Engine misfiring after fixing leaky injector pump valves
 

I bought 1993 300d, 5 cyl turbo, with 115k a couple weeks ago. It ran very well but had a leak from the valves on the injector pump. Last night, I changed all the o-rings and crush washers as described in the wiki diy here. After restarting, it is running very rough (chug-chug) and smoking. I am guessing that maybe one of the injectors got clogged during the work on the IP, or is it possible that I did some other damage to the IP or it's timing. From researching similar issues on this site it looks like the smoke is unburned fuel from misfiring. I was thinking about backing off the fuel lines on each injector while its running until I find out which injector/cylinder is having the problem and then inspecting/cleaning the injector and compression test the cylinder. I was also thinking it might be trapped air in the line and not sure if I should just keep running the engine for a while or if this might make something worse. Any ideas?

are the injector pipes in their correct places on the pump?

Good question. I want to say yes. But I'll check again when I get home. That would be a nice easy fix.

Pass on the cylinder compresion test if the engine was not smoking before you changed the element seals. A very much higher percentage factor is it has something to do with you changing those seals.

Good point Barry. I'll keep the compression test lower on my priorities. I'd rather wait to take the injector out also. I think I'll bleed the air more carefully when I get home and go from there.

That is a good idea.

Loosening the Fuel Injection Hard Line Nuts at the Injectors and Cranking is also a fast way to get rid of any Air in the Fuel Injection hard LInes.

Another note as suggested above: Anytime your vehicle displays new issues after you just worked on it, there is a good chance that the new problem is related to what you just did. This doesnt always mean you did something wrong, but perhaps need to do one more step.

Question: why is it so common that people want to check compression right away? Perfectly good running motor, something simple happens and compression is thought to be an issue. I understand that this is a standard test, but if something ran fine a while ago, it takes a HUGE failure to have a sudden effect upon compression.

OK. I opened all the injector ends of the hard lines and cranked it for a few 20 second cycles, then re-tightened the lines. There was good fuel from each hardline. Then I started the car and slightly opened each of the injector ends of the hard lines for 20-30 seconds after I saw some fuel.

It still has the same smoke and rough running sound. I made a video and put it on youtube for you guys. The quality of the sound was distorted but I hope you can hear the pattern of the firing when I open it up a little. You can also see the color of the smoke- a mixture black and grey http://www.youtube.com/watch?v=CGeh9fRriSE It had zero smoke 2 days ago.

I'm new to diesels, and I'm used to working on cars alone, trial and error. So, the ideas are much appreciated. Using my old way of thinking, at this point, I'm inclined to remove the injectors and try inspecting and cleaning them. I have reopened the valves on the injector pump to make sure everything set right. I'm not sure what else could have gone wrong beside debris getting into the injectors; one of old IP valve o-rings was falling apart. Other ideas?

-Mark

Did you torque the delivery valves properly???

When you re-opened the Valves to check them that means you should have also changed the Copper Crush Washers again; they are supposed to be a 1 time use item.

I thought you had a handle on the Injector Cut Out test?
Loosen one Fuel Injection Hard Line/Tubing Nut at the Injector and pay attention to the reaction. If the Injector is good the RPMs should drop and it should be noticeable.

Tighten that one up and go to the next one and do the same.

When you find the bad one when you loosen the Line Nut at the Injector the RPMs will not change much. So now you have a suspect Injector.

Take the Suspect Injector and exchange positions with a good one. Re-do the Injector Cut Out test on those 2 Injectors. If the Problem followed the Suspect Injector to the new position there is a problem with the Injector.
If the problem stayed on the same Position as before you have a problem in the Engine or in the part of the Fuel Injection Pump that supplies that Cylinder.

If you just changed out the delivery valve seals and crush washers, you would not have changed the timing of the IP. If it ran well before you did the job, then I would suggest that your re-do the delivery valve crush washers and seals. It is a cheap re-do as far as parts go, and just takes a little bit of time. It should also be noted that even when the job is done correctly, it sometimes takes a little time for all of the air to work its way out of the system. Parrot of doom put out an excellent DIY on changing out the delivery valve seals, and I would suggest you pay particular attention to the part about making sure that the lower elements are installed correctly, and the part about installing the valves by hand and rocking them as you screw them in. As explained the the piece, it helps center the elements as the valve is screwed back in. Once the valve is hand tightened, then you can torque it to specifications.

Somebody has the torque specifications for delivery valve seals?

Right you are sir, I missed it on my first trip through that thread. Thanks.

Thanks for the ideas. I couldn't work on the car tonight because I had to take one of my boys to baseball practice. To address the torque issue, we stopped on the way home and got a torque wrench that will fit correctly inside this part of the engine compartment. I gather from the responces addressing torque that an exact torque may be very important.

Thanks to diesel911. I like the way you are thinking on this. I did notice one hardline not having as much of an effect on the idle, but I was looking for NO effect and ignored this until I read your reply. So, we picked up a 27mm deep socket on the way home also to remove an injector or two.

Renntag also had a good point, among others, "This doesnt always mean you did something wrong, but perhaps need to do one more step." I thought about this on the ride to work. I was hoping that bleeding the air carefully was that last step, but hey, even if it is a big job, a big job is just a collection of small steps. Not to get too philosophical, but we never know when it's going to be our last step either, so let's enjoy the ride.

It costs 37 dollars at my local dealership to pick up 5 o-rings and 5 crush washers for this job. So, I'm dragging my feet on replacing the washers. I did read, on a reply to parrot doom's post on the subject, that changing the washers had a great effect on the idle. Anyone got a lead on some reasonable washers?

Here's what I wrote about my experience. Pay special attention to the part about the chintzy little plastic fuel fitting.

http://www.peachparts.com/shopforum/diesel-discussion/88574-finally-step-step-glowplug-change-98-99-e300-post2406544.html#post2406544

As for the parts, I got mine from peachparts, they were cheap. I bought 7 the extra was insurance against losing one.

I had to redo my o-ring/cursh washer job twice before mine ran well. I had the exact sypmtoms you are describing. The o-ring would actually tear on the threads as it was going in, thus not making a seal and allowing air into the system. Basically you need to determine which cylinders are getting air, then go back and redo those o-rings and crush washers, and you should be fine.

"It costs 37 dollars at my local dealership to pick up 5 o-rings and 5 crush washers for this job. So, I'm dragging my feet on replacing the washers. I did read, on a reply to parrot doom's post on the subject, that changing the washers had a great effect on the idle. Anyone got a lead on some reasonable washers?"

If you did the job the first time without a torque wrench, your crush washers are worthless. When I did my delivery valves, a number of posters on this forum said the copper washers did not need changing. I did the job the first time, without changing out the washers, and the car ran like crap. Prior to doing the job, the engine ran as smooth as silk, and after I did the valves the first time, it was nailing and idling rough. I redid the job a second time, only this time changing out both seals and washers, and the engine once again ran as smooth as silk. As another poster has said, if the engine ran fine prior to you doing the job, if the job was done properly, it should be running just as fine now. If I were you, instead of messing around with the injectors and hard lines, start from the starting point of the trouble, which is the delivery valve job you did, and redo it according to specs. If that fails to remedy the situation, then move on to the other areas mentioned in prior posts. Good luck

I took pimpernell's advise and re-did everything with new o-rings and washers and torqued it all to spec. It still ran poorly. I added a bottle of Diesel Kleen. I took it out on the highway for an italian tune-up with no luck. So, I stopped and added another bottle of diesel additive that I found and drove some more. No luck. Drove some more and added a third bottle; nothing.

Then, the next day, I tested the injectors by backing off one hard line at a time. I found that two injectors (1 and 2) had a slightly different effect on the idle when I loosened the lines. So, I took them out and switched them. When I restarted the car, it was still misfiring but it seemed to be misfiring differently. This seemed like a positive sign. I also found some residue/dirt/grease inside the end of one of the hard lines.

So, today, I took out these two injectors and soaked them in carb cleaner for half the day, and then replaced them. I blew the hard lines out with carb cleaner. I started the car and it still ran poorly, maybe worse.

I decided to take a crack at the fuel injector pump valves once again. I'm very skilled at this now and reset everything with still more confidence. I pulled out the pieces beneath the crush washer on valves 1,3 and 5 to check them because I accidentally pulled one out last week the first time I did the job. Everything was fine. But, the fuel in the pump seemed darker or more oily than i remember.

Nonetheless, I reattach everything, I slightly over torque the valves hoping to get a little extra crush out of the twice used washers. Then, I bleed the lines and crank it up. Here's the kicker. It hesitates to start, and I am about to stop trying when it fires up and runs wide open all by itself- no pressure on the pedal. I shut the key off and it is still running wide open. I get out to push down the stop level and the lever is already down (as it should be when I turn the key off). I check to see if the pedal is stuck down; it's not. With no other options, I pull the fuel line at the inline filter and in runs out of fuel 20-30 seconds later (it felt like 5-10 minutes later).

Feeling totally confused now about how I turned this easy job of fixing a leak into an engine raging out of control. I decided to write a note to the forum, being new to diesels, maybe all this makes sense to someone else who is not new to them.

My first guess is that the fuel additives I added didn't mix well with the fuel and entered the fuel line in concentrated form. My second guess is that the fuel injector pump just failed somehow, related to me pulling out the parts beneath the washer and/or possible oil in fuel. Although the oily fuel might have been related to the additives. When it was wide open, it was pouring out black smoke to the tops of the trees, and it was running good.

Classic RUNAWAY ENGINE!
 

Save yourself some money, time and heartache. Locate a known working IP and replace your pump. Better, pay somebody to replace your pump. Or send to get pump calibrated by someone familiar with these pumps and then pay somebody to replace your pump. You are where I was about $17,000.00 and 10 years ago EXCLUDING LABOR trying to learn these cars on the cheap.

BenzDiesel

Thanks BenzDiesel,
I really appreciate the response; it shed light on the confusion.

After I did the related searches on runaway engines, I'm inclined to follow your advise straight away just to save the time except that I want to understand a little more. How exactly do I know the IP is dead? Could it be a lift pump or bad fuel? or does the IP take the lift pump with it on on the way out?
Thanks again.

fuel lift pump causing runaway engine and rough idle?
 

Having a chance to think about own question again, a better way to ask it might be, could a malfunctioning fuel lift pump have caused both the poor running and the runaway engine? I didn't bleed the air out of the lines well the first time I changed the valves, and gave the engine a very hard time trying to start it. So, could I have damaged the fuel lift pump which may have already been suspect? Is there a way to test/check this?

Thanks.

"Nonetheless, I reattach everything, I slightly over torque the valves hoping to get a little extra crush out of the twice used washers."

I don't mean to be harsh, but you started out doing the delivery valves on a perfectly good running engine. You finish doing the job apparently without a torque wrench, and the engine runs poorly. You have been advised to put in "new" crush washers, and to torque to specs. You decide to "slightly over torque" the valves and use the old crush washers again. Now your talking about the lift pump, removing the I.P., flipping hard lines and the like. If you used the same crush washers, and over torqued the valves, there is a good chance the run away engine was caused by your hit or miss approach as it pertains to the I.P.. The delivery valve job was a piece of cake compared to changing out the I.P., which requires special tools and a good grasp of how to set the timing up correctly. Good luck.

I think you went too far into the elements of the injection pump and they are no longer indexed internally to the element control mechanisim. In otherwords the elements are somehow or other wide open.Or off their control mechanisim. Should only be the ones you removed though. It is not the lift pump or bad fuel for sure. When you shut the shutoff down it was unable to control those parts you had out. Thats why it took removal of the fuel feed line to stop the engine.

gsxr wrote up an explanation on the proper sequence for installing NEW sealing washers:
"the 3-stage torque procedure for the barrels is 30Nm, release, 30Nm, release, final torque 30-35Nm. Some people have mentioned that later FSM procedures don't mention this requirement, but it won't hurt to do it. Don't overtorque, you can warp the pump body. Better to be under-torqued."

When it says "release" he means back off on the driver to remove the previously applied torque. In other words tighten to 30nm, loosen, tighten to 30nm, loosen then finally tighten to 35 nm spec and that's it.

I think your problem is not the lift pump for sure, some crud got in there during the work you did previously and is responsible for full fuel flow. AND the dirt in the end of the line could have left crud in that injector, the tiny opening in the nozzle is probably plugged, so I would start looking to have your injectors pop tested and spray checked, AFTER you find the problem in the IP.
To kill a runaway diesel, keep a pair of metal snips or similar nearby and if it runs up like that start cutting the lines one by one until the engine dies, thats faster than trying to loosen nuts and cheaper than a replacement engine.
DDH

This was a cake job and not my finest moment. Out of respect to the suggestions from here that I appreciate so much, I want to be clear that I did buy new washers and more o-rings and torqued everything perfectly, as suggested, I even used the torque sequence provided. The engine continued to run poorly. It was the final time I went into the IP- to explore my original mistake of messing with the lower parts of the valve- when I "slightly over torqued" 2-3 lbs. At that point, I had given up on a perfect torque being the extraneous variable, but I may have been wrong.
Thanks for all the ideas.

Internal IP
 

I am now thinking this was the problem all along. When I did the first valve, the first time I opened the IP, I pulled out the lower parts because I didn't see the spring at first and thought everything was lower in the aparatus. I used a magnet to get it all out- 3 pieces. I remembered the diy saying not to mess with the lower pieces, but it was too late by the time I realized what I had done. So when I went back into the IP the final time, it was to check the lower parts on the other valves to make sure I re-inserted the parts correctly to that first valve I had worked on. This time I noticed a hole on one side of the longer, lowest piece and had a bad feeling, i.e. it probably needs to line up correctly inside. Could the answer be as simple as aligning these lower pieces correctly ? or is this IP toast, or is there another option?

Yes.
 

That is the risk that you take when you go into the pump. Many get lucky and their car runs great after replacing the parts in the pump. But you and I weren't so lucky and experienced run away engines. Once the pump gets into the run away mode, it will remain in run away mode until you replace the pump with a known good one. There is not much information to tell you how to fix the pump after it starts running away on each start, which I think is your question that MANY have asked here and at other places. I can tell you that you can use a wooden board to shut the engine down when it starts running away, but it will eventually blow your seals, especially at the front crankcase, which is better than blowing at the rear crankcase. That is why I say save yourself some money, time and heartache and cut your losses now by letting somebody else fix your car, who understands these diesel cars.

BenzDiesel

Wanting to understand more is what got me in trouble.
 

How do you know the pump is dead? The car runs away each time you attempt to start the car and you don't have any answers as to why. Could it be a lift pump or bad fuel. It could be. It could be anything. But the fact that the pump was opened and the car had no problems before then, lets me know that your first problem is related to the work done on the pump. If these cars are going to run away, it will happen upon start up unless you are REALLY unlucky and the fuel over filled the oil crankcase and the car goes into run away mode and you are in drive gear and forget to put the car in neutral and run away from the car yourself. I haven't heard of anybody experiencing a run away in drive, but one guy's crankcase a couple of years ago was over filling with fuel and we/me suggested that he not drive that car until he resolved the problem. Hopefully, he is still amongst us, the living.

BenzDiesel

I read that thread if it was the car with 16 quarts of fluid in the crankcase, luckily the car stalled. I am catching up on reading more about the subject and looking for a replacement IP.

Thanks to Barry123400 also for your comment which helped confrim the diagnosis for me, in addition to reading some of your other posts on similar threads.

Quick follow-up. I had the IP replaced after struggling for a few weeks to find someone with a reference and experience. I had it towed ~70 miles. I bought a used IP after sending a message to diesel911 to get his opinion about the symptoms the donor car had. Now, the car runs well. Working on shifter bushings and the auto trans fluid and filter. After that, the cruise control wiki fix and then I have until next summer to get the A/C running.