2001 Jetta ALH slow start/difficult to start w/o ECT, worse with ECT

Car is new to me, one previous owner, southern US state driven 2001 late production (10/00) Jetta ALH 5-speed with 194k (presumably mostly highway/freeway) miles currently. It is in very good condition overall for age and mileage. As far as I know the car is stock, with no modifications.

Issue with the car is that when it hasn't been run recently it takes a long time to get started, like 3-8 seconds of cranking, and that's with the ECT sensor disconnected. If the ECT sensor is connected, it can get started, but it takes a very long time, like 10-30 seconds of cranking, and quite a bit more smoke when it finally starts.

In March 2010 the previous owner took the car in for service. His issues were timing belt service was due, car is hard to start/having to pull ECT, and losing diesel fuel on the ground. The shop did timing belt service (including water pump) that was due, and replaced the fuel injector pump and fuel filter due to old pump leaking, and replaced a broken oil dip stick tube. I purchased the car 2 months and 2k miles later. Obviously the start issue wasn’t resolved.

If the car has been run recently (even if not all the way up to normal ~90c operating temperature) the engine will still start much more quickly; Less than 1 second to 3 seconds of cranking if ECT is disconnected, and 1-6 seconds (?) if ECT is connected.

I haven't tested extensively to determine after what time frame it no longer starts up quicker, but I do know the effect lasts for a while, like an hour or two at least, and the engine starts better the more recently the engine was last run. So I’m thinking issue is possibly either engine temperature related, or something to do with some fluids getting moved around that stick around for a while enabling an easier start, or some sensor reading being correct after the engine just ran, but these are just blind conjecture at this point.

Once started, with ECT either connected or disconnected the car runs and drives fine, no idle issues, no power issues, no smoke issues, etc.

I've been through the TDI FAQ slow start troubleshooting page, and a couple of forum posts and at this point could use a pointer in the right direction...

Here's what I've checked and done so far:
Replaced ECT - ECT was disconnected by previous owner to start car easier, and was likely also bad (dash gauge was non-working even after I plugged the connector on again). I replaced with new ECT sensor, dash gauge now works, and VAG-COM readings look good, both ambient and at running temperature.
Replaced engine air filter – Old one was dirty, replaced with new Mann filter.
From the TDI FAQ page section on ‘engine is difficult to start, requires a long period of cranking before starting’:
Weak battery - Old one was pretty good, replaced with new OEM sized one anyway due to upcoming winter.
Corroded electrical connections - Battery terminals look very good, I haven't checked ground straps, but starter cranks nicely at 250-300rpm, and car starts quickly when recently run, so I don't think this is an issue.
Air in fuel system – Replaced fuel filter, checked and replaced the o-rings on T fitting, vacuum pulled fuel through injection pump, loosened injector fittings and verified fuel was running through them before tightening. Clear tube shows 1/2" to 1" bubble when engine off, but I have read on forum that this is normal. No bubble(s) visible while engine running. Checked and tightened all connections. I see no leaks on either supply or pressure side.
Fuel filter clogged or filled with water - Fuel filter was 6 months and 2k miles old, replaced with new VW filter as troubleshooting step anyway.
Incorrect fuel injection timing - Checked with VAG-COM, timing was advanced out of envelope after timing belt service with rough ~800rpm idle speed (ran fine, just a lot of vibration, like a large displacement two cylinder motorcycle engine). I adjusted the timing to be midway between the middle and upper line (at just under the top line the idle was still a bit rough) of VAG-COM timing graph. Idle is now smoother at ~900rpm, and the engine now has power to 3500rpm instead of 3000rpm. No difference in starting.
Anti-shudder valve problems - Checked that valve moves freely and is spring loaded to be in open position.
Poor quality fuel - I have only filled up twice at the same name brand station, but the previous owner had the same issue multiple states away, so not likely.
Cold temperature - Not an issue, ambient temp has been between 60-90F (16-33C), VAG-COM reports normal fuel temperature.

I've checked the codes with VAG-COM, there were a few in there that I saved to log and then cleared when I first hooked it up. There is an ECT sensor intermittent short code that comes up when I disconnect the ECT. If I leave the ECT connected and drive/use the car with it connected I do not believe I have seen it come up. No other codes, other than one for my remote key fob.

I've checked the other group measurements I thought/knew to look at and I believe all measurements were within spec, but can check anything again and provide measurements/screenshots/logs if desired.

Coolant and oil levels are good. The only thing I haven’t done yet that I was going to anyway due to it being a new to me used car is to change the oil and oil filter.

I’m now to the point where I do not know enough about the car to be able to continue intelligently troubleshooting. Help appreciated!

Check for paint marks on the timing belt wheels. If you find some, you should at least check the basic cam timing. Try to ascertain if all the bearings (water pump, idlers, tensioner) were replaced during the last service. If you can't find positive evidence of proper service, I recommend a timing belt re-do.

With the ECT connected, read group 007. It should match all the other sender values as well as the approximate outside temperature. Obviously, the car has to have been sitting for a while.

Begin logging group 001 with the ignition on, engine not running. Start the engine. Save the log. Check the RPM value (~250 new, about 160 is the lower limit) and the IQ value. It should be about 25 mg/H (it will increase to nearly 40 mg/H during extended cranking). Observe for excessive white smoke on starting.

With the engine warmed up to 80-90C, turn off all electrical loads, and read the IQ value from group 001.

Thanks for the reply. The service invoice lists parts of toothbelt, water pump, tensioner, three rollers, ribbed belt. They all look pretty new, and I have no reason to think they didn't do the service. Whether they adjusted everything appropriately or simply copied the exact positioning of the old belt, I do not know.

I'm not sure I know what you're talking about with regards to checking cam timing (where/how?), I've only skimmed the timing belt replacement guide so I'm not super familiar, and I thought to change the actual cam timing you have to remove the head and adjust the positioning of the belt with the cam nobs 1 and 2 locked pointing up in the proper position, or something, ha. There are a few white paint marks on the edges of the sprocket that is attached to the cam. There are yellow paint marks across the bolts related to the new injection pump, the sprocket bolt, and like 3 bolts on the top of the pump.

This http://imgur.com/NuXu1.png is one long image of screen shots of the VAG-COM readings. The timing graph on 2010-09-26 the ECT must have been disconnected, looks like the rest of them it was connected. The set from today (2010-10-08) the ECT was always connected. The cold temperature readings you requested are accurate +/-1 degree C of the actual ambient temperature. After connecting the ECT I cleared the error code related to it having been disconnected, and none reappeared.

The cold start with the ECT connected there was not an excessive amount of white smoke, some for sure, not any still being produced after initial puff. The idle seemed more like it was, very noticable two cylinder engine like vibration. Pretty buttery smooth at like 1,300rpm, but I don't think I'll set it to that, ha. Here is the log of the cold start, and two warm starts using group 001:
I guess the tach was lying to me about cranking at 250-300rpm, appears the starter motor isn't exactly the speediest thing, but as you can see, the warm start is pretty darn quick.

Thanks again for helping me out!

The starter looks borderline, but if the group 13 shot was taken with 15 and 12 at the same time, cylinder 2 looks a little off.

For some reason, cyl. 2 is producing relatively more power than the others. 1,3,4 look well balanced.

Is the car somewhat sluggish?

I'd start with the starter - I got a Bosch reman for $140 at AutoZone. I don't know how slowly mine was turning when it went bad, but it was noticeably slower than normal.

Second, the #2 injector may be leaking or streaming. Given the time on the unit, at least a spray test is indicated. There's almost a 100% chance you'll need new nozzles.

Yes, screenshot of group 12, 13, and 15 were all at the same time. Other than the nozzle(s) being old and likely needing replacement, any other things I could/should look at in regards to cylinder 2? I presume it's far better that it's making more power rather than less, which could indicate something like damaged piston rings or cylinder walls or damaged or really dirty valves, etc., yes? Or are there some equally troubling problems that could be indicated by more power in a single cylinder?

I don't think it's sluggish, I mean for what it is. My other car has 290hp and 300ft/lbs torque and does 0-60 in under 6 seconds, so a car that takes 12 seconds for 0-60 is going to seem rather slow to me. But it builds power up from about 1,900rpm, pulls strong between 2,200 and 3,600rpm and then drops off suddenly, which is what I'd expect of its power curve. Other than cold starting and the issue you pointed out with cylinder #2, there doesn't seem to be anything wrong with it.

I'll check the connections and cables first before replacing the starter, but I'll put it on the possible to replace list.

I'm just concerned that I could drop $150-$200 on a starter, and $50-$300 on nozzles (used stock-bosio PP), and still be no closer to an answer.

Should the diesel just be that much more difficult to start cold than warm? I mean it starts pretty much instantly when it's warm, even with the marginal starter... And why does it start so much easier with the ECT disconnected, that one concerns me the most... 3rd question is what's the deal with the really strong vibration at idle, I mean it's like you were sitting on a Harley, except, you know, in a chair behind it, ha. Only comments on that are that it seemed much smoother after adjusting the injection timing but now seems mostly back to the way it was; And that just playing with the throttle at idle the engine was very smooth (for a 4 cylinder diesel anyway) at somewhere around 1,300rpm, if that helps at all in pointing to or eliminating possibilities. I'll have to get an answer or go research possible causes for the idle vibration, not the end of the world, but it's quite noticeable.

It starts a little better with the ECT removed due to the long preglow and initially higher injection quantity. Normally, the engine will start immediately hot or cold.

Is yours automatic or 5 speed?

5 speed.

So you're thinking pretty simply then that heated chamber + more fuel = faster start? If that's the case, for what reason would it be failing to start with a cold chamber and a normal amount of fuel? You know, as it should be doing. Thinking that slow rotation of engine isn't heating enough through fast compression, and perhaps in combination with worn/dirty injectors that have poor ineffectual spray patterns...?

Slow starting speed - check the battery. What's the date on it? Is it fully charged?

Can you measure state of charge using the open circuit voltage and electrolyte specific gravity?

Bad nozzle performance can also have an impact. They have to have very poor flow rates for it to be a major problem.

I'd have to see what the old battery date was, it was fully charged, and was only marginally increased in turn over speed by jumping it with another battery. The current battery that I just put in is brand new, manufactured September 2010, regular OEM 94R size, and also fully charged. Again, only a minor difference between it and the old battery. It would be cables/connection or starter motor at this point for the slow turn over speed.

I can get you a baseline for the starter to battery cable resistance using a clamp ammeter and reading the DC voltage drop across the cable. I can see about checking the windings, but I don't have a high current industrial ohmmeter.

I'm not certain of the steps you're discussing, you're primarily interested in testing the resistance of both the cable / cable + connectors, and additionally the windings of the starter motor, correct? I don't see how a clamp meter would be used for that, that's used to detect the amount of current flow, yes? I have one that works on AC, doubt it is capable of DC. And otherwise just have regular multi-meter.

I use the second meter to estimate the cable current, so I can approximate the resistance with the voltage drop information.

I think I get you, you're reading the voltage/voltage drop of the battery, while also measuring the current flowing to the starter, if you're missing current (which you will be) you're doing some math to figure out how much resistance there is, rather than measuring it directly with an ohmmeter, yes? If I'm right, I'm treating myself to an ice cream cone. Because I shouldn't be thinking this hard at 1am on 3.5 hrs sleep and an empty stomach

Anyway, short of that equipment and math, can you not simply red-neck it and hook up some cables to the starter directly and see what kind of RPMs are generated?

mkr said:

Anyway, short of that equipment and math, can you not simply red-neck it and hook up some cables to the starter directly and see what kind of RPMs are generated?

Click to expand...

Yes, if you have a large enough cable.