So polishing the faces of the DV's works.

I took the 2 DVs from Cyls 4 and 6 that were causing such loud nailing and gave them a polishing session. Cleaned well and swapped them into 1 and 2. I had no nailing before the swap and have no nailing after the swap. I'm currently running straight pump #2 without any additives, so any nailing would be very obvious.

I'll keep driving it and give a status report after a few days worth of driving and some cold starts. Seems promising though and could explain quite a lot of nailing issues that seem so hard to resolve.

It is important to note that the DVs I used were the last thing replaced to eliminate irritating nailing. Injectors were tested, DV seals were redone 3x, ran diesel purge, adjusted timing, added 2-stroke oil to the fuel, tried cetane booster. Nothing worked. As the engine warmed up, it would start to nail and just get louder until the engine reached temperature at which point it was ridiculous. Swapping in different DV's killed the problem in both cylinders. Now those DV's that I pulled got polished and reinstalled in different cylinders that were NOT nailing and continue to NOT nail afterwards. Far from scientific, but certainly seems to suggest that polishing/lapping the sealing faces helps.

__________________
Current stable:
1995 E320 157K (Nancy)
1983 500SL 125K (SLoL)

Gone but not forgotten:
1986 300SDL (RIP)
1991 350SD
1991 560SEL
1990 560SEL
1986 500SEL Euro (Rusted to nothing at 47K!)

Gone and wanting to forget:
1985 524TD 167K (TotalDumpster™) [Definitely NOT a Benz]

Awesome! Thank you for going out of your way to do that; I realize that it wasn't necessary since you've already solved your problem. But that knowledge will definitely be helpful to others in the future.

I'm looking forward to hearing how they hold up after some extended driving. That would be great if the Cerium treatment is all that's needed instead of having to procure new or used DV's.

__________________
Current rolling stock:
2001 E55 183,000+ Newest member of the fleet.
2002 E320 83,000 - The "cream-puff"!
1992 500E 217,000+
1995 E300D 412,000+
1998 E300D 155,000+
2001 E320 227,000+
2001 E320 Wagon, 177,000+

Prior MBZ’s:
1952 220 Cab A
1966 300SE
1971 280SE
1973 350SLC (euro)
1980 450SLC
1980 450SLC (#2)
1978 450SLC 5.0
1984 300D ~243,000 & fondly remembered
1993 500E - sorely missed.
1975 VW Scirocco w/ slightly de-tuned Super-Vee engine - Sold after 30+ years.

Just wanted to pop in and thank everyone for starting and contributing to this thread. I've been chasing noises all over my engine for two years and was at wit's end. Now have another possible avenue that suits my symptoms better than other options I was pursuing. Will be pulling DV's this weekend.

__________________
617 swapped Toyota Pickup, 22-24 MPG, 50k miles on swap

Just a word of advice and caution to anyone willing to be guinea pigs with DV face polishing:

USE THESE INSTRUCTIONS AT YOUR OWN RISK!!! PLEASE BE CAREFUL AND DON'T TRY THIS ON YOUR ONLY MEANS OF TRANSPORTATION!!! EVERYTHING BEING DISCUSSED HERE IS THEORETICAL AND EXPERIMENTAL!

If you do wish to try polishing your DV sealing faces, use common sense and follow the instructions below:

1: Make sure you have your DV oriented correctly. If it has a groove on the barrel, it should be pointing down, away from the plunger part of the DV.

2: Before you do anything to the DV, clean it in solvent, such as lacquer thinner to remove any varnish or contaminants.

3: Dip the entire DV in a thin oil such as sewing machine oil, or even WD-40 as a lubricant.

4: Use Cerium Oxide as your abrasive. It is extremely fine (finer than makeup powder) and is a very "soft" abrasive.

5: Use a battery operated drill and fasten the tip of the plunger part of the DV that the spring attaches to in your drill chuck.

6: Use a toothpick dipped in oil to transfer some of the cerium oxide to the beveled face of the DV. Try not to get it anywhere else. Keep applying until you have a coating of the cerium oxide on the face.

7: Insert the plunger into the barrel and press the two pieces together firmly and run the drill at medium speed for ~15 seconds.

8: Pull the parts apart and clean in your solvent. Inspect the faces and look for a polished appearance. If they need more polishing, repeat 3-7 above until you're satisfied with the final appearance.

9: Use a toothpick dipped in cerium oxide to clean up the entire beveled face of the plunger. I held the drill up and ran the tip of the toothpick back and forth over mine. This will help get any other oxides off and give you a fully polished sealing face.

10: Clean both halves of the DV in CLEAN solvent. Dry and repeat in a new pot of CLEAN solvent. Use a tooth brush or a Q-tip to really scrub everything perfectly clean.

11: Dip the DV in thin oil and reassemble to check fit. It should move freely without dragging. Press the DV halves together and twist. It should not feel "rough".

Keep in mind this process and all of the anecdotes attached to it are theoretical and experimental. If you screw up your IP or break your only car, you have yourself to blame. Not I, Pelican Parts, or anyone else is responsible for you or your engine. As the fine print always claims - "Your results may vary"

__________________
Current stable:
1995 E320 157K (Nancy)
1983 500SL 125K (SLoL)

Gone but not forgotten:
1986 300SDL (RIP)
1991 350SD
1991 560SEL
1990 560SEL
1986 500SEL Euro (Rusted to nothing at 47K!)

Gone and wanting to forget:
1985 524TD 167K (TotalDumpster™) [Definitely NOT a Benz]

Power lapping seems a safe approach. My only comment is if you find one with a leaking valve to seat issue. The others may not be in that great shape either.

Perhaps not bad enough to cause noticeable issues. The other delivery valves should actually clean up easier as they are not in as bad of a condition.

Kind of a general maintenance issue then in my mind. . Past repairing the really problematic one.

Anyways congratulations on the newest if not only delivery valve reconditioning approach. As the member cautions just use your head when doing it. Do your best at keeping the valve and seat in equal contact when lapping.

The risk is minor. If you some how manage to mess up. Just get another new or used delivery valve.

I live in a small area. Can you mention a source of the light abrasive you used?


I was impressed with the way you tackled this. I knew the valve was too small to reliably lap it on glass in combination with dealing with the seat as well separately. These two parts probably had been finished lapped against each other in production as well.


Original life cycle of the check valves is great with no valve rotators present as well. There are no specific water separators on these cars either. I wonder since there is probably always a little water in the fuel we buy.


The only water trap is the secondary fuel filter and I have always wondered why they recommended such frequent fuel filter changes.


Over the years the moisture adding a little surface corrosion to those seats and valve faces might be what is being polished out. When enough is present it creates a leakage situation.


Fifteen seconds with that abrasive is not going to cut hard metal very much if at all. Polishing or cleaning of softer corrosion is more probable. As that was loosened up it would mix with the new lapping compound perhaps.


Many more people will be doing this now. Perhaps someone could examine the valve face with a microscope. Post what the observe.

Can you please wright how to disassembled and assembled after cleaning. I heard horror stories from mechanics , that if you moved them it will never run a same again.

I changed fuel injector nozzles, and result same as you experienced even louder or same. Than I added 2 -stroke oil, and it's get better sometimes it's so good I can hear soft clucking.

https://youtu.be/maiKdps41i4

__________________
91' MB 350 SDL, with 961 engine and #14 valve head
09' ML320 Bluetec

On the M pump that the SDL uses, the delivery valves come right out once you take the holder off with the special spline tool. Unlike the MW pump that the 617's use, the Delivery Valves have no bearing on how the engine runs since you are not disturbing the pumping elements.

__________________
Current stable:
1995 E320 157K (Nancy)
1983 500SL 125K (SLoL)

Gone but not forgotten:
1986 300SDL (RIP)
1991 350SD
1991 560SEL
1990 560SEL
1986 500SEL Euro (Rusted to nothing at 47K!)

Gone and wanting to forget:
1985 524TD 167K (TotalDumpster™) [Definitely NOT a Benz]

Did you ever look at prime pump? May be this nailing sound is coming because the fuel flow is not enough. Just a thought. It came to me after added 2 stroke oil, a "thicker" diesel is creating better flow to pump and fuel nozzles.
Now why is I'm thinking, because before I added 2 cycle oil, my would like fluctuating at stop light, like skipping one cylinder. I know for sure my nozzles are at right pressure, since I change all previously installed spacers, they were all not enough to get right pressure at nozzle.
It's little expensive part , prime pump . But this pump is working all the time, even start up, right ?
I got springs from Pelican and will get kit from Mercedes source.

__________________
91' MB 350 SDL, with 961 engine and #14 valve head
09' ML320 Bluetec

I bought my cerium oxide from Amazon. I originally used it with a power buffer to get rid of tree sap on my Honda's windshield after it got baked on and literally nothing would take it off. It will be sold as a glass polish. As a useful sidenote, it's also GREAT for removing hard-water spots from windows.

The reason I chose to use cerium oxide is because it is such a mild abrasive. Far less than the carbide and carborundum usually used in valve lapping compounds. The goal was to polish the faces and it did just that. One of the DV's I polished had noticeable pitting to the face. Apparently it wasn't very deep because it polished out.

Until we have some anecdotal evidence from a few other brave souls, it would be wise to hold off doing an entire rack of DV's until we can gather some data points.

__________________
Current stable:
1995 E320 157K (Nancy)
1983 500SL 125K (SLoL)

Gone but not forgotten:
1986 300SDL (RIP)
1991 350SD
1991 560SEL
1990 560SEL
1986 500SEL Euro (Rusted to nothing at 47K!)

Gone and wanting to forget:
1985 524TD 167K (TotalDumpster™) [Definitely NOT a Benz]

Hey Diseasel - you may want to put a caveat somewhere in your post.

All kidding aside, thank you for the comprehensive instructions!

(Now I just wished that the DV's were as easily accessible on a 606 vs 603.)

__________________
Current rolling stock:
2001 E55 183,000+ Newest member of the fleet.
2002 E320 83,000 - The "cream-puff"!
1992 500E 217,000+
1995 E300D 412,000+
1998 E300D 155,000+
2001 E320 227,000+
2001 E320 Wagon, 177,000+

Prior MBZ’s:
1952 220 Cab A
1966 300SE
1971 280SE
1973 350SLC (euro)
1980 450SLC
1980 450SLC (#2)
1978 450SLC 5.0
1984 300D ~243,000 & fondly remembered
1993 500E - sorely missed.
1975 VW Scirocco w/ slightly de-tuned Super-Vee engine - Sold after 30+ years.

Of course holding off on doing the whole rack of them. Until some overall experience is gained. Just something to consider now.

Why I considered it safe was for true lapping both surfaces have to be held parallel to each other. Usually requiring some form of valve guide to hold the relationship. You cannot achieve this with a hand held drill alone. Where cleaning up corrosion and other hard deposits is not nearly as demanding. What you established is far too important to ignore.

There will be failures if there is an obvious flow erosion channel across a seat or valve for example. Or the corrosion is massive. Common sense will have to be used. Personally I think your approach is going to work out well for many members.

I am old enough and experienced enough to recognise an approach that will usually probably work. A seat or valve warped from the heat for example is never going to occur. So I just suspect you have developed a real winner of a more constantly occurring problem all the time here.

I was poking around and found this

Delivery-valve problems are cylinder specific. If the valve sticks open, no fuel passes to the associated injector. Leakage is harder to diagnose. White smoke that persists after all bases have been touched—pump pressure and timing, new injectors, and engine compression—suggests that one or more delivery valves may be at fault. Disassemble one valve at a time, clean the parts thoroughly, and test for leaks by blowing through the outlet port. Older, simpler valves can sometimes be resurfaced. Check piston fit by depressing the valve and placing your finger over the inlet port. You should feel the vacuum as the piston falls.

__________________
92 e300d2.5t
01 e320
05 cdi
85 chev c10

This is another good point for mine, I've had smoke the entire time I've owned the engine. I would have thought swapping a different injection pump with different delivery valves in would have helped or at least changed things significantly, but with a used pump who's to say what condition all of that was in.

__________________
617 swapped Toyota Pickup, 22-24 MPG, 50k miles on swap

It's important to note that there are a lot of other things besides DV's that will cause smoking. Low compression, injection timing, spray pattern, and of course the SOURCE of smoke are all important factors to consider. If you swapped the IP and had the same behavior, it's probably safe to say that the IP wasn't the source of problems. Used or not.

I'd be looking for oil consumption problems, worn turbo seals, or IP timing before blaming the DV's. It's very common for people to claim they have "grey" smoke when it is in fact "blue" from oil burning.

__________________
Current stable:
1995 E320 157K (Nancy)
1983 500SL 125K (SLoL)

Gone but not forgotten:
1986 300SDL (RIP)
1991 350SD
1991 560SEL
1990 560SEL
1986 500SEL Euro (Rusted to nothing at 47K!)

Gone and wanting to forget:
1985 524TD 167K (TotalDumpster™) [Definitely NOT a Benz]

Actually sticking open is an interesting possibility. I was trying to grasp the possibility. A totally dead cylinder seems to be the suspected result.


Over the years non soluble particles in the fuel might get compressed and stick to the seat or valve face. Providing enough of a compromised surface to leak. It would not take much leakage to disable the function in a hydrostatic situation like this. Volume of fuel flow is probably also low so a washing type action may not be strong enough to erode the compressed and sticking contaminates.

I was more concerned with a finding a testing method to find out if leaky delivery valves are present in place. Especially since this threads originator proved we could address the delivery valves at home.


Perhaps taking an old hard line cutting it and brazing or silver soldering a usable attachment on the cut end.

Apply perhaps 25 pounds of air pressure with a gauge in the rubber line that feeds it. . With a shutoff at the end and an air input at the extreme end. A good delivery valve should hold the pressure for days I expect. For testing a more sane period could be established. I think we could quickly develop what a problem delivery valve acts like. Or establish the leakage factor that is acceptable.

You might have to run the injection pump out of fuel and rotate the engine to expose the port in the injection pump element. Giving a route for the air to easily escape if the valve was leaky. May require a few magnets and a degree wheel stuck on the harmonic balancer. As an indicator when a specific fill port was exposed on each element.

All four of my diesel 616 and 617 engines run very well. Although the 616 engine that has rebuilt or all new parts including the basic engine runs just a little better. That engine still has only a very few thousand accumulated miles since it was installed by a dealer.

It is possible in my mind because these cars are so old that there are a lot of sub standard delivery valves out there in service. Not causing any signifigant problem but degrading the engine just enough to notice a difference like I observe. There is no fairly quick test to see what shape they are in that I can think of quickly.

Other than the sitting under a head of air for a period yet to be established. Or taking an old hard line and teeing in a high pressure gauge with an injector on the other end. Building pressure by rotating the engine and seeing if it holds.

Once the delivery valves are removed I think my vacuum pump and one of my better vacuum gauges that reads down to low microns could be used.

If a hose could be clamped on the delivery valve casing. It could quickly pull down to 100 microns and sit there when the vacuum pump is isolated from the circuit with a valve. The test could be done very fast and be very conclusive.

I also think with any leakage it would be obvious with my stronger pump in only a very few seconds. You could evaluate each delivery valve this way establishing if it needs attention. Plus testing after you repair it to verify it is now good to reuse. This to me is a very fast way to deal with them.

What is a little scary is the majority of injection pumps on these cars are still the originals. Very old now with high miles on many of them. I suspect even a little leakage will have consequences. Nothing you might be able to put your finger on but still something. Could be for example fuel milage. A slightly leaky delivery valve might delay the injection for example. Or under fuel a cylinder. Or both.

The much maligned milli volt system should indicate it as one of many possibilities if a cylinder is running cooler. At the same time if an engine is producing no more than .1 milli volts difference between all the cylinders in a power balance test the overall delivery valves are probably very good.

We did experience on this site and in private messages and emails. That some engines settled to a perfect milli volt test after what was wrong was repaired. All the cylinders were exactly the same after in the milli volt power balance testing. Not that many but there were a few of them.

Years ago now when I joined this site there was little known about these pumps by most members in comparison to what now is. Pumps got changed out very frequently when they were either not the issue or we could have addressed the issues.

There are much fewer injection pumps changed out today in comparison. Certain things are and will still remain beyond what we can do. These cars like myself have continued to age over those same years. Delivery valve issues seem to be increasing.


These are just a few quick thoughts. Others can come up with better and easier hopefully. Or perhaps I can with some thought.