It started after a routine maintenance operation. At 105,000 kilometres, the cam belt of the VW 1.9 litre PD turbo diesel in my Skoda Roomster needed to be replaced. I gave the car to a specialist workshop and asked that the water pump, coolant and brake fluid also be changed - the sort of things it seems reasonable to do about every 100,000 kays.
The bill was relatively painless (although this sort of service will never be cheap!) and the job was done in about half a day. I paid my cash, and drove out the door.
But the trouble is, the car didn’t seem to drive the same.
After the new intake, new rear muffler, upgrade intercooler and ECU reflash (see series starting at Powering-Up the 1.9 litre TDI, Part 1), the Roomster had, I thought, been very good to drive. Tractability extended down to idle revs (you could drive around in fifth gear at literally 1000 rpm), power came on strongly at 2000 rpm, and the engine revved freely to its 4500 rpm redline. In city driving I typically short-shifted at about 2500 rpm, making use of the extra bottom-end torque.
But now, after that cam belt change, the car seemed to have lost its bottom-end urge. Instead of being strong at 2000 rpm, the engine started to really work only at about 3000 rpm. In fact, hold the accelerator at (say) 50 per cent, and the car could be felt to suddenly pick-up as 3000 rpm was passed.
And after the cam belt change, fuel economy was worse. On the ~80km country drive home from the workshop, economy came in at 6.5 litres/100km – previously, I was getting about 5.7 on the same route.
So what was wrong? The most obvious answer seemed to be that the cam timing was now incorrect – perhaps the workshop had got the cogs one tooth out? I rang the workshop and they were bemused: if the timing was a tooth out, they said, the car would barely start and, in their words, “you’d sure know about it”.
“Well, I do know about it,” I replied. “The car now drives differently.”
We arranged a time for the car to be returned to the workshop for the cam timing to be checked. The earliest I could make the car available was two weeks away, so I had plenty of time to feel how the car behaved.
What I additionally found was that the car went like a cut snake at the top end – from 3000 to past the 4500 redline, the car was plenty strong. In fact, it didn’t feel like a diesel at all; instead, it felt like a pretty well-tuned 2 litre naturally aspirated petrol engine with a 7000 rpm redline!
But fuel economy (important to me as I do a longish country commute) was clearly way poorer than before. On my country roads I also found that I had to downshift one or even two gears to get the overtaking performance I could previously get in fifth or fourth gears.
Cam timing check
So back the car went to the workshop. And what did they find? Well, the came timing was ‘spot on’, they said.
And where did that leave me? Pretty dissatisfied!
By this time I’d done some research and found that, intriguingly, the factory provides the facility for small adjustments to be made in cam timing. This facility takes the form of slotted holes on the camshaft belt pulley: loosen the three bolts, rotate the camshaft and then tighten them up again.
Even better, to allow easy access, the top part of the cambelt cover is removable by simply undoing two spring clips.
But let’s take a step back: why would this engine be so susceptible to tiny variations in cam timing?
In the Volkswagen PD (Pumpe Duse) engines, common rail diesel injection is not used. So rather than a single mechanical high pressure pump and separate, electronically-controlled injectors, the PD system uses what are called ‘pump/injectors’. As the name suggests, these combine the functions of pressurising pump and electronic injector. One of these pump/injector units is provided for each cylinder. (Click on any of the images to enlarge them.)
Extra lobes on the camshaft act through roller rockers on these pump/injectors – in the case of the 1.9 litre engine, there are four extra lobes. In order that fuel pressure is developed rapidly, the lobes have very steep leading ramps.
Now if the cam timing is altered, it appears that, even with no change in the electronic action of the injectors, the amount of fuel that is actually injected may be altered. That’s because it’s the combination of the pressure developed by the pump injector and its electronic actuation that will influence the amount of fuel that’s injected (and possibly also exactly when it is injected).
So when cam timing in these engines is altered:
• as a result of the difference in intake and exhaust valve timing, the volume of air that is flowing through the engine may be changed
• as a result of the difference in pump pressurising, the timing and amount of fuelling may be changed
Diesels are extremely sensitive in performance to the amount of fuel being injected: pour it in and the engines go harder, take it out and performance suffers. (There’s more to it of course than just this, but compared to petrol engines, it’s much more the mass of injected fuel that governs output.)
Playing with cam timing
So after mulling over all this, I decided to have a go at altering the cam timing. It seemed to me that if the cam belt was on the right teeth, the adjustment offered by the factory via the slotted holes was very unlikely to result in pistons hitting valves.
So I took off the top cam belt cover and had a look.
What I could immediately see was that the adjustment was right at one end of the slotted adjustment range: the cam was advanced. Now if the cam belt had previously been stretching, and if the car had been serviced by a very good mechanic, then perhaps the cam belt timing had been advanced to compensate for belt stretch that had occurred during the 80,000km the car had done before I bought it? If this theory was correct, when a new belt went on, the cam would be too advanced.
The theory seemed reasonable, so I marked the existing cam position with a pencil, then loosened the adjustment bolts and retarded the cam timing a bit. How much? I don’t know: nothing was being measured.
The car started perfectly, so I went for a test drive. Strong power was now clearly coming in at slightly lower revs – say 2800 rpm rather than 3000.
I retarded it a little more, and got the rush of power down to 2500 rpm.
A little more retard, and power came on at 2300, then with more adjustment, 2000 rpm.
Then I went too far and the engine wouldn’t start (ie it cranked fine but wouldn’t fire), so I advanced cam timing a fraction from that point.
Everything seemed fine: the car drove pretty much as it did before the cam belt change, and the slotted adjustment was very close to the middle of the range.
But the next (cold) morning, the car wouldn’t start. Again it would crank fine but not fire. This time, working from dead cold, I advanced the cam timing just far enough that the car started fine.
On the road
I was aiming to bring performance (and fuel economy) back to the starting point: that is, the results achieved after the intake, exhaust, intercooling and custom dyno reflash modifications. But it appears that by very carefully adjusting the cam timing within the ‘slotted holes’ range, the results are actually better than previously achieved.
Performance is very close to what previously occurred: I can short-shift around town, using only moderate accelerator pedal openings to keep up with traffic. I can also overtake in fourth and fifth gears. The car ‘comes on turbo’ at about 2000 - 2200 rpm, but can still be easily driven in fifth gear at 1000 rpm.
However, fuel economy is clearer better than previously. Fuel economy comparisons are hard: when the car was first modified, it was summer – and this diesel gets better fuel economy in summer than winter (where I live, summer has typical maxima around 30 – 35 degrees C; winter has maxima more like 10 – 15 degrees C).
On my country commute, summer fuel economy was typically about 5.2 – 5.3 litres / 100km, while winter fuel economy was about 5.6 – 5.7 litres / 100km. After the fine tuning of the cam timing, I am achieving in winter the fuel economy I achieved in summer – ie, about 7 per cent better than before.
Downsides? Yes, there are a few.
In cold conditions (around 0 to minus 5 degrees C), the car does not start as well. It still start, but it isn’t as absolutely seamless as the standard car – there’s a hesitancy and a little stumble after the engine first fires. When the engine is dead cold, the torque just off idle is also clearly lower: the clutch needs to be slipped a fraction more when moving away from a standstill. When the engine warms up, this issue disappears.
If you like to play and you have a PD diesel Volkswagen, it certainly appears that trialling slightly different cam timing settings has the potential to make quite major differences to:
• where the power is developed in the rev range
• the fuel economy
• the starting behaviour
In all these aspects the ‘PD’ design appears to be far more susceptible to cam timing variations than common rail diesels and petrol engines – so if you’re after an optimal outcome, consider making some very small adjustments!