1998 E300 will not pass CA smog because of bad glow plug. What to do? - Page 3 - PeachParts Mercedes-Benz Forum

ROLLGUY · #31 01-30-2017, 01:59 PM

I got the car today and checked all the glow plugs, and they all seem to be working. I would have thought that th glow plug was the problem, but it may be a bad relay instead. I took off the manifold so I could isolate each wire for testing, and also each plug. I tested each wire between the connector and each plug, and there are no shorts or open leads. I then tested each plug with my meter on the 200 ohm scale, and I got between 1.4 & 1.9 ohms on all of them. The glow plug light is still lit is still lit, and the "check engine electronics" message still appears. I am not sure what to do now, other than grab the GP relay off of my friends parts car for testing. The car I am working on is three months past due on the tags, so I would like to help him get the car smogged before he get pulled over for expired tags.

funola · #32 01-30-2017, 02:14 PM

You cannot measure accurately the resistance of a glow plug with an ohmmeter. Easiest way to check while installed in an engine is measure the power they draw with an ammeter. Put the ammeter on the 20 A DC scale, connect one lead to each glow plug, other lead to battery + terminal. If the glow plug is good, it should surge to 18 - 20 A, then steadily drop to 8 -10 A as the glow plug heats up. The amps drop because the Nichrome wire resistance increases when heated.

Edit: If the 1.4 and 1.9 ohm reading is accurate, it indicates a bad glow plug. A good one should be around 0.7 ohm.

compu_85 · #33 01-30-2017, 02:33 PM

Another way to test glow plugs is to get a high wattage lamp, like a headlamp, and ground it through the glow plug.

-J

ah-kay · #34 01-30-2017, 03:03 PM

Hi Rich, questions and some tests.

1) what is the exact CEL number?
2) If you reset it, would it come back on again immediately or after a while?
3) I would imagine GP would cause it to come on immediately and if it is a while then it is probably not GP.
4) You can't smog by resetting it. I am not sure do the smog people check for 'running' data. They do it on gas car. Resetting the CEL would clear all 'running' data and smog will fail.
5) If you start the engine with 1 or 2 GP disconnected, can you check whether you get different CEL code? This is to check whether the computer is telling the real fault. Sometimes the computer throws out code unrelated to the symptom.

Just a thought.

ROLLGUY · #35 01-30-2017, 04:34 PM

Quote:

Originally Posted by funola

You cannot measure accurately the resistance of a glow plug with an ohmmeter. Easiest way to check while installed in an engine is measure the power they draw with an ammeter. Put the ammeter on the 20 A DC scale, connect one lead to each glow plug, other lead to battery + terminal. If the glow plug is good, it should surge to 18 - 20 A, then steadily drop to 8 -10 A as the glow plug heats up. The amps drop because the Nichrome wire resistance increases when heated.

Edit: If the 1.4 and 1.9 ohm reading is accurate, it indicates a bad glow plug. A good one should be around 0.7 ohm.

I don't have an ammeter.

ROLLGUY · #36 01-30-2017, 04:35 PM

Quote:

Originally Posted by compu_85

Another way to test glow plugs is to get a high wattage lamp, like a headlamp, and ground it through the glow plug.

-J

This would be fairly easy to do, so I will try it.

funola · #37 01-30-2017, 04:47 PM

Quote:

Originally Posted by ROLLGUY

I don't have an ammeter.

Most DVM's has amps. The high amps scale has a separate input from the V, ohms, ma input. My $20 HF DVM has a 20 A input, the free HF DVM has a 10A input which is not high enough.

97 SL320 · #38 01-30-2017, 07:09 PM

You need to use a scan tool to determine what codes are being set. Without this info random parts will be changed and not fix the problem.

Good factory shop manuals ( not being MB specific here. ) will list trouble codes and the conditions that will set them, this is vital when diagnosing a failed system.

ROLLGUY · #39 01-30-2017, 08:23 PM

I borrowed a scan tool, and the code is glow plug/heater circuit A. If I clear the codes and restart it, the glow plug light goes out, and almost immediately comes back on. After a couple more starts, the CEL comes on. I will get a glow plug relay from my friends parts car tomorrow and try that. I am almost certain all the glow plugs are good. I replaced all but two last year, and I can't imagine that they all are bad because they test at 1.4-1.9 ohms. Especially since I can just turn the key without glowing and it will not start. If I start it normally, waiting for the light, it starts right up.

jay_bob · #40 01-30-2017, 08:29 PM

I would definitely pull codes to see what is wrong. A cheap $20 reader from Amazon will work for this particular task (ECU codes, and very limited transmission codes are the only things passed to the OBDII port on these cars). Hate to see you chasing your tail on GPs when it could be something stupid.

Also note that the GP can suffer a failure where it ohms correctly but still does not get hot at the tip. The heater is a coil of wire running the length of the tip. Well once that coil gets to the end of the element it has to get back out to find ground. So the ground wire runs back down the length of the element and out to the shell. If the coil shorts to the return wire at the halfway point, the glow plug will still show plausible resistance, but will not heat at the tip. I have seen posts on this forum showing this effect, the plug is clamped in a vice and hooked to a car battery, and the middle of the shank is glowing but the end is dark. Proper functioning glow plugs have a dimly glowing shank and brightly glowing tip.

I say plausible resistance, because the only way to accurately determine the resistance of a glow plug is with a meter that costs multiple orders of magnitude more than a HF meter. Your meter's basic accuracy, and contact resistance at the probe tips, will overshadow any variances in the resistance of the heater coil. We have to do this at work, to check the integrity of bolted bus bar joints. We have a special meter (called a Ductor, I think that's a brand name) that injects several tens of amps across the joint and measures the resultant voltage drop and thus calculates the contact resistance (in milliohms).

An ohmmeter test won't tell you anything about the health of the plug, other than if it has gone total open circuit.

jay_bob · #41 01-30-2017, 08:39 PM

Our posts crossed. The behavior of the GP light and CEL is consistent with glow plug failure indication.

Hey just thought of something. Our friend tjts1 had this problem. The GP relay on these cars has a power connection, a glow plug harness, and a two wire harness back to the engine electronics. It's not like the 124 and earlier, where you have terminal 15 (run) and terminal 50 (start) and a dumb connection to a light bulb. The two wire harness has a one wire data connection to the ECU. It communicates back and forth across this one wire, the ECU tells the relay it's time to glow, and the relay communicates back that it is glowing or if it has a problem using serial data. One wire is brown, that is the ground, (Check this!) and the other pin has 2 wires, one wire is white which is the data connection to the ECU and the other is violet which is terminal 50 (start signal).

Tjts1 had a problem where the data plug connection was failing and causing problems with the preglow system. I know he posted on this about a year and a half ago.

Mxfrank · #42 01-30-2017, 08:51 PM

The way the relay detects a bad plug is by measuring current draw across each GP circuit and then comparing groups of two or three plugs. If the current draws don't balance, then it throws the signal. A variation of half an ohm at the plug may result in a significant imbalance.

jay_bob · #43 01-30-2017, 09:31 PM

I surmise it is groups of 2 since the OBDII codes on this car are in groups of 2 cylinders.

Later engines (like my OM642) have an individual code per plug.

ROLLGUY · #44 01-30-2017, 09:50 PM

Quote:

Originally Posted by jay_bob

I surmise it is groups of 2 since the OBDII codes on this car are in groups of 2 cylinders.

Later engines (like my OM642) have an individual code per plug.

That could be why the fault code is for circuit A. I suppose that means that the problem is in cyl one and two? I can still do the glow plug in the soup can trick on 1&2, and hope that fixes it.

ROLLGUY · #45 01-30-2017, 09:53 PM

Quote:

Originally Posted by jay_bob

Our posts crossed. The behavior of the GP light and CEL is consistent with glow plug failure indication.

Hey just thought of something. Our friend tjts1 had this problem. The GP relay on these cars has a power connection, a glow plug harness, and a two wire harness back to the engine electronics. It's not like the 124 and earlier, where you have terminal 15 (run) and terminal 50 (start) and a dumb connection to a light bulb. The two wire harness has a one wire data connection to the ECU. It communicates back and forth across this one wire, the ECU tells the relay it's time to glow, and the relay communicates back that it is glowing or if it has a problem using serial data. One wire is brown, that is the ground, (Check this!) and the other pin has 2 wires, one wire is white which is the data connection to the ECU and the other is violet which is terminal 50 (start signal).

Tjts1 had a problem where the data plug connection was failing and causing problems with the preglow system. I know he posted on this about a year and a half ago.

I suppose I could check the connection on these wires, that is easy enough to do.