientation). This and other random error effects on measurment will pose unmitigatable measurement uncertainties. I thus believe the most accurate way to adjust these is with flow measurements.
Image Unavailable, Please Login

 

I haven't done this test yet, but this setup seems an appropriate one. I've seen some videos of other vehicles doing similar testing as well as for bench testing setups. My only issue is that the available volumes are not suited for much more than idle-speed sensor displacement (fuel flow from a pressurized open injector will fill the cylinders quickly). It would be nice to see if I could place flow meters in each injector line and optimize over engine speed with a running engine. Perhaps the flow control setup would be similar to something attached to the CIS distibutor top with all the lines going to their respective injector, having some analytical flow measurement device at a centralized hub while conserving access to the adjusters.

 

Now time to take my beautiful wife to a well-deserved dinner and watch a movie afterwards. I will complete this thread tomorrow.

BTW, testing the CIS will have its own thread.

Have a good evening, All.

 

Rob not what I was saying. Do not measure the spring height as the measuring device will cause some spring tension. Remove the spring and cap and depthmic to the spring base which will be solid. I have a suspicion that if you set the spring tension at 1.5 pounds with the cap flush to the face of the FD half you will.be correct. After all the only reason for the adjuster is to get even tension on all the springs without having to fiddle with shimming the spring base for each port which would be very time consuming.

Probably Larry is laughing at me thinking I am fuller bull.

Rock

 

This test should never be done with the injector(s) putting liquid fuel into a non-running engine -- ALL of the injectors served by the fuel distributor must be removed and sprayed into an individual measuring container. At idle, you need to use a long enough time for the amount of fuel collected to be not infintesimal (I used 3 minutes). At greater airflow plate deflection, the collection time can be reduced to keep the total fuel volume collected reasonable (for example, partial load = 40~50 mL in one minute, and full load = ~50 mL in 20 sec).

 

I think I see your point...hard to measure spring tension on a length measurement with a finger and good intentions...now I think we are all fulla bull.

BTW, my assessment of direct length measurement error is 5-9 mils; same error with screw turns and a 3mm deflection (optamistic), at 11 mils per turn (single turn magnitude) => ~3-6 mils error per half turn. When including random errors in measurements, these are insignificant in difference.

My car runs and drives well, the injectors inject (I measured that) and I don't get fuel fumes at idle. All I have left to fix now are the annoying exhaust leaks I discovered after removing tha cat...air injection will disappear soon enough.

 

Sorry Steve...I meant engine running...no filling the crankcase with fuel allowed. Idle testing with containers is what you are describing, and seems acceptable for just above baseline checks; I was looking for a running engine flow metric analysis for flow equalization over the RPM ranges. This is where the stronger inequalities will reside.

Maybe I'll invent this or see if the Bosch Gods have already done so in a lab somwhere...probably stored it in a vault in Germany along with some of Max Plank's old testing apparatus.

 

voice from the peanut gallery: DAMN nice thread! WOW!

Great work, thanks for documenting and sharing!

 

+1

And the impromptu "peer review" system seems to be working well

I am wondering if it would be possible to build a bench test rig for balancing the CIS...

 

I've read the Probst book twice and now with these pictures I'll read it again so I can really see what's going on. I do know however that I'll never do this. I have enough trouble getting carburettors to work correctly.

What I will do is try to drive my car frequently so I'll keep these passages as clean as possible.

Thanks for sharing your process and pictures. Thankfully for me I know when something is far out of my league. I just like understanding how things work.