TDi300 EDC Cold Start Issue - timing is everything...

[DiscoTdi]

OK, it’s been a few minutes since we had a EDC cold start thread so I thought I’d start one….:biggrin:

We poor old afflicted EDC owners need to get to the bottom of this issue once and for all, it’s gone on too long and the non-EDC community are getting too cocky and smug….personally I think the non-EDC folk are only jealous that they don’t have an excuse to play with their Discos and feel like they are missing out ;)


Just like others, I've spent a bit of time on my auto trying to improve the cold start performance and gone through all the usual suspects to no avail. Glow plugs, earth straps, battery power, starter motor, compression, fuel lines etc.

One theory I initial subscribed to was the lack of sufficient fuel on cold starts, however when mine does eventually fire it’s accompanied by copious clouds of white smoke, so this tends to indicate the opposite in that plenty of fuel is being injected during the cranking.

So, based on some interesting experiments my latest theory is that the injection timing advance is insufficient at cold temperatures and the cause of the cold starting issues.

The EDC takes input from various sensors including the coolant temp sensor fitted to the cylinder head. On mine this is nominally around 4K Ohm at 1deg C and around 600 Ohm at running temp.

This sender input to the EDC may have an effect on fuel quantity delivery, but definitely has an affect on injection timing. I have proved this by rigging up a variable potentiometer and adjusting the value of resistance with the engine running.

With a cold engine I can fool the ECU into thinking the engine is actually hot by suitable adjusting the resistance. This greatly reduces the diesel knock to almost inaudible levels, accompanied by smoke. This tells me that the EDC is retarding the injection timing based on input from the coolant sensor and would conversely advancing the injection timing on a cold engine.

The injection timing is controlled by the EDC using an inductive pickup on No 4 injector coupled to the crank rotation sensor and therefore it is not adjustable in the traditional sense. If you adjust the pump timing mechanically it will be overridden by the EDC.

However there must be a limit to the range that the EDC/pump can adjust and one possible explanation is that when the timing belt is replaced, because the pump does not need to be critically timed/aligned (because we are told the EDC takes care of that) it could be so aligned that even when the EDC demands max advance, it is unable to provide this and comes up against the stops.

All theory of course, but this could explain this phenomenon of some the EDC equipped cars not starting like they should and indeed would have when new, as most would have had the timing belts changed…

I will be checking the timing on mine at some stage and report back.

Regards - Martin.

[thebiglad]

OK, it’s been a few minutes since we had a EDC cold start thread so I thought I’d start one….:biggrin:

So, based on some interesting experiments my latest theory is that the injection timing advance is insufficient at cold temperatures and the cause of the cold starting issues.

The EDC takes input from various sensors including the coolant temp sensor fitted to the cylinder head. On mine this is nominally around 4K Ohm at 1deg C and around 600 Ohm at running temp.

This sender input to the EDC may have an effect on fuel quantity delivery, but definitely has an affect on injection timing. I have proved this by rigging up a variable potentiometer and adjusting the value of resistance with the engine running.

With a cold engine I can fool the ECU into thinking the engine is actually hot by suitable adjusting the resistance. This greatly reduces the diesel knock to almost inaudible levels, accompanied by smoke. This tells me that the EDC is retarding the injection timing based on input from the coolant sensor and would conversely advancing the injection timing on a cold engine.

The injection timing is controlled by the EDC using an inductive pickup on No 4 injector coupled to the crank rotation sensor and therefore it is not adjustable in the traditional sense. If you adjust the pump timing mechanically it will be overridden by the EDC.

However there must be a limit to the range that the EDC/pump can adjust and one possible explanation is that when the timing belt is replaced, because the pump does not need to be critically timed/aligned (because we are told the EDC takes care of that) it could be so aligned that even when the EDC demands max advance, it is unable to provide this and comes up against the stops.

All theory of course, but this could explain this phenomenon of some the EDC equipped cars not starting like they should and indeed would have when new, as most would have had the timing belts changed…

I will be checking the timing on mine at some stage and report back.

Regards - Martin.

Interesting thought Martin, I will be doing a cambelt on both of ours within the next couple of months (not now, no garage and it's minus 8°!!).

I wonder if my Difflock cambelt kit will help in the accuracy of setting the EDC pump, or whether the pulley on it is different to the one on the manual one??

I just know that mine should start more readily than it does.

The manual TDi300 fired up yesterday first pull, 2-3secs of cranking and it had been -10°c !!!!

Cheers
Dave

[Jon v8]

A slack timing belt will cause retarded timing,BUT the advance is not set by the needle lift sensor,its done by the crank sensor.The needle lift sensor just gives feedback - closed loop timing,if you like. So if you "modify" the pump timing too much all you will get is a check engine lamp as the needle lift sensor will tell the ecu when it starts to inject and if the pump cant retard the point of injection enough to get back where it wants you will put the light on.
If you are getting lots of white smoke I'd have the injectors out and have them tested for pattern and pressure.While the injectors are out do a compression test too.
Remember these engines are no longer young,and because they are so good people think they will last for ever.(Plus most have now had several years of neglect,rubbish fuel and water through the fuel system - none of which helps.)

[DiscoTdi]

Interesting thought Martin, I will be doing a cambelt on both of ours within the next couple of months (not now, no garage and it's minus 8°!!).

I wonder if my Difflock cambelt kit will help in the accuracy of setting the EDC pump, or whether the pulley on it is different to the one on the manual one??

I just know that mine should start more readily than it does.

The manual TDi300 fired up yesterday first pull, 2-3secs of cranking and it had been -10°c !!!!

Cheers
Dave

Hi Dave, -10 deg! that's as good a test as any :biggrin:

Agreed the EDC should start more readily.

I've not done the cambelt on one of these, but from what I have heard the Difflock kit is supposed to be execellant quality and value, so should give the necessary accuracy.

Just been reading the difflock article and the pump pulley has slotted holes, so I'm assumung that was for a manual, so yes be interstesting to see if the EDC is the same. After two cam belts you'l be an expert!

Hopfuly we will get to the bottom of this.

Cheesr for now - Martin.

[sotal2]

I had major issues with mine over winter last year.

I purchased the Disco at the end of August, and had a few issues which I sorted. All was then well until the temp dropped in October. When this happened it no longer started and required a jump from another battery to get it started.

First attempt at fixing was to buy a new battery, the highest amp which would fit. No difference

Next attempt was to replace one duff glow plug and one questionable glow plug. No difference

Next attempt was to remove the fuel cut off solenoid. No difference

Next attempt was to add extra earths in the form of jump leads from negative to various parts of the block - No difference

Next attempt was to play with the pump timing. After many changes - NO difference.

Next attempt was to have the Bosch Starter Motor Reconditioned. The recon place phoned me to say it had performed flawlessly on the test bench, would they like me to still look inside. I told them to anyway as they had it. They ended up changing a "com bush" as it was slightly questionable and they were in there anyway. When I came to but the starter motor back I took the earth leads off and cleaned them properly. I used a wire brush in the drill and cleaned all contact points to bare shiny metal. I also moved the lead which went to a bolt near the starter to actually go on the starter mounting bolt. - Seemed to be sorted

By the time I sorted it, it was February, but 2 weeks after we had a cold spell, and it still started straight off every morning. Even had a day with 3 inches of snow on the bonnet and it started perfectly. After that the weather improved and I ended up selling it in May but it started perfectly every day no matter what the weather. No idea how it's performed this winter.


I still think it's down to the condition of the basic electrical circuits. When I measured the voltage across the battery it was at aout 12.5v at rest, on cranking the voltage would drop below 10v at this voltage the starter would turn but the electronic fuel injection pump wouldn't squirt the fuel.

After sorting the earth leads / starter motor the voltage never dropped below 10v on cranking.

I think it needs really good earths. Adding Jump leads as earths wasn't good enough for mine to start, and I had "cleaned" the earths before - which was just a quick scratch with a wire brush where I could easily get to. When I did the proper clean it sorted it properly.

Hope this helps, and try a Voltage Drop test when it's refusing to start and let us know the results.


If it helps when mine refused to start there was no way it would start - didn't matter how long I left it cranking for, without an extra battery it would not start up. I could however leave it until later in the day when the sun came up and warmed the vehicle up and it would sometimes start.

[DiscoTdi]

thanks for the pointers Jon.

Regarding the injectors, she runs dead sweat with no smoke once started, even when flooring it, it's amazingly smoke free, so I don't think there's anything wrong there. The engine is in nice fettle with 44K on the clock.

The only time I get the smoke is after the extended cranking at cold start.

I might be being a bit thick here, but I'm not sure a slack timing belt would cause the timing to retard as any slack in the belt will be on the unloaded side of the belt and therefore have no influence on the angular relashionship between the two pulleys ?

Yes the injector sensor and crank sensor work together, but I'm not looking to adjust the timing outside of this closed loop function. By using the coolant sensor to get the EDC to adjust the timing, it is all under closed loop control and was used to prove the cause and affect of this particular sensor. I agree if we go outside of this loop it could cause the EDC to flag up an error.

The theory is that the EDC is unable to advance the timing as required, due to the pump not being aligned in a neutral enough position to allow sufficient +ve or -ne adjustment by the EDC within it's capability.

Of course it may be the cooland sensor is duff, so I am looking at some other tests to eliminate that.

Cheers

[DiscoTdi]

I had major issues with mine over winter last year.

I purchased the Disco at the end of August, and had a few issues which I sorted. All was then well until the temp dropped in October. When this happened it no longer started and required a jump from another battery to get it started.

First attempt at fixing was to buy a new battery, the highest amp which would fit. No difference

Next attempt was to replace one duff glow plug and one questionable glow plug. No difference

Next attempt was to remove the fuel cut off solenoid. No difference

Next attempt was to add extra earths in the form of jump leads from negative to various parts of the block - No difference

Next attempt was to play with the pump timing. After many changes - NO difference.

Next attempt was to have the Bosch Starter Motor Reconditioned. The recon place phoned me to say it had performed flawlessly on the test bench, would they like me to still look inside. I told them to anyway as they had it. They ended up changing a "com bush" as it was slightly questionable and they were in there anyway. When I came to but the starter motor back I took the earth leads off and cleaned them properly. I used a wire brush in the drill and cleaned all contact points to bare shiny metal. I also moved the lead which went to a bolt near the starter to actually go on the starter mounting bolt. - Seemed to be sorted

By the time I sorted it, it was February, but 2 weeks after we had a cold spell, and it still started straight off every morning. Even had a day with 3 inches of snow on the bonnet and it started perfectly. After that the weather improved and I ended up selling it in May but it started perfectly every day no matter what the weather. No idea how it's performed this winter.


I still think it's down to the condition of the basic electrical circuits. When I measured the voltage across the battery it was at aout 12.5v at rest, on cranking the voltage would drop below 10v at this voltage the starter would turn but the electronic fuel injection pump wouldn't squirt the fuel.

After sorting the earth leads / starter motor the voltage never dropped below 10v on cranking.

I think it needs really good earths. Adding Jump leads as earths wasn't good enough for mine to start, and I had "cleaned" the earths before - which was just a quick scratch with a wire brush where I could easily get to. When I did the proper clean it sorted it properly.

Hope this helps, and try a Voltage Drop test when it's refusing to start and let us know the results.


If it helps when mine refused to start there was no way it would start - didn't matter how long I left it cranking for, without an extra battery it would not start up. I could however leave it until later in the day when the sun came up and warmed the vehicle up and it would sometimes start.


Thanks for the useful info, sounds like all the things I have checked !

It's worth noting that mine will always start, it just needs a lot of cranking during which time it splutters, fires and coughs then eventually catches. Battery terminal voltage drops to 10v during cranking, but fuel is going in during the cranking as evidenced by the white smoke when she eventually starts.

If I repeatedly energize the glow plugs prior to starting to really warm things up it helps a lot.

Cheers

[Stu2985]

1996 300 tdi auto with aircon edc

started great in summer come colder days I had issues starting it but turned out 2 of the glow plugs were burnt down

new glow plugs and I have been lucky its never failed to go first time

Have u checked for any air getting into the fuel system - even a pin ***** on somewhere like pipes that come from fuel tank on a cold day I was led to belive could cause a problem. When its warm or a warm day the tiny amount wont make any difference.

worth a check maybe

[Tom Mepham]

Timing is everything.. as you rightly say! White smoke on start up often means that the injection timing needs to be advanced a fraction. We had a similar problem on a Perkins 4.108 we had in a truck some years ago, this had bevelled timing gears that made it very difficult to time accurtaly.
Another example with cold starts and advancing the timing is the perkins prima found in maestro and montegos. We have one fitted up in our boat(with 37hours on the clock from new!) The fuel pump (manually controlled) on this has a cold start advance solenoid that is controlled by a thermistor screwed into the block so its in the water jacket.
This was on wired via the thermisor when we got the engine and we used to get white smoke on start up, with it wired correctly its a puff of black and its away.
The Prima uses very similar Bosch fuel pumps and injectors (and its direct injection) to the LR Tdi, it was developed at the same time as the 200tdi.
Tom

[Jon v8]

Timing is everything.. as you rightly say! White smoke on start up often means that the injection timing needs to be advanced a fraction. We had a similar problem on a Perkins 4.108 we had in a truck some years ago, this had bevelled timing gears that made it very difficult to time accurtaly.
Another example with cold starts and advancing the timing is the perkins prima found in maestro and montegos. We have one fitted up in our boat(with 37hours on the clock from new!) The fuel pump (manually controlled) on this has a cold start advance solenoid that is controlled by a thermistor screwed into the block so its in the water jacket.
This was on wired via the thermisor when we got the engine and we used to get white smoke on start up, with it wired correctly its a puff of black and its away.
The Prima uses very similar Bosch fuel pumps and injectors (and its direct injection) to the LR Tdi, it was developed at the same time as the 200tdi.
TomI think you may be missing the point that this thread relates to 300 TDI Auto's fitted with the EDC Bosch pump - EDC = Electronic diesel control,apart from the basic shape of the pump and the high pressure side nothing is the same as the manually governed VE pumps.The problem is that the EDC pump has a lower start quantity than a manual VE,(Therefore dumb in terms of atmospheric conditions.) so it relys on working glow plugs,a decent cranking speed and correct sensor inputs to start in low temps.Its as simple as that,the only problem with them is poor earthing from the factory,fix that and they start with no problems.

[DiscoTdi]

Timing is everything.. as you rightly say! White smoke on start up often means that the injection timing needs to be advanced a fraction. We had a similar problem on a Perkins 4.108 we had in a truck some years ago, this had bevelled timing gears that made it very difficult to time accurtaly.
Another example with cold starts and advancing the timing is the perkins prima found in maestro and montegos. We have one fitted up in our boat(with 37hours on the clock from new!) The fuel pump (manually controlled) on this has a cold start advance solenoid that is controlled by a thermistor screwed into the block so its in the water jacket.
This was on wired via the thermisor when we got the engine and we used to get white smoke on start up, with it wired correctly its a puff of black and its away.
The Prima uses very similar Bosch fuel pumps and injectors (and its direct injection) to the LR Tdi, it was developed at the same time as the 200tdi.
Tom

Hi Tom, thanks for the interesting info, 37 hours, probably not even run in then :)

Did you have to do much conversion to make it marine worthy - cooling system, exhaust etc ?

Cheers

[DiscoTdi]

1996 300 tdi auto with aircon edc

started great in summer come colder days I had issues starting it but turned out 2 of the glow plugs were burnt down

new glow plugs and I have been lucky its never failed to go first time

Have u checked for any air getting into the fuel system - even a pin ***** on somewhere like pipes that come from fuel tank on a cold day I was led to belive could cause a problem. When its warm or a warm day the tiny amount wont make any difference.

worth a check maybe

Thanks for the tips, I must admit I have not specifically checked for air leaks, I'll add that to the list. Regarding the glow plugs, I had them all out and checked on the bench. I did have one faulty one which has been replaced.

Cheers

[DiscoTdi]

Quick update on progress.

This morning I increased the resistance fed to the ECU from the coolant sensor by an additional 5.6K ohm, bringing the total to around 10K ohm on a cold engine.

This made an immediate improvement and she started straight off perfectly, no hesitation or spluttering or smoke. I did wait for the glow plug light to extinguish, so the acid test would be to try starting without waiting for the light, but even so it seems this has cured the problem (crosses fingers).

Obviously this is a temporary fix and I still need to identify the root issue. It could be the thermistor sensor is faulty and reading low or this timing advance hack is masking another faulty component.

However the test also confirms that the injection timing is able to be advanced sufficiently by the ECU and so rules out possible pump mis-alignment issues discussed earlier, which is good.

Progress at last, watch this space...

[DiscoTdi]

Quick update on progress.

Started fine again this morning, less than 1 sec cranking :o

[thebiglad]

Quick update on progress.

Started fine again this morning, less than 1 sec cranking :o

Hi Martin, couldn't attach the pics to your pm, so here they are. Incidently it's a '98 Tdi300 auto EDC.

Note that there are NO sensors on the ally elbow of the thermostat housing.

[Tom Mepham]

Jon and DiscoTdi
Sorry to be a bit late, I was thinking that perhaps that the problem might lie with the sensors that supply the fuel pump with information for it to base its timing strategy on.
Would it be possible to fool the injection pump into advancing the timing more to see if that improved the cold start?
As for the Prima in my boat, it was "professionally" marinised by a Muppet and installed in the boat by a complete Muppet! That was before I got my hands on the boat. I have subsequently removed, cleaned, painted and re installed the engine.
It has a heat exchanger, raw sea water to engine coolant to cool the engine. It has a custom water cooled jacketed exhaust manifold that has a port that takes the sea water from the engine heat exchanger and discharges it out of the exhaust pipe.
This is another hobby, even more expensive than landrovers! Lol I will try and up load some pics of what we have done for those of you interested.
Tom

[Tom Mepham]

Jon and DiscoTdi
Sorry to be a bit late, I was thinking that perhaps that the problem might lie with the sensors that supply the fuel pump with information for it to base its timing strategy on.
Would it be possible to fool the injection pump into advancing the timing more to see if that improved the cold start?
As for the Prima in my boat, it was "professionally" marinised by a Muppet and installed in the boat by a complete Muppet! That was before I got my hands on the boat. I have subsequently removed, cleaned, painted and re installed the engine.
It has a heat exchanger, raw sea water to engine coolant to cool the engine. It has a custom water cooled jacketed exhaust manifold that has a port that takes the sea water from the engine heat exchanger and discharges it out of the exhaust pipe.
This is another hobby, even more expensive than landrovers! Lol I will try and up load some pics of what we have done for those of you interested.
Tom

Yes that is 90psi of oil pressure in that photo!
And the dip stick showing the original engine oil from new.

[andymach23]

Jon and DiscoTdi
Sorry to be a bit late, I was thinking that perhaps that the problem might lie with the sensors that supply the fuel pump with information for it to base its timing strategy on.
Would it be possible to fool the injection pump into advancing the timing more to see if that improved the cold start?
As for the Prima in my boat, it was "professionally" marinised by a Muppet and installed in the boat by a complete Muppet! That was before I got my hands on the boat. I have subsequently removed, cleaned, painted and re installed the engine.
It has a heat exchanger, raw sea water to engine coolant to cool the engine. It has a custom water cooled jacketed exhaust manifold that has a port that takes the sea water from the engine heat exchanger and discharges it out of the exhaust pipe.
This is another hobby, even more expensive than landrovers! Lol I will try and up load some pics of what we have done for those of you interested.
Tom

Nice looking boat Tom. Is that a Colvic hull? You're right that boats are more expensive than Land Rovers. Just as much fun to work on though. It would be interesting to see some more pics of what you have done.

Cheers

Andy

[DiscoTdi]

Jon and DiscoTdi
Sorry to be a bit late, I was thinking that perhaps that the problem might lie with the sensors that supply the fuel pump with information for it to base its timing strategy on.
Would it be possible to fool the injection pump into advancing the timing more to see if that improved the cold start?
As for the Prima in my boat, it was "professionally" marinised by a Muppet and installed in the boat by a complete Muppet! That was before I got my hands on the boat. I have subsequently removed, cleaned, painted and re installed the engine.
It has a heat exchanger, raw sea water to engine coolant to cool the engine. It has a custom water cooled jacketed exhaust manifold that has a port that takes the sea water from the engine heat exchanger and discharges it out of the exhaust pipe.
This is another hobby, even more expensive than landrovers! Lol I will try and up load some pics of what we have done for those of you interested.
Tom

Your on the money there, I've proved that I can fool the ECU into advancing the timing and makes a huge difference (see previous posts in this thread).

Nice clean engine and pretty tight by the looks of that oil pressure !

I remember the Prima being a bit noisy in the old Montegos/Maestros, I expect installed in the boat with all that wood she sounds a bit more muted
?

Thanks for the info and pics Tom, very interesting.
Cheers

[Tom Mepham]

Andy and DiscoTdi
Yes that is a Colvic 20 hull, about the only thing that remains from before we got the boat is the engine and bear Fiberglass hull (as in we stripped it back to bear hull!). As for the noise, it is very distinctive, we have a fitted ply engine box that has 25mm heavy rubber sound proofing inside it. Have not tried it with the sound proofing fitted yet but just the ply box made quite a difference. We did some trials in it back in October, we have not finished fitting out the inside yet so it back out of the water at the moment.
It will be interesting to see what the fixed timing is like on your engine, do we know if this was this a problem on these engines from new?
Tom

[DiscoTdi]

Andy and DiscoTdi
Yes that is a Colvic 20 hull, about the only thing that remains from before we got the boat is the engine and bear Fiberglass hull (as in we stripped it back to bear hull!). As for the noise, it is very distinctive, we have a fitted ply engine box that has 25mm heavy rubber sound proofing inside it. Have not tried it with the sound proofing fitted yet but just the ply box made quite a difference. We did some trials in it back in October, we have not finished fitting out the inside yet so it back out of the water at the moment.
It will be interesting to see what the fixed timing is like on your engine, do we know if this was this a problem on these engines from new?
Tom

Yes I bet she will sound quite nice, probably a nice deep throb and should give the boat a good turn of speed I'd have thought ?

I had a TDi300 EDC auto from new back in 1995 and I can't say there was a problem with starting in the winter, if there was I'm sure I would have remembered.

I'm hoping that it's a case of the EDC models coolant sensor goes bad after a few years, so I've ordered a new sensor (the one on teh cylinder head that feeds the EDC) as they are quite cheap. If nothing else I can compare the readings and eliminate if not the root cause and look else where (after cursing profously).

I started her up the other evening in the cold and didn't wait for the glow light to extinguish and still started within a couple of revolutions.

Cheers

[Tom Mepham]

Yes it dose throb a bit, so far we have had nearly 12knots out of it! The Engine is light weight, high speed and 60BHP, it makes the whole thing lift- really the boat is quite over powered but it has some interesting results!
As for the main topic of conversation, I fitted new heater plugs in my 300tdi(Not EDC) in October 2007. Four new ones as all of them where duff.
I noticed during the cold spell we have had in over the last week or so that the engine was winding a bit before starting at cold but still starting well.
Tested the heater plugs this morning and their all dead again!
Ordered 4new ones from paddocks just now- the new ones I fitted have been in for 15months and 14,500 miles, not lasted well in my books!
Tom