Glowplug relay on 2003 TDI Jetta. Where?

The glow plugs get no juice and hence no starting the engine. So I found a relay panel above the left knee of the driver but it is confounding me as it is different from as shown on my "generic" Chilton circuit diagram.

Also, could not find any group of 4 black wires with red, green, dark and light blue tracers as they come off the glow plug harness on the engine and then pass through the fire wall to the instrument cluster. The "100" labelled relay I assumed to be for the glow plugs has two heavy wide spade terminals for like 40A and two smaller ones for the solenoid. It ohm-ed out OK and clicks when powered too but I found no continuity from the sockets the glow plugs. Also, there were no mice chewing marks or burnt smell. My diagrams however show 2 two more but otherwise functionally similar relays with 4 contacts but my panel has some larger multicontact relays. Googling showed a longish rectangular relay, plus a squarish one, so which is which in the age of fake news? The car was Mexico assembled, if this helps t ID it.



Mirrors did not help me picking up this color scheme wiring below the steering wheel but I did trace the wiring from the GP's on the engine block to underneath the battery tray and then into a covered plastic channel about 2" wide. This hid a 4-pin connector with identically color coded wiring which then dives into the firewall to the instrument cluster area. All wiring appears to be intact and this problem only started after I replaced the turbo and cylinder top as the cam bearing were seized. The engine now starts and runs ok but only if I briefly McGiver 12V to the glowplugs directly. Only then will it will fire up.

Is this relay out of sight perhaps, accessible from above only? Too difficult to see as the harness is wrapped and there is other stuff in the way. There is also a small fuse-like 2-pin relay, which on opening up appears to be a thermal timed delay type, presumably for the glowplugs. The relay is is both quite visible and accessible with two heavy red wires to it.

On the 03 the relay is mounted up where the windshield meter is you’ll have to pull the wipers and remove the valance. It a 180 relay. Check idparts.com for a picture.

Thanks TDIjarhead, I will check that info out and let you know. The car is
not right with me and it will only be next week some time when I will be able to get to it.

Did you check if the fuses on top of the battery are OK?
I think there are 2 fuses, 1 for two glow plugs.

wonneber said:

Did you check if the fuses on top of the battery are OK?
I think there are 2 fuses, 1 for two glow plugs.

Click to expand...

Wonneber is right. A no power glow plug issue is more likely that fuse panel on top of the battery.

There are 2 50A fuses there. Pull them off clean them up, check for any melting of plastic or other damage. Very common with that fuse panel.

You know, I did check but only with my eyeballs and not with meter. (My bust!)

I've repaired and worked with electronics most of my life and those fuses looked pristine, as did the connections. No burnt or melted plastic and IMHO there should have been connection to 12V just by looks. High current ( like the plugs would draw ) does do funny things at times, so thank you for this tip. Per my available circuit diagram other stuff should also have failed if those big fuses had failed, or any of them, so I did not check further.

The little fuselike relay on the outside of the underdash relay board was o/c however, until I pried off the aluminum shielding can to wiggle the contacts. I thought I'd solved the problem at that point but no luck! Still don't know what that one does and it may well be a fail-safe part of the GP relay circuit, to prevent them staying on and burning out in case the main relay contacts welded shut.

Back the Fail Train up a bit.... why do you think you have a glow system problem?

The car won't start? Unless it is about 10 F, the car will start without the glow system working. It may be cranky and smokey and noisy if it is below 40 F or so, but it WILL start.

Do you have any DTCs? The ECU is VERY good at detecting and flagging faults with the glow system. If you have no DTCs related to that, you can rest assured your glow system is fine.

Oilhammer, there is no voltage at the plugs, zip. Only when I directly wired 12V to the plugs with 14gauge wire does the thing start. On my boat I had a similar issue that the 3-cylinder Perkins did not start without the plugs working.

It may be a timing issue on the Jetta as well, as the head and timing belt were replaced and no software tune-up for advance/retard was done afterward, but even repeated attempts at cranking did not get things to run. Once the 12V was applied, no problem. ( The car is at 3300' altitude, so the air is a little skinnier there )

Might be the engine/controls are not in good shape. What is the temperature there?
I'm confident you can get the glow plug thing sorted out. Relays, fuses, wires all occasionally fail.

I don't care about the voltage at the plugs, I care if the ECU cares. If the conditions are not calling for a preglow, then there will be no voltage.

If there are no glow related DTCs, then your glow system is conditionally intact. The ECU monitors this. You need to find out what the ECU thinks the coolant temp is, and go from there. You are attempting a diagnostic strategy based on a false assumption. Perhaps there is some other issue with the engine, and an aggressive preglow is allowing it to start when it otherwise would not.

You can pull the CTS plug off, and the ECU will default to a full preglow, about 10 seconds or so (I forget).

You say you are at high altitude. That's fine. The ECU should know this. Does it? This data is available. Computers are dumb. Bad input means bad output. DTCs mean the ECU has flagged something, you still have not answered that question.

Good point about the ECU calling a glow need. My cheepo code reader only gave me a single I think 0380 or was it a 0382 reading, can't remember so ignore this. The little code chart I have said something about a #3 cylinder misfire for that code.

However, the 12V did make the difference. Now, if that negated ANOTHER issue, I really cannot say. It may well be so but I am unable to tell. Do not have a laptop or fancy reader to monitor the ECU ( dont know about the CTS plug either, sorry! )

Figured if I can get the car into at least a limp mode a more competent or better equipped guru can then fine tune it. Ich Dummkopf vill chust sticken die hands in den pocketsen, votch z blinken lites und zen leaven den adjusten to den Experten.

BTW Oilhammer: the coolant thermometer light was on previously. Adding a little coolant did fix this issue though and air temp was well above 85F.

It is very difficult, impossible really, to troubleshoot any modern vehicle without a suitable scan tool. There is just only so much guessing and parts swapping you can do.

The threshold for the ALH's preglow period is around 45 F, afterglow will happen at higher coolant temps, but that is after the engine is started, and that is a different strategy for cold engine emissions reduction and not for starting aid. But the parts that make it work are the same, and the ECU monitors this just the same.

The glow system is very simple, and pretty reliable, on these cars. Relays rarely fail. Plugs fail, and sometimes the harness, but both would give the appropriate DTC and are unlikely to cause a no start all at once because each plug would have to die at the same time, which simply does not happen.

If you have a really sick engine, or some really pooched injectors, or some other such issue, that will not allow the engine to make sufficient heat in the piston bowl during cranking, and adding an "artificial" preglow period where there would normally be none, and it lets the engine start, then that is not a glow system problem. That is an engine or fuel problem.

But even half million mile ALHs rarely have issue starting in moderate cold temps, let alone whatever temps you are experiencing in your area right now.

Very true Oilhammer, I sure wish I had such a tool but then I'd also need the diagnostic code crossreferences etc etc etc. This all adds up!

The compression after the head xchange I measured at about 380psi, which seems to be about in the ballpark for a like 180k engine with just ground in valves that have not had a chance to fully seat yet. ( guessing here mostly and I only tested #one cylinder, just to see if I could get #1 firing with a hot glowplug. )

What do you suggest i should do at this point? There does not seem to be a real idling issue now but I did note a gentle slow surge and relax. Not all the hoses of air filter etc are on and this may not help that ECU to diagnose it all fully. Ck engine light I believe is also still on.

If you are anywhere where the ambient temp is above freezing and it "won't start w/o pre-glow" there is something really wrong. Either the compression is (very) low, injection timing is very retarded (not advanced enough), cranking speed is sub par, etc. Using/needing pre-glow in such temps is masking an issue not related to the GPs.
As Oilhammer said, if there was a fault in the GP system the ECU would be letting you know about it.

I call timing retarded.
Glow plugs allow enough heat to ignite the fuel.

I recommend looking at the pump bolts and seeing if you have a lot of room on either side of the oblong holes for timing.

If they're off one way center them up and try starting it then.

If the bolts were loose enough to move it wouldn't start.

Sent from my Armor_2 using Tapatalk

The OP needs a scan tool. Period. Can't properly see what DTCs are present, can't see ANY data, can't check/verify the ECU is getting the correct inputs, can't check/verify timing, can't do a whole lot of anything.

VCDS is CHEAP. Any old PC will work it, you can find brand new cheapo laptops for $300, you can find used ones for half that. I am running it on a 12+ year old Asus Eee PC that was $220 brand new, could probably find one for $50 now.

If you can't, or won't, get the proper tools, then just take the car to someone that does.

Thanks Jettawreck. ( how DO you come up with such names??? ) It is quite possible
that both you and Oilhammer are totally correct. Trouble is my $40 code scanner did
not help me very much and I really do not know at this point how to fiddle with the
pump timing short of rotating the pulley vs a fixed TDC on crankshaft.

Not sure if the fuel pump is locked in mechanically or can be advanced/retarded via software?

Yes they are, but it is the interweb, you'll need to track down your issue using the advice.
And yes you can adjust timing using VCDS (or equal), guessing most don't. But you're getting ahead of yourself.

OK, I get it. Scan tool, then the manual on how to use it etc. Best to take it
to someone who has that learning curve behind him already.

Checking the 3-bolt orientation in the pump seems the quickest and easiest
fix to me at this point to get things into the ballpark. ( thanks eddilee and bob....OH and others for all your help ) The car sits at an off-grid location right now with very poor cell/web contact but I will let you know when I can get back to it and try some things.

Not just a generic scan tool. If you are going to own VW (perhaps especially a TDI) and work on things like injection timing it's almost hopeless to not have VCDS. The timing graph makes it so easy. The logs, the adaptations, etc, etc. It's not just for reading codes.

Ivoskis said:

Thanks Jettawreck. ( how DO you come up with such names??? ) It is quite possible
that both you and Oilhammer are totally correct. Trouble is my $40 code scanner did
not help me very much and I really do not know at this point how to fiddle with the
pump timing short of rotating the pulley vs a fixed TDC on crankshaft.

Not sure if the fuel pump is locked in mechanically or can be advanced/retarded via software?

Click to expand...

Well, the first TDI I bought was a 2001 Jetta. It was only a couple years old w/33k miles. Prior owner had rolled it into a ditch and it was pretty well wrecked. Body damage in multiple places and it turned out the engine was ruined when the front center engine mount broke and derailed the TB and also broke a sizable piece out of the engine block all the way up to the bottom of the head. The damaged engine is my avatar pic.
It was THE jettawreck!!
Replacement engine, replacement hood, grille, fenders and plenty of bodywork. Live and learn.
Still have and drive it.

You don't want to just start randomly moving the IP sprocket. Tiny increment of movement makes a huge amount of timing change to the point where it won't start/run at all. Really need to know where it is at before you start adjusting and how to get back if needed.
Injection timing needs to be set mechanically before making any software adaptation changes as the mechanical (static) timing is what the engine uses to start up on. After it's actually running the electronics from the ECU start making adjustments.

Ivoskis said:

The compression after the head xchange I measured at about 380psi, which seems to be about in the ballpark for a like 180k engine with just ground in valves that have not had a chance to fully seat yet. ( guessing here mostly and I only tested #one cylinder, just to see if I could get #1 firing with a hot glowplug. )
[some deleted]
Not all the hoses of air filter etc are on and this may not help that ECU to diagnose it all fully. Ck engine light I believe is also still on.

Click to expand...

A bit more info please. Maybe I missed some.

Did the car run at all after the head change?

Why was it changed? Maintenance or belt problem?

In the clear fuel line going to the injector pump do you see a small air bubble at the highest point?

Crack the fuel lines at the injectors loose and have someone crank the engine.
Does fuel come out?
Have a few rags wrapped around the area.

Before shutting the car off scan for codes (even with your code reader)
Post the codes.

Good info there, thank you so much! Yes, I did suck the fuel thru the pump with a hand vacuum pump until I thought I had cleared everything. Also tested that it pumped fuel by cracking the #1 union nut to the injector. My lines are totally web covered so no bubbles are visible.

The turbo failed originally and then I failed to realise that metal junk had gone into the head at that time. Exchanged the turbo ok but very little power and then realized the camshaft was seizing up. That's when a whole Chinese head assembly went in with a new gasket ( earlier posts from 2014, and yes, the car stood idle that long! )


That is the history then. New TB with the head and I tried to match up the old timing as much as I could but it is hard to do. One thought was changing the idler roller diameter with some tape, as that would also change the pump timing by effectively shortening the belt but this then starts to become a science project. It might show incremental improvements though but may be hard to measure.

The CD I got with a supposedly VAG lookalike diagnostic program never really worked out for me, but I can see the value for having a working one. Needless to say I have many other projects going on and travel a lot plus a sailboat with two diesel engines, so I try to keep things simple. At 71 I don't want to have to learn too many new tricks all at once.

NO NO NO NO NO to any of that.

First, by hand roll the engine over to the TDC mark in the bellhousing with a wrench. Jam a screwdriver in there to make very sure it does NOT move. Remove the vacuum pump (and valve cover if you have the old style lock plate) and see if the lock plate goes in. If it doesn't STOP. Note that if this check is off more than ~7 degrees of crank rotation or so you HAVE had valve contact and you're in serious trouble. BTW if this check (crank/cam) is off "a bit" post below and I'll tell you what you have to do to fix it PROPERLY, but if it IS off you had BETTER NOT be off by more than that 7 degrees -- because if you are then you're set up to drop a valve a few thousand miles from now and that'll be the end of your fun.

Let's assume that's NOT the case and the cam and crank relationship is bang-on.

Leave the lock plate and crank locked and see if the pin (for the IP) goes in. If it doesn't then leave the other two locked and FIX THAT (three bolts on the IP sprocket; careful, you only need to move it a TINY bit and leave the IP pulley itself alone -- you want to turn the center, large nut on the hub but ONLY with those three bolts for the pulley cracked loose -- once the IP pin is in then re-torque those three bolts to 18ft/lbs) DO NOT play with anything in or on the idler pulley or similar; anything that gets into the timing belt system or causes anything in there to lock up will cause the engine to jump time and *destroy* the cylinder head.

Then remove all three locks (pin, cam plate and screwdriver in the bellhousing) and roll the engine over CW on the crank bolt TWO full rotations and return it to TDC in the bellhousing, turning ONLY CW. Put the plate and pin back in -- they should both go back in without a problem and without moving the crank position at all. Assuming it all checks put the vacuum pump back on and button it up.

If all three of these are indexed correctly (cam, crank and IP pin) it should start, even without the glow plugs working, assuming the outside temps are even a little bit above freezing. It may not be perfect in terms of where the timing is but it WILL be in the ignition window if those are all properly lined up.

Then find someone with VCDS to figure out what else is going on.

If I'm reading this correctly, the car hasn't run since the head replace?
Follow instruct by Genesis.

Ivoskis said:

Good info there, thank you so much! Yes, I did suck the fuel thru the pump with a hand vacuum pump until I thought I had cleared everything. Also tested that it pumped fuel by cracking the #1 union nut to the injector. My lines are totally web covered so no bubbles are visible.

The turbo failed originally and then I failed to realise that metal junk had gone into the head at that time. Exchanged the turbo ok but very little power and then realized the camshaft was seizing up. That's when a whole Chinese head assembly went in with a new gasket ( earlier posts from 2014, and yes, the car stood idle that long! )


That is the history then. New TB with the head and I tried to match up the old timing as much as I could but it is hard to do. One thought was changing the idler roller diameter with some tape, as that would also change the pump timing by effectively shortening the belt but this then starts to become a science project. It might show incremental improvements though but may be hard to measure.

The CD I got with a supposedly VAG lookalike diagnostic program never really worked out for me, but I can see the value for having a working one. Needless to say I have many other projects going on and travel a lot plus a sailboat with two diesel engines, so I try to keep things simple. At 71 I don't want to have to learn too many new tricks all at once.

Click to expand...

Fail train rolls back to the start.... whooop...

Did get the timing mark at the #4 cyl (far end) of the block centered into the transaxle(?) window, but did not see any other alignment marks. ( supposedly there are some per Chilton )

A perfectly sized pin gauge locked the pump and some gauge block gizmos locked the camshaft as per instructions at the vac pump end. Larger errors may be possible if the timing mark was supposed to be at one corner of the sight hole.... bummer on that then.

Yes, engine did run with this set-up and no valves binding. This was chkd by hand.

The 3 pump pulley bolts seem to be centered in their spaces OK per eyeball.

Did turn the engine over by hand with no big differences noted, but again, if the
timing mark is off by 1/4" that is probably quite a severe error of a degree or more.

At what point do you then slip on the timing belt and what if a groove/ridge do not line
up perfectly?

Don't mean to bring a cloud of doom over this, but with a Chinese head on it I think the fail train will roll into the station eventually sometime after you get this sorted out and running again. Or, perhaps I read or comprehended that part all wrong.