If compression is good. Then it is time to test to locate the exact issue. It is an education in itself. Each cylinder must burn at a close temperature to the others to have a power balanced engine.


I already looked to make sure you had the pencil glow plugs. Proper mili volt testing is reading the operational temperatures of each cylinder. Each glow plug develops a small voltage that reflects the temperature present in that cylinder.

Certain rules must be obeyed when doing it. No short cuts allowed. You are doing a comparison check basically. So even if the glow plug is being used as an uncalibrated thermocouple. It does not matter.

It not only finds the issue. You use it also to correct the issue in most cases. It is an unbelievable sensitive system when used with no short cuts. I stopped advocating it when too many where taking short cuts. Reduced to using this in my opinion it is better than finding a weak compression situation. It is harsh but you will always prevail if you follow all the rules.

Far too many people hunt long and hard for the cause of an issue where this system can really narrow it down pretty fast. Rule number one is in most cases like yours. You must have a record of the compression check across the engine. Or what is commonly known as a baseline.

Equipment needs are a simple cheap digital voltmeter with a 200mv function or smaller. You may also have to buy two new glow plugs from the same batch if yours are no longer the same individually with age. First a swap test of two of your older plugs is done. If the voltage they produce in the new cylinders compares well with them before the move. Meaning if they are close enough for voltage output from the heat input. Then usable. It sounds more complicated than it is really.

Keep us abreast. That's an area for which I've read too many posts warning against disturbing. In this instance, I will allow more experienced hands to prevail.

__________________
1981 240D 4sp manual. Ivory White.

I've read a lot about your millivolt method but very little about how to actually perform it.

Installed are 4 matching bosch glow plugs, replaced as a set about a year ago.
I have a volt meter.

Is the engine running while I perform this test?
Do I use the respective body of each glowplug as a ground, or is the battery terminal or valve cover adequate?
What % margin between glow plugs is considered acceptable?

__________________
1981 240D 4sp manual. Ivory White.

Subscribed .

Interesting post .

__________________
-Nate
1982 240D 408,XXX miles
Ignorance is the mother of suspicion and fear is the father

I did then what I knew how to do ~ now that I know better I do better

Speaking from experience (and confirmed compression imbalance issues) the "millivolt method" won't tell you anything useful. Do a compression test before tearing into anything injection related. DV's would be causing nailing and knocking (again....lots of experience talking here...) which your engine is not doing.

Start with the basics. It's amazing the lengths people will go to in order to avoid the simplest and most important steps that have the potential to save huge amounts of time, frustration, and money.

__________________
Current stable:
1995 E320 149K (Nancy)
1983 500SL 120K (SLoL)

Black Sheep:
1985 524TD 167K (TotalDumpster™)

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

Couldn’t agree more.

Compression test is penciled in for Saturday. I set my valves about 6k ago so I’ll give them a look first.

Worn delivery valves causing shake, however, is documented several times in the archives. Members Cho and Eaton(9999?) cured their shakes entirely with new DVs alone. Granted, n2 does not a hypothesis make. Still, an interesting idea worth pursuing once compression numbers have been ascertained (and hopefully laid to rest).

__________________
1981 240D 4sp manual. Ivory White.

Well the first hurdle is the compression test. This you could not get many people to do in earlier cases if you tried. The poster has a new set of glow plugs and that is a real break. As for application of the system. You read each glow plug in isolation. This is very easy and fast with the five cylinders that have a plug in glow wire harness.


My 1983 240d has a parallel wiring set up to the glow plugs and so will the posters 1981 model I think. So he would have to get the engine warmed up and remove the wire connections from the glow plugs. Then start it up again in all probability without the working glow plugs to do the readings. Just disconnect enough wires that no plugs are connected each other.


The firm rule is to use the ground wire of the meter on the engine. If you try to use the battery ground the excess wire in the ground loop picks up undesirable things. Of course the other meter lead to the glow plug terminal.


Usually one glow plugs voltage will not be like the others. You already know that cylinder has good compression. Because you did the compression check. So already you know it involves the injection pump or the injector. So you swap another injector in the bad reading cylinder and verify you still have an off voltage. Many people had old glow plugs and as time goes on they do not produce a simular voltage response. So you think you have it narrowed down. Now swap two plugs to make sure it is just not an odd glow plug. If the voltages repeat again in the same order it is an almost absolute you know the element on the injection pump that is causing the problem. You are now a long way out of the woods. You are no longer speculating.


The worse situation is if the voltages are all over the place. Usually this has not been the case. In this posters case he does not have a knocking sound so it is probably not an element advanced in the sequential order. All I know that if all the common sense rules are applied it works very well. Of course you swap with another one of the injectors. To clear a bad injector as well. In the off reading cylinder. Even though you had it rebuilt and tested.


As time moves forward the proper approach of this system will become appreciated. As soon as you cut a corner you are lost.


For example if you suspect an issue with a delivery valve. You know the one that might need your attention. Far better than shot gunning them all. If it is repaired the voltage will change to normal on that cylinder.. This is just for general information at this point.


I sadly feel the chances of a low compression cylinder may be present. I would not feel this way if there was evidence of a knock. Yet only the compression test will really tell.


Since the time is minimal. Removing the glow plugs on a warm engine is thought better than on a cold one. When you go to do the compression check. Warm the engine. Remove the wires to the glow plugs.


If the warm engine will restart without the glow plugs connected. Start it and read the four voltages. Record the voltages. Then remove he plugs and do the compression check. It gets you familiar with the system even though the readings are not done properly.


They may show something abnormal. If you have a substantial compression issue present You will not be proceeding with the milli volt testing after.


I know your glow plugs are current. What I do not know is how much of a voltage drop would be present if a cylinder was low in compression. Voltages to be expected are in the very low milli volt range of 7-12mv usually. Any questions just ask. I will try to answer them.


Remember this is a dynamic testing system. It was patented by the same fuel system manufacture many years ago.


For some time there was so much flack. Guys where dealing with this system through my private message function. Yet it is just a comparison system. So simple in concept. It just indicates the comparison of combustion temperatures across an engine.


People that wanted to disagree pointed out things that really had no signifigance. For example since a glow plug is not a calibrated thermocouple how do you know the actual temperature? That information is not needed. They were right in that I could not establish it though.


When an engine has balanced cylinder combustion temperatures. It has a balanced power factor overall. When a cylinder is running hotter or colder than the others that power balance is upset.

The hope is with your increased fuel pressure your 0-60 time is so good you do not have a cylinder down in the mud.

Thanx Barry ;

Cylinder balance testing was an old school thing I learned in my GM/DELCO training close to 50 years ago, it was very helpful but few ever took the time to understand or do it .

I oughta give it a whirl just to see what I get as I have the parallel glow plugs on all three W123's .

__________________
-Nate
1982 240D 408,XXX miles
Ignorance is the mother of suspicion and fear is the father

I did then what I knew how to do ~ now that I know better I do better

Just putting this out there since the nonsense of the "millivolt testing" keeps coming up again and again as some sort of useful test:

Thermocouples (and that's basically what you're using the glow plug as) operate on current, not voltage. The voltage they produce is a function of a fixed resistance (load) that they act against.

For any sort of meaningful data to come from a glow plug used as a thermocouple, you'd need to know the exact measurement it should produce at a given temperature into a known, calibrated load. Any other measurement you do is absolutely useless. It may tell you "something", but what it is is arbitrary and trivial. Take the results with a grain of salt. Even if you do find an imbalance in the results, what does that tell you? You have an imbalance in the cylinder. It doesn't tell you why or really give any clues on how to fix it.

__________________
Current stable:
1995 E320 149K (Nancy)
1983 500SL 120K (SLoL)

Black Sheep:
1985 524TD 167K (TotalDumpster™)

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

Okay. Some data.

Cylinder 1-4:

360
320
350
380

Cylinder one (360) and four (380) were tested through the glowplug holes.
Then the elbow broke (harbor freight) and I was forced to test cylinder two (320) and three (350) through the injector holes. Fortunately for me, I had exactly 2 heatshields lying around. I mention this just to be thorough and to potentially account for the discrepancy (lowest two cylinders tested via different port).

I don’t know what this means. But my math does suggest more than a 10% margin between cylinders. Perhaps that’s enough to create a wobble. Thing is, it definitely gets worse when the engine is hot (sometimes). And once, a few months ago, when it was night and near 32 and I had a tank full of b20 and cold started to an engine so smooth I could balance a wine glass on the valve cover.

Thoughts?

__________________
1981 240D 4sp manual. Ivory White.

I've restored to filling my SDL with a gallon of cheap motor oil to stop the shake, smoke and noise.

__________________
1998 Ford Escort ZX2 5 speed - 279,000 miles My Daily

1992 Mercedes 300D 2.5 202,000 - Pure junk
2000 Mercedes E320 Black - 136,000 miles - Needs repair

Don't forget to grease the screw and threads on the spring compressor.

Well unfortunately you have to do the compression test again. I suspect part of the varience may have been you using two different test points. Two cylinders through the glow plugs and two through the injector ports. There is a chance the cylinder volume read was different between the two . Simply because of variations in the adaptors. Or sometimes their is leakage with the harbor freight ones. Anyways it has to be eliminated as a possibility. All cylinders with one adaptor.

Fortunately there is a chance the compression might be more even. For example the mass of the glow plug displaces more cylinder volume than the glow plug hole test fitting. Glad you mentioned this as it could have resulted in a set of milli volt readings that did not make sense. If gets that far.

Testing compression is similar to milli volt testing. The need for absolute accuracy of numbers is far less important than them being similar in a case like yours. Milli volts is cylinder performance comparison. Compression testing is cylinder condition comparison. I do not know what a consistant test will show. Adapter design may or may not be signifigant.

As for the b20 test at 32 degrees. The b20 I suspect has a slower flame front. Plus at that temperature a higher viscosity perhaps. So no real conclusions are possible. Other than the wine was chilled. It does add the potential of another issue but not worth consideration yet.

You burnt the car? I read you post too fast and replaced restored to with resorted to.

Adding a gallon to the fuel I can see it slowing down the combustion process though. Technically the engine should return to what it was after that fuel was consumed though? Still adding the oil and what viscosity. ?

That's a fair bit of sideways movement shown in your video, its very similar to hunting in petrol carb cars where the mixture is reduced down too low.

Inside both the MW and M pumps exists a mechanism for controlling tickover, in essence it parks the centrifugal weights using a spring loaded buffer - ordering a second repair kit which I am hoping contains at least one of those parts.

__________________
David


1996 Mercedes S124 E300TD - 129k - rolling restoration project -

1998 Mercedes W210 300TD - 118k (assimilated into above vehicle)

What do you make of the compression numbers?

I'd like to do it again, but for the sake of consistency, entirely via glow plugs or entirely via injectors. From what I've read, my lowest cylinder isn't necessarily "bad" but the difference between 320 and 380, is. Not entirely sure what to do next.

__________________
1981 240D 4sp manual. Ivory White.