3.2 Motronic engine running rough backfiring
 

I wish to thank people associated with Pelican Parts for the efforts in service I got during the building of my Project Sixer car. The car was started to day after more than five years of building from the excellent parts at Pelican. Today was a great day!!!!

I do need some more expert advice and guidance since I am getting into the fuel injection which is over my head.

I have been building a 914-Sixer with a 1987 Carrera engine. I know it is not a 911 but I do need the expertise of the 911 crowd. The engine has been setting for about five years during the time I have owned it. The oil was primed by hand and then by starter over a period of a month. The fuel pump was verified in working order by jumping prior to starting without the ECM installed; the same for the coil; found reference sensor was not holding its electrical connection; etc.

The engine has 145,000 miles with excellent compression and leakdown numbers. The compression numbers are140-145 across the board at sea level; leakdown numbers show "0" in four cylinders, 1% in one cylinder, 3% in one cylinder.

It appears that I over serviced the engine and will be having to waite for the excessive oil to burn out. Is this alone enough to cause the engine to run rough, backfire and miss on one bank, appear to run lean, etc????

Or, it this something else such as fuel injectors???

As far as I can tell the injectors have been on the engine a long time and have been sitting without running for considerable time. The oxygen sensor is new along with the rotor, cap, plugs, etc. It appears the engine wants to run occasionally since it appears to run on all six cylinders several times for 30 seconds or so.

In sum, I started the car with starting fluid; got it running rough; I did over fill the oil tank; one bank misfires with flames; engine was taken to 1800-2000rpm; wants to run smooth at times then starts running rough and missing/backfiring.

Definitely over my head on this one. Need advice!!!!

Steve Hurt

Try the "in the Dark Garage and use a spray bottle of Water and mist while car is on" this will determine if you have errand sparks especially at the miss-firing cylinder.

Drain off the oil instead of burning it off

If you have a Cat By Pass on the 3.2, she will back fire.

While sitting did it have old fuel in the lines? If so that may have clogged the injectors. A test, move the injectors around to see if you can cause the miss-fire on another cylinder.

Jim

You may want to consider throwing some BG44K into the fuel system to help rid some contaminates. Make sure it is proper ratio, do not over do it. The Cleaner needs to go through a full tank to be effective.

Check for vacuum leaks as well.

My 2 cents

The reference sensor must send magnetic pulses to the dme or it will misfire. Exactly where are you seeing "flames"? How much did you overfill the oil? Is the motronic in stock form, or modified? Is the fuel pressure regulator vacuum line connected? What was the fuel pressure when you jumpered it? Old gas? Correct 3.2 fuel pump, or a 914 pump?

DRACO A5OG and rusnak:

Thank you for your rapid response. I will try to answer your questions below:

DRACO A5OG:
Oil will be drained off instead of burning off.
Does not have a CAT bypass; running headers through short exhaust mufflers
Engine was not connected to any fuel supply during storage; drained lines and plugged
Will try to move injectors around to see if the problems change
Will use BG44K cleaner into fuel system following exact directions
Will keep checking for vacuum leaks; have not found any thus far.

rusnak:
Reference sensor and speed sensor are both new.
Flames out of one tail/header pipe (911 installation left; 914 installation right side)
Over filled the oil tank by I am guessing 3 quarts; it appears I cannot read the dip stick
Fuel pressure vacuum line??? Will check; this is something new to me.
New gas - Chevron 91 octane
Correct 3.2 fuel pump

I like these suggestions and will also follow them closely.

Thank You very much

My guess is sticky fuel injectors from sitting for 5 years. And remember when reading the oil dipstick the motor needs to be up to temperature and running/idleing.

Thanks, rfuerst911sc.

I am starting to think along the lines of vacuum leaks and sticky fuel injectors. I woke up the morning (4.00 AM) and I think I left one end the line open at an added vacuum gauge. Also, I have the Auxiliary Breather System to add after getting the engine running. I may have left one of these vacuum ports open.

I have drained the tank down below the engine breather line in the tank and also still showing on the dip stick. I will be watching this level to see if I need to take any more oil out.

Is there any way of getting the oil out of the intake/breather system inside the engine other than burning it out???

I am going to try some injector cleaner as suggested but in the long run I think I will pull the injectors and send them to a qualified person for cleaning, testing and general reconditioning.
Does anyone have any suggestions on where to send them??

Any and all suggestions/recommendations are greatly appreciated.

Thanks again,
Steve Hurt

Some strange comments. Be sure you are using an oil tank meant for a 914-6 with a properly calibrated dipstick. Get the whole engine together and vacuum leaks fixed before attempting to start the motor.

are we absolutely sure the spark plug wires are hooked up properly?

Ok, here we go.

I am using a 911 1987 Carrera oil tank so the fittings etc are made for the 3.2 engine I am using. Fact came from same engine; there was considerable oil in the intake breather area and out the tail pipe. Plugged the vacuum to the vacuum gauge and started car. The roughness and missing went away after running for five minutes. It was still backfiring on the one bank; let the car run longer and backfiring appears to have gone away. smoke from the oil has been substantially reduced with very little coming out of the tail pipes. Engine sounds strong but still needs to be run to clear out the oil

The two problems appear to have been the vacuum leak and the over servicing of oil. I am still going to pull the injectors and sent them for cleaning, testing, etc. I thought I had all the vacuum leaks plugged before starting, but by starting i found one along with the oil over servicing. The suggestions made on this thread will be checked individually to make sure the car does not have any problems.

Thanks everyone for the comments and suggestions. It put me on the right track to get the engine running.

I have the entire 1987 Carrera engine, oil system, front oil cooler with fans, etc being used on this car. Car is running headers with M&K short mufflers. The fuel pump, DME, altitude sensor, harness, DME/fuel pump relay, etc came with the engine.

Just a shot, but mine would have muffled backfires/gurgling on deceleration - I thought it was normal, until I found/fixed an ICV (return path xsistors on DME) problem, now they're no where as near frequent/pronounced.

Here are a couple of places for injector cleaning. I used Witchhunter. Ken

RC Fuel Injection

WitchHunter Performance - Injector Cleaning & Flow Testing Services

not an attempt to steal the thread, but I'd like to know more about how your problem was solved.

I agree with Kidrock, I also would like to know more about your problem and how it was solved. This looks very interesting.

Steve Hurt

me 3

also...

I dont think it is resolved yet... this is a new thread.. but I too am watching :p

As for the OP's problem, I think it would be good to borrow a stock 3.2 911, and observe the original fuel hoses, vacuum hoses, and learn how the system works. Sounds like the fuel pressure is not equalized from left to right.

Sorry I'm 'late' - I didn't come back until now. If your ICV vibrates, this may not be the problem.

I only owned the car a couple of years and was convinced of a cold start issue. I worked a lot of things (vac lines, injectors, etc) with no luck / eliminated CHT as a prob etc. I had no idea what the ICV vibe was supposed to feel like - figured it was muffled by the engine vibe. I tested the +12v to it and it was ok, but looking at the schematic I found that the ground is switched. (+12 had no place to return to).

So I made up a fake ICV coil (load resistor) and looked for a signal and one side was missing indicating no drive to the ICV. So it was either stuck in one position or the other all of the time.

I found a thread that helped, and started my own thread, but interest seemed nil, so I never updated it. I post both threads here, and can provide more info if Steve thinks it's helpful.
One transistor needed to be replaced.

Let me know if you want and I will fill in the blanks on my thread or PM me.

HTH

landwerlen's
http://forums.pelicanparts.com/porsche-911-technical-forum/346667-motronic-idle-surge-solution.html


mine
http://forums.pelicanparts.com/porsche-911-technical-forum/573245-3-2-cold-start-issues.html

Damn Dan,

That is Awsome, will to talk my DME friend about this.

Wow,

Jim

Thanks Jim - but I'd be lyin' if I thought that this is Steve's (only) issue - I never had flames, and all my stuff was stock to begin with. BTW I replaced both original transistors as a precaution. My icv certainly vibes now. I'll shutup now :-)

Oh No Brother, Keep Writing, this is very intriguing!

Need to open her up and take a closer look inside.

WAAAYYYYY too technical for my pea brain. Methinks I may just send the whole board in to a specialist just to have everything checked out at some point in the future....after all, I'm pretty sure the DME isn't built to last forever. Thanks for the info anyway, Steely.

Dan, thanks for a considerable amount of information regarding the DME. I am still proceeding along the suggestions I have obtained from this thread. I am in complete agreement that I may have problems other than the over servicing and vacuum leaks I have found. I am always suspicious of corrections that seem far too simple; I always am looking for other causes and will proceed through the thread suggestions with this fact in mind. I may have other causes as has been suggested. Just let me know what you think may be causing the flames and backfiring.

I am going down the list and will make sure everything on the engine is operating correctly. This is where the project building process takes on interesting situations. Thus the challenge is to learn, understand, and keep going until the engine problems and related problems are solved.

What I have done is to build an entire 911 Carrera front suspension, engine, clutch, transaxle, brakes, drive axles, rear bearings, etc. into the 914 chassis. This includes the 911 Carrera oil tank. I am running headers with short M&K mufflers until I get the engine straightened out and then I will have special mufflers and cats put on the car. I am building this car to operate in California since I will eventually move it there.

Dan, I find your information very detailed and somewhat informing. I think it is something that we all have to address since our cars are getting older and the DME is definitely getting more scarce. I have seen further developments come out of the information that you have been writing about thus creating more understanding and practical application.

I am going to take the information given in this thread and digest it including all the possibilities that could be causing my problem. I am going to the Parts Heaven Parts Swap this weekend. The traveling from Flagstaff, AZ along with the information provided in this thread should give me time to think about the problem(s).

Please keep the comments and ideas coming since they are definitely helping me to under the DME system and its components.

Steve Hurt

Somewhere I've got the Dme Test Plan. Might make good airplane reading material. You need an ocilloscope though.

I have been going over items that are not stock and what I have changed. The engine is running two separate headers matched to the engine. The right bank (911 installation left) is operating with the oxygen sensor and this is the one that is not backfiring with flames; the left bank (911 installation right) is operating without a oxygen sensor (I am setup to change the oxygen sensor from one bank to the other) is the one backfiring with flames. In addition, I have built in ports for gas mix measurement on both sides.

The original 911 exhaust had both banks going into one exhaust tube with the oxygen sensor before the CAT.

Wondering if this change in the exhaust system could be contributing to the problem??

Thinking of changing the oxygen sensor from present location to the other side. What do you think??

Thinking of doing a complete gas analysis for both sides. Again, what do you think??

Just trying to get as complete foundation to the problem as possible before concentrating on the problem.

Steve Hurt

The 914 has a fairly long tailpipe. I'd install the bung in the tailpipe, before the muffler. I doubt changing side will do anything. You should have a complete loop between the fuel rails. Do you have this? Is the fuel pressure regulator and fuel pressure dampener vacuum control working? Idle position switch? Wot switch? These things get checked every engine installation. Etc etc. I would at minimum connect an air fuel ratio gauge and check closed loop operation, and do a lean stop test, and rich stop test.

Thanks, rusnak.

Those are some of the checks i want to hear about. I have a complete loop between the fuel rails. I like to know more about the tests for the fuel pressure regulator, fuel pressure dampener vacuum control, idle position switch, WOT switch, etc. I have done somewhat a minimum tests on the fuel pressure and idle position switch. I am glad you mentioned these and the others. Yes, I agree they all need to be checked in greater detail. This now has become one of my initial objectives. I am sure there may be other items to check. Will be looking for these; included will be the air fuel ratio, close loop operation, lean stop, and rich stop.

Thanks,
Steve Hurt

I see. You might check out the "Porsche 911 Carrera Service Manual" by Bentley Publishers, and "Bosch Fuel Injection and Engine Management", by Charles Probst. Two excellent, excellent resoueces. Usual suspects - Amazon or Ebay are you best bet. And beg borrow or steal an LM-1 wideband fuel air instrument by Innovate.

An update on this thread since many valuable sources of information and ideas were provided by many.

The engine is up an running without backfiring out one of the banks. It appears I had several vacuum leaks involving the systems. Engine runs strong and with authority. I have been going through the checks involving the Motronic systems and they are checking out. I will continue to do these complete checks including all the ones suggested in this thread.

I have one characteristic develop: the engine starts and runs very well during cold and hot status. It now has taken on a characteristic of wanting to shut down both under load (running down the highway) and when in neutral. The engine seems to go to shutoff position and then comes on strong when it "catches itself" in the shutoff position. It appears to be getting gradually worse.

Need some input on the above characteristic problem.

Thanks again for all the foregoing information and I really liked the way the thread took off and people contributions.

Even got a source of information involving the DME at high technical levels and maybe I can find a source for total repairs of it.

Thanks again.

Steve Hurt

Check your dizzy rotor for slop.

You may also want to check the AFM, to insure you are are gettig proper readings through out the travel.

Also, I assume she is trying to shut down when gas pedal is not depressed, correct. I would check the idle position micro switch at the throttle body.

Check also the Fuel Filter strainer is not clogged with debris.

"The reference sensor must send magnetic pulses to the dme or it will misfire."

- rusnak -

Totally incorrect as usual! The 3.2 DME ECM only needs the reference signal
to start. Once the engine starts, the ECM counts the speed signal teeth for
all the needed info. If the engine requires the reference signal after it starts
to run properly, then the speed signal is most likely bad, e.g. bad speed sensor.

Loren, I suggest that you actually try it with a second reference sensor out of the car, as I have. You will have to retract your statement. The dme needs signal beyond startup from the reference sensor. You can also reference JW's thread on his "no start" related to a missing ref signal. He also confirmed that idle was misfiring without reference signal, so you may step down from your high horse now, your Royal Highness......

Oh No, you didn't? :D

Shoot, I'll even make a Youtube video and prove my point if he makes me.

I've got your back, stud. :cool:

"I suggest that you actually try it with a second reference sensor out of the car, as I have."

- rusnak -

Sounds like someone's located a major problem with their engine management
system!

Been doing that for over 15 yrs with all types of BMWs (528s, 533s, 735s),
Porsche 944s, and 911 3.2s, as they all have the basic same early type
DME ECMs, i.e. running the engine without the ref sensor and having it
run with a perfect idle and rev perfectly also.


Bottom line: READ THE OTHER THREADS WHERE THE FLYWHEEL PIN
PROBLEM OCCURRED AND A SECOND REF SENSOR IMMEDIATELY
STARTED THE ENGINE BY JUST A MOMENTARY WAVE NEXT TO METAL
AND WASN'T NECESSARY ONCE THE ENGINE STARTED.!!!!


"You can also reference JW's thread on his "no start" related to a missing ref signal. He also confirmed that idle was misfiring without reference signal"

- rusnak -

Right, because the timing of the ref signal was NOT at the right TIME.
If he had done it again and luckily found the right timing, it would
run perfectly. The DME ECM needs the correct point (ref point) to
start its count for the next TDC, if not then the timing will remain
either too advanced or too retarded.

Good luck finding your problem, as this should help you!!!!!!!!!!!!!!!!

So much mis-info gets posted here where DIYs just don't fully analyze
and understand what actually happens and GUESS or re-post some
other hyperbole they found.

Loren,

Now I caught you just making stuff up. Talk about misleading!! You're a hypocrite!!!!!!!

JW had NO SIGNAL FROM THE REF SENSOR!!!!! THAT WAS HIS PROBLEM!!!! (recall that his customer flywheel had a non-ferrous pin. He chose to re-drop the engine and change the pin back to original)

Note, that when he artifically supplied signal by pulsing the ref sensor, the car started.

Then, in his words, he had to keep pulsing the ref sensor AFTER THE CAR WAS RUNNING to make it run smoothly!!!!!

Now, again.....you are caught in either (1) an outright lie, or (2) you are being lazy and shooting casually from the hip. In either 1 or 2, your tech info is misleading, incorrect, and should be avoided at all costs.

(how's ^this for my imitation of Lorenfb?)....haha!


edit: Here's the original thread for entertainment value: http://forums.pelicanparts.com/porsche-911-technical-forum/588695-86-wont-start-after-new-flywheel-no-injector-coil-pulse.html

"Then, in his words, he had to keep pulsing the ref sensor AFTER THE CAR WAS RUNNING to make it run smoothly!!!!!"

- rusnak -

WRONG!!!!!!

The 1st pulse did not provide the right timing reference.
A few more random pulses finally got the correct reference
point and then no more pulses were needed.

Stop being such a hard-headed DIY and try unplugging the
reference sensor on a good running car and NOT yours which
needs to be properly worked on by a qualified Porsche tech
who understands Porsche electronics!

Loren,

WRONG!! stop being a hard-boiled egg-head and get out into the real world and look at cars, not books. While you're at it, actually read JW's post.