Other foum members have documented that adjusting the Engine Valves made the Engine easier to start. The assumption is that the need for Valve Ajustment had an effect on Engine compression.

So if the Valves needing to be adjusted has an effect on starting it could also have an effect on the idle.

And, then there is simply that since doing the Valave adjusment cost little it is a good thing to eliminate for before you replace Fuel Injection Pump Delivery Valves.

Also some have found that Engine Compression caused by something like sticking Piston Rings was the issue with their rough idle.

That is why I say the Valves may need adjusting.

__________________
84 300D, 82 Volvo 244Gl Diesel

At a minimum, "rebuild" means replace the O-ring (if you injection pump has them, some do not) and the copper washer. The FSM recommends replacing the spring as well, so I do that too. I've found the old springs to be shorter (weakened) compared to new springs, on the three or four injection pumps I've been into.

Some have reported an increase in combustion noise with new springs, and one fellow I know even reverted back to the old springs because he didn't like the new sound. Many have not replaced the spring and had no problems.

The key point is that the copper washers wear and leak by, and idle quality is affected.

__________________
Respectfully,
/s/
M. Dillon
'87 124.193 (300TD) "White Whale", ~392k miles, 3.5l IP fitted
'95 124.131 (E300) "Sapphire", 380k miles
'73 Balboa 20 "Sanctification"
Charleston SC

Most times when a problem is intermittent it makes it harder to find. Only the poster knows if it is thermally or temperature related.

If for example when you start the car in the morning the idle is normal but as the engine heats up the idle gets really bad. Or it seems to be so especially on a hotter day when the engine and injector pump are running hotter.

You did mention on a hot day with your a/c not working and sitting in traffic it rattles your teeth. So I thought the common variable might be the reduction in the fuel viscosity. This from the normally generated heat being higher in the injection pump if that where the case. The reduced viscosity could increase the leakage rate past a delivery valve or elements piston if worn.

Have you ever considered that the problem is worse on hotter days? Can you start up your car and let it idle on a hot day until the really rough idle develops? At idle these engine will only burn a gallon of fuel probably over eight hours. When the idle is really rough play a small stream of water from a garden hose on the pump. In a few minutes it may reduce the internal fuel viscosity inside the pump enough to restore normal idle. Or even a serious reduction is a good indication. You did mention the problem is worse on diesel fuel so perhaps it may be best used for a test. I have not read the factory service manual but if they recommend replacement of the springs when servicing the pumps. To me there is the additional thought that they never envisioned these pumps to be still in service 30-40 years later. What they were aware of is the springs do sag in a far less period of time though apparently.

This is actually an old test. In my thoughts I just could not find a way to separate beyond that point. To establish if it was a worn element or a leaky delivery valve creating the issue. normally if you could read the plug voltages you might just deal with the one delivery valve. Then if that did not cure it you would know it was that cylinders element instead.

I did find a way to determine what element or leaky delivery valve it may be though. On the later 123s that have a plug in glow harness. You start the car and unplug the harness at the relay. You apply the negative probe of a meter with a milli volt scale to only the engine as ground.

Every few minutes you note on a piece of paper the voltage the glow plugs are producing from every terminal in the disconnected plug. Each glow plug because of the disimular metals the element is made from produces a readable voltage in milli volts.

Depending on how hot it is. At some point when the engine starts to roughen up you should see one plugs voltage reading starting to drop. Simply because the running temperaturte of that cylinder is getting cooler. This will tell you what element or delivery valve has the problem. This unfortunately only applies as a test to those without a constant problem.

Also unfortunately this poster does not have the plug in glow plug harness on his car as it is an early 123 engine. So it would be much harder to do this test on. At the same time I thought for general knowledge this application of the milli volt testing I thought of should be mentioned for general knowledge. Although with some difficulty it could still be applied to his situation.

Also if the poster is bound bent and determined to do the delivery valves. I would remove the current ones and inspect them carefully as we have apparently had people hand lap them.

Also as another poster mentioned it may just be the copper washers worn anyways. Just going out and buying new delivery valves probably at a fair cost first may not solve the problem anyways. I have tried to estimate if it is a delivery valve issue or a worn element situation if the pump responds well to thermal testing. I just cannot calculate the odds.

Sorry about the long post but anytime I can think of something that may not have been considered before I feel I should post it. Basically it amounts to if you have a problem like this gentleman you might be able to locate it to only one cylinder to deal with.

Quite easily in fact perhaps on the plug in glow harness cars. It does not matter what voltages you would be recording from the glow plugs. By recording them you would be looking for abnormal falling voltage changes in any one reading compared to the other readings as the problem developed.

I did say years ago that over time I felt we could discover many ways to deal with issues with these injection pumps.. In reading other posters posts on this thread I have discovered yet another one perhaps to think about. Relief valve springs that are partially collapsed with lost tension with age may affect overall injection pump timing. . I noticed this from the post that stated someone had put stronger delivery valve springs in and the engine as a result was creating more noise. They reinstalled the old ones and it was quieter.

There is a possibility that the stronger springs just sealed the delivery valves better but I think not. Just quicker perhaps. What the stronger springs may have done is to keep the residua pressure present in the hard line higher and in this way actually advanced the injector timing. So someone should put a new set of normal springs in any high milage example first to see if it improves things.

I can see the day where it might be considered as a general tune up item at some point. Especially if those springs are pretty cheap. Also we may have to start paying more attention to the delivery valves as these cars age. Engines in general not running as smooth perhaps as they should or lower on power than possible. This may especially be more relevant on the four cylinder 616 engines. Simply because you are using the engines total output power far more often. Any spring under tension and these ones always are even when the engine is not running will sag and lose some tension over thirty years has to be a given. Also they might suffer a little flow erosion as well.

If what I am thinking is plausable. We have had many cases where the sequential timing of injection pumps was not what it should be by more advanced and involved milli volt testing. It is my guess now that in most cases. A proper refurbishing of the delivery valve system and springs might have corrected the issues. You do have to do a compression check and have fairly even compression to follow this course though. Or any improvement may be marginal.

Basically the pressure retention in the individual hard lines once the injector has closed may have real signifigance. We perhaps should look into this. The sequential timing adjustments are locked down. So unless there is a change in the element piston to bore clearance the timing will stay the same in relation to other cylinders unless there are issues further on. .I never read the factory service manual but it is certain that if they were seeing sagging springs way back when these cars where much newer. They never envisioned these pumps still running thirty to forty years later with the original springs still in them doing their jobs properly.

You are probably referring to my video and the thread I started. Like you, I went through the entire gamut of diagnosis and nothing had any positive effect. The previous owner of the car in the video had run straight veggie oil in it for an unknown amount of time; which likely caused the DVs to wear out or stick. I found the replacement valves from a junk yard car that looked like it had not ever seen an alternative fuel. The DVs are pretty easy to replace. You can probably get them new from a Bosch shop but I happened to get lucky going the used route.

__________________
1991 F250 super-cab 7.3 IDI. (rebuilt by me) Banks Sidewinder turbo, hydroboost brakes, new IP and injectors.
2003 S430 - 107K
1983 300SD - Tanoshii - mostly restored ~400K+.
1983 300SD - Good interior. Engine finally tamed ~250K.
Monark Nozzle Install Video - http://tinyurl.com/ptd2tge

Thanks for all the responses folks.
barry - Are there any obvious signs of leaking DV seals/rings besides overall poor IP performance? I dont have any visible leaks, infact I even tried sticking paper towels around all the lines to see if anything is maybe leaking at a very slow rate, nothing yet. I ask because I m not sure how to determine a leaking DV seal if there arent any visible cues.

I will however try measuring the temperature of each cylinder today to see if any one particular DV is the culprit.

I spoke to a fellow knowledgeable forum member who suggested running some wvo to increase overall fuel viscosity and see if that makes a difference once the temperature goes up. He also gave me his contact for IP/DV rebuilds close to my area. If all else fails, I might take it there.

eatont9999 - Yes it was your video. I watched a bunch of your videos, great stuff! I will give the junkyard a shot, although I doubt I ll have much luck since most of the W123s I ve found have either been severely gutted or look like they were treated poorly.

FWIW, would DVs from a turbo engine work? I did notice some 300SD at the local yard that looked like they were in much better shape. Either way, all this may result in no difference at this point since most of this conversation is speculative. I have a valve adjustment scheduled soon. Hopefully that will make a difference and/or eliminate potential sources of diagnosis error.

I will post a new video soon showing how things are.

You could check with any injection repair shop. Personally I suspect there is a very strong chance that the delivery valves are the same. Actually if they turn up to be your issue or not. I knew how they functioned before. At the same time if not functioning well might give various effects I had never considered before. Besides the obvious.

As for any temperature comparisons remember they must be with and without the problem present to have any validity. This is the only way these engines can be dynamically tested with cheap equipment. Your problem being not there all the time at least to me indicates it may be temperature dependent.

Here are a couple of videos from earlier today.

First one shows the lopey idle, I also tried the temperature comparison between glowplugs while idling - https://www.youtube.com/watch?v=XDKWdgoB9Y0

Here is a 0-60. Based on the video time code looks like 60 happens at 24-25 seconds.
https://youtu.be/kYosE2nqM-Q

If you observed the 40-50 degrees less temperature as you know it is a good indicator that cylinder is running much cooler. The test makes sense as that cylinder would be also making less voltage if you had been able to employ the milli volt method as well. Personally I expect it might have dropped to zero voltage after the engine got really warm.

Well at least you know where the problem is now or a good part of it. One reason I did not suggest using a thermal reader. Was the cool plug casing will reflect the coolant or block temperature in the area I suspected perhaps as well.

The milli volt method indicates the glow plug tip temperature. That cylinder may even be totally dead and by loosening off the injector nut a little you may notice no difference in idle quality. When the idle is really rough of course.

Anyways you are on your way now. You made me start to think about the many effects delivery valve problems could result in. I am toying with the issue of engines running still reasonably well but power being down.

I have had a very hard time selling the running temperature of the individual cylinders as our most definitive test. Plus it being about the only dynamic test available.

I assume you will only take this one delivery valve out for inspection. If it looks good perhaps swapping the position with another one as a positive test. Or repairing or changing only it. Of course the change between a hot engine and a cold one primarily in your case is the viscosity of the fuel in my opinion. If you feel a compulsion to do them all right away avoid it. If you can get this delivery valve working and it is the problem. Do the rest later if you must. Doing them all at once may introduce something else into the equation by accident. Also it may still be the pumps element for that cylinder remember.

I am sorry for my approach to trying to help you out. In the right hands using tools like the milli volt system got me a lot of flack years ago. The shortcuts some people where taking negated any value in too many applications. Without taking shortcuts it was and remains deadly accurate.

Eventually I decided that to further pursue it might result in members getting themselves in trouble. Or discrediting it by improper application as I was observing. So I only suggest it in a few cases where they will not. It hurts me to see problems posted where using milli volts it is about the only sane way to determine what is really wrong with great accuracy and fairly quickly. Yet I cannot mention it. It has to still remain in the future for now.

It takes a pretty much dead cylinder to get usable results with a thermal reader I suspect but you did it. Others should remember that it is a very crude approach compared to using milli volts at the same time. What gets anyone through the problem is fine by me though. Sometimes a nightmare otherwise. .

Your 0-60 time is reasonably consistent with the issue you have. Plus you have indicated you have enough knowledge to tune up your fuel system later if needed.

Your time is about what a four cylinder 240d automatic would do. My 77 N/A 300d is quicker than that or I would have noticed it.

Also in some of these situations it makes me think again. I never found the reason for the fuel milage discrepancy between 240ds with manual transmissions for example.

Most get around twenty five miles per American gallon on the highway. Yet some get thirty or a little better. This differential was too consistently reported by too many members to ignore.

Part of it may be weak delivery valve springs. To me there is a possibllity you drip time the injection pump but at higher rpms the delivery valves have to wait until the pressure wave from the injectors closing off to seat them. Good springs may shorten this time as the valves slam shut quicker. The quicker or earlier they close the more residual pressure is retained in the hard line and there is a good chance the elements fill better.

Much of my thinking along this line goes to the fellow that changed back to his original weaker springs because his engine was noisier with the new springs. I have to wonder if he drove it when noisy and found he had more power? Certainly if the fuel milage was impacted in any way was the last thing on his mind. Weak springs might have a retarding effect on the injection timing.

I have been carefully looking for an indication of what causes this discrepancy for years now with no positive indicator yet.

Well anyways the lawn tractor at the cottage blew the head gasket this weekend. So I better go get a replacement gasket and gather up the tools needed to change it out there..

Clean your tank strainer.

To address your initial question, I doubt it is due to using the new Diesel HPR. I use it since my engines seem to run quieter and slightly smoother with it, plus cheaper than D2 and good for the earth. But I have used D2 in between and neither causes a rough idle.

I would also first suspect the fuel supply to the IP, either restriction from the tank (clogged screen) or a failing lift pump. The later seems more likely since a restriction should cause more problems at higher flows. Some people have installed a pressure gage on the lift pump output, some tapping a banjo fitting, but you could also install a tee if you changed from factory nylon to a rubber hose. I recall the spec is 12 psig at the IP inlet.

__________________
1984 & 1985 CA 300D's
1964 & 65 Mopar's - Valiant, Dart, Newport
1996 & 2002 Chrysler minivans

Thanks Bill, I have the repair kit for the lift pump but havent installed it yet.

FWIW, I tried using D2 last week and it seemed like occasionally my idle wasnt as shakey; however once the engine was warm enough or the ambient temperature was high enough, the shaking returned.

You are correct, HPR makes my car run quieter and smoother as well but I think (from recent experience) there is definitely merit to the idea that D2 has slightly higher viscosity than HPR; therefore it takes longer to "thin" out at higher temps.

I have sourced some delivery valves - they are not cheap. I think fixing the Lift pump first might be the more prudent thing to do since I already have the kit.

Your thermal read outs of the glow plug temperatures in the video indicated it is not the lift pump although the kit cost is cheap. Doing just the one delivery valve at least first is important in several ways.

You appear to have the issue generally verified and can get a conclusive indication by watching the milli volts decline from that glow plug at the pluggable harness if desired but not really needed. Your method employed to find that cylinder was more than a reasonable indication for further investigation. The problem is serious or well advanced for you to have located it the way you did. You do have the older series glow plug system. This would mean you would disconnect one wire from the suspect cylinders glow plug. And read the milli volts with the engine running from the open glow plug terminal to the solid wire exiting the same glow plug. If the voltage remains high initially compared to any other glow plug. Yet declines to nearly zero as things get hot. You have absolute proof in your case.

You might even have a weak injection pump element. You can prove this either by swapping out the number three cylinders delivery valve that is running cool after warm up. With any other delivery valve in the injection pump. If it is the delivery valve the problem will move. Or with inspection might see the build up on the surfaces.

There is no way in my opinion that a lift pump can kill off only one cylinder in the engine. Or anyway I can think of after it warms up.

I almost never spend money on parts until I prove the need is really there. At the same time your choice and doing it your way is a personal decision. At the same time it is reasonable to refresh a lift pump. Best done after the need is proven though. The last owner could have had it done for all we know with these cars.

It has reached the point that new delivery valve springs should be installed on a cleaning and overall checking of the delivery valves. . I am starting to really wonder about the true impact old tired springs may present. It is still early yet to draw any firm conclusions. Hopefully the springs are inexpensive.

This has made the need for an easily applied test to determine the delivery valves general condition in place. We have to find if they leak down and if so what is acceptable.

Many must be quite dirty and crudded up by now after thirty or more years in service. There to me is little doubt that old sagged delivery valve springs will have some overall effect. It is important enough that Mercedes service literature mentions to replace those springs during any rebuilding of the injection pump. I base my suspicions on their being a very short period when the pressure in the element and the hard line are equal.

The spring closes the valve during this short period. Weak springs will naturally close it slower. Then the massive residual pressure remaining in the hard lines slams it shut harder. Rather than just flowing by it back into the element longer. Not exactly rocket science to make a list of all the things this slow closure problem could enable.

To me in some ways similar to leaky delivery valves themselves. Apparently they are expensive enough we should develop a real test that indicates the need to do so is really there. We may need it as a part of a general fuel system tuneup we can do ourselves.

It appears some members have also hand lapped them at home to restore the delivery valves sealing surfaces. A piece of glass with a small hole drilled in it and a vacuum pump attached can indicate the good sealing surfaces from the bad ones on the delivery valve components themselves.

We have a few members picking up some money by rebuilding injectors etc. I think they could also offer a service of checking and repairing delivery valves where indicated as a mail order item. That is particularily good if the new delivery valves are really expensive. Or depending on the time factor offering exchange sets of checked and repaired and verified good used ones.

I am just too old to get involved in new ventures. Surface lapping machines should be a dime a dozen if you can locate a good used one. I have watched them go as scrap. They could save the elbow grease component. Even the present injector rebuilders could lap the mating surface of their rebuilt infectors on them as well.

I may as well mention just why these surfaces can crud up. Dirt etc in the fuel flows through these delivery valves. Just before the valve is initially closed by the spring the fuel pressure has reached a state of near equilibrium. Then the residual pressure of the fuel in the hard line really slams the surfaces harder together. With weak springs the displacement of what is between the two surfaces by the slower displacement of the fuel between both surfaces and what is in it may be an issue over time.

So in effect they are not really good self cleaning surfaces. In all too many cases what they may capture may slowly erode away enough to remain basically functional. On alternative fuels the capture may be far more signifigant depending on what is being dealt with.. Once they become leaky the self cleaning option is no longer enough . Volumetrically the flow rate is little even if they are good or marginally bad. Bad enough and the cylinder will no longer fire properly or at all. When we run a cleaner through the injection pump how it acts on the build up if any on the delivery valves sealing surfaces may be the most important function of it.

Yet once again a member posted an effect his friend observed when installing new delivery valve springs. His engine made more noise at idle. So he put the old semi collapsed ones back in to eliminate the noise. To me there are two possible reasons I can quickly think about that caused this. There may be others.

My prime thought but not conclusive is he restored the relationship between the timing intent by the drip method. So the new noise was just a restoration of the original injection timing being restored with the new springs. That a reduction of the delay in the springs closing the delivery valves is going to do this is pretty obvious. I suspect if he had test driven the car with the noise it would have been faster. That reported experience in itself plus mercedes requesting the replacement of these springs during rebuilding the pumps was enough to justify having at least a further look at these springs and their possible overall impact. Perhaps something for others smarter than myself to think about.

.

If the Dv's are crud-ed or plagued with some residues, service can be done,but
if those are worn I'm strong believer that service can not be done successfully due to shape and design of the DV's surface.

Attempt to do so by other members I spoke (on STD) failed miserably.
Delivery valve holders on the other hand can be easily done or even fabricated
cheaply.

cheers

ChO

.

__________________
w126 500SEC gen II euro, powered by OM617 turbo stolen from 84 300SD 2.88 diff,EGR blinded
next wish/project: w114 coupe OM603 powered