PD Engine Torsion Value - Solid Information About Why and How To Optimize?

bmwman91

Member
Joined
Sep 17, 2019
Location
CA, USA
TDI
2006 Jetta (BRM) w/ DSG
While there are countless threads discussing the setting of the Torsion Value on PD engines, I have yet to find a good technical description of why it even needs to be messed with (versus zeroing it) and why it is a thing at all. Granted, I mainly come from the world of engines with timing CHAINS which have their own quirks, but basically if everything is locked to TDC, there's no mystery. Even every engine I have encountered with a belt-driven valvetrain is about the same, with one simply aligning timing marks. Yes, the PD engines differ in that the cam also drives the high pressure injector pumps, but it is well accepted that the Torsion Value tweaking has zero to do with the fuel injection, and it is all about timing of the valves.

So, why exactly is it that every PD engine shipped from the factory with varying torsion values scribbled onto the label on the belt cover? If VW had thought it best to set it to zero in all cases, I assume that they would have since that is basically what every other car maker seems to do. For instance, mine has "-1|62" written on it, with the bar not being quite that tall, but pretty long for a comma. I presume that it means that the cam was retarded 1.62° from the factory. At this point, the car has had 2 belt changes, as well as a new cam and lifters, so I wonder what the best value is now. The previous owner had just changed all of that before I bought it from him, and I have only had it for about 6 months. This gets to why I am posting this at all. The torsion value in VCDS reads -5.5° at idle, which is pretty far off of zero, but not out of the (wide) range of values that people in various forums have set it to and claimed improvements in power or fuel economy. Anyway, should I just set it back to the original value of 1.62°? The car smokes a little at startup, feels like it runs out of juice above 3000RPM and only gets ~39.5MPG (measured from odometer and gallons pumped when filling) when I did a few hundred miles on the highway with cruise control locked to 65MPH. Around town it gets 32-34MPG, and I feel like this is a little low. The car is a 2006 Jetta with DSG and 209K miles.

So, if anyone has thoughts on my explicit case above, I am all ears. Beyond that, I'd like to see if anyone knows the true technical basis for the big variance in torsion values that these cars shipped with. From a manufacturing standpoint, the pistons, rings, camshafts and valves are all going to be within a couple of thousandths or better as they come off the manufacturing line, so I do not see any way that variance in those bits would be the reason. The timing belt seems like the only item that could have anywhere near the kind of variability that might lead to each car having a unique torsion value optimum. Some may be a little stretchier than others, or have different thermal expansion characteristics, different tooth wear-in etc. Other vehicles with belts don't seem to worry about it, and I cannot imagine that the belt in our cars is super special, except for the additional tension in it from driving valves AND injector pumps. Other than my guessing, has anything ever been published that goes into detail about why torsion value (which I am calling "valve timing" in my mind) needs to be adjusted, and how one is supposed to determine the optimal value for their engine? Did VW get a dial indicator on a lifter and set the cam to a specific position at TDC (or some other angle) when they built these things? That's about the only way that I can think of that they would have been able to set the cam timing to an optimal value fast enough to be in a mass production environment.

Thanks folks! With any luck, maybe we can put this one to rest for good.
 
Last edited:

Tdijarhead

Top Post Dawg
Joined
Nov 10, 2013
Location
Lawrenceville PA
TDI
2003 TDI Jetta Daughters Car, 2001 TDI Beetle, Wife’s car, 2005 Golf TDI Mine, all 5 spds

oilhammer

Certified Volkswagen Nut & Vendor
Joined
Dec 11, 2001
Location
outside St Louis, MO
TDI
There are just too many to list....
PD engines' torsion value is not unique. Most modern engines employ some form of cam-crank relation zero point calibration. GM was doing it years (decades) ago with their pushrod V6s and V8s (4.3L, 5.0L, 5.7L, 7.4L). You adjusted it by looking at a value in the scan tool during start up, and turning the "distributor" and locking it down. It often took five or six tries. It was their way of accounting for chain stretch, and if it was off, the engine would not run as well as it could/should, and in extreme cases would set DTCs.

VAG, Ford, Honda, GM Toyota, and Hyundai all have values now that are learned within the cam-crank correlation, VAG even displays this as "chain stretch" in the EA888 engines. This is how they can work the vario-cams they most all employ. The ECU needs a fixed, solid, zero point basis to start from. All of the engine's timing control, be it for cam phasing, fuel injection, ignition timing (on gassers, obviously), and even variable intake phasing, all have to take into account this data.

Getting back to the PDs: my experience has shown me that BEWs and BHWs seem to be happiest at around 0.00, while BRMs seem to be happiest around +1.00. But that is just an average, because it also tends to be a seat of the pants thing as well as an individual preference. To me, I would rather have the best performance, even if it is at the cost of a little fuel economy. All TDIs get stellar fuel economy. But a ~3000 pound car with only 100hp could use an optimized for power tweak whenever possible.

You can actually go in and watch the fuel quantity and play with the torsion setting and sometimes find a sweet spot where the quantity stays the same as you change the torsion. This is a super granular way of tweaking it, and you may or may not even notice a difference in your daily driving.

You may also want to read through my sticky in the Fuel Economy section, as I would bet your 14 year old 200k+ mile car has a few things that could be addressed to squeeze a little more MPG out of it, if that is what you are after.
 
Last edited:

bmwman91

Member
Joined
Sep 17, 2019
Location
CA, USA
TDI
2006 Jetta (BRM) w/ DSG
You’ve probably seen this thread Franko6 is the guy who is the PD expert.

http://forums.tdiclub.com/showthread.php?t=336279&highlight=problem+setting+torsion

When my torsion was -4.9 if I recall I had 5 of 8 worn cam lobes. I had originally set the torsion to +1.5 it seemed to have moved on its own as the cam wore.

You may want to check your cam.
Thanks for the link, it seems to totally explain the whole situation. Wide tolerances in the hole for the locking pin mean that one is not likely to be locking the crank at true TDC, hence the need to adjust the cam. So, the physical reality is that a "proper" torsion value setting means that the cam and crank are perfectly set for 0deg TDC relative to one another (well, the crank has piston 1 at TDC and the cam lobes are at their ideal positions for that condition). Short of Frank's suggestion to pull injector #1 and get a dial indicator onto the piston crown, I can now see why one has to experiment with this.

So, assuming the 1.62° mark on the cover was correct, and the previous owner never unbolted the crank sprocket, I assume that a torsion value of -1.62° is what I want?

When I first bought the car I pulled the valve cover and checked all cam lobes and lifters. Everything was fresh, with no scuffing on lifters and all of the lobe chamfers intact. The PO had said that he had replaced it all, and it appeared to be true.

PD engines' torsion value is not unique. Most modern engines employ some form of cam-crank relation zero point calibration. GM was doing it years (decades) ago with their pushrod V6s and V8s (4.3L, 5.0L, 5.7L, 7.4L). You adjusted it by looking at a value in the scan tool during start up, and turning the "distributor" and locking it down. It often took five or six tries. It was their way of accounting for chain stretch, and if it was off, the engine would not run as well as it could/should, and in extreme cases would set DTCs.

VAG, Ford, Honda, GM Toyota, and Hyundai all have values now that are learned within the cam-crank correlation, VAG even displays this as "chain stretch" in the EA888 engines. This is how they can work the vario-cams they most all employ. The ECU needs a fixed, solid, zero point basis to start from. All of the engine's timing control, be it for cam phasing, fuel injection, ignition timing (on gassers, obviously), and even variable intake phasing, all have to take into account this data.

Getting back to the PDs: my experience has shown me that BEWs and BHWs seem to be happiest at around 0.00, while BRMs seem to be happiest around +1.00. But that is just an average, because it also tends to be a seat of the pants thing as well as an individual preference. To me, I would rather have the best performance, even if it is at the cost of a little fuel economy. All TDIs get stellar fuel economy. But a ~3000 pound car with only 100hp could use an optimized for power tweak whenever possible.

You can actually go in and watch the fuel quantity and play with the torsion setting and sometimes find a sweet spot where the quantity stays the same as you change the torsion. This is a super granular way of tweaking it, and you may or may not even notice a difference in your daily driving.

You may also want to read through my sticky in the Fuel Economy section, as I would bet your 14 year old 200k+ mile car has a few things that could be addressed to squeeze a little more MPG out of it, if that is what you are after.
Good info, thanks. I guess I have not wrenched on quite enough other cars to run into this. I fully get why one needs to double or triple check that everything is in proper phase after removing the locking tools and rotating the engine a few times since the tensioners and such will pull things into different positions (same deal with the timing chain in my old BMW). With all of the other cars where you have to do trial and error to get a value in software to look right, is it also the case that one is simply trying to get the cams to their TDC lift position relative to the crank when at TDC? So, in theory, lie with the BRM if one had a dial indicator on piston 1 to make it sit perfectly at TDC, they could then set the cam perfectly with the alignment marks as well? The valve cover and injectors are pretty easy to remove (already replaced the VC gasket and injector loom), so I am tempted to see about using the dial indicator to verify what my current TDC torsion value will be. Maybe I will first try setting it to the value scribbled on the belt cover since that only takes a few minutes...

I'll check out the MPG FAQ as well. Thanks again.
 

Mozambiquer

Vendor , w/Business number
Joined
Mar 21, 2015
Location
Versailles Missouri
TDI
2004 VW Touareg V10 TDI, 2012 Audi Q7 V6 TDI, 1998 VW Jetta TDI. 1982 VW Rabbit pickup, 2001 VW Jetta TDI, 2005 VW Passat wagon TDI X3, 2001 VW golf TDI, 1980 VW rabbit pickup,
I know that my bhw sure started better when I adjusted it to 0 degrees! It was at +5 degrees, and it didn't start hard, but it started a lot better after changing the torsion setting.

Sent from my S60 using Tapatalk
 

Tdijarhead

Top Post Dawg
Joined
Nov 10, 2013
Location
Lawrenceville PA
TDI
2003 TDI Jetta Daughters Car, 2001 TDI Beetle, Wife’s car, 2005 Golf TDI Mine, all 5 spds
I’ve not heard of a 1.62 value, 1.6 or 1.5 but I’ve just not seen the value carried two decimal point to the right before. I would do some experimenting at 0.0 and 1.5 or 1.6 and see how it starts and runs.

You know where it’s set now so you can always put it back if you don’t like the way it.
 

bmwman91

Member
Joined
Sep 17, 2019
Location
CA, USA
TDI
2006 Jetta (BRM) w/ DSG
So with your note that they never specified 2 decimal places (which makes sense considering how sensitive the cam position is and realistic tolerances), I did my best to carefully clean the label on the belt cover and get a better look. The thing was pretty filthy and hard to read. As it turns out, there are two hand-written values on there, with the "new" one making a lot more sense for a torsion value. Here's the best shot I could get in the dark.



I assume that the "0.5" is the original torsion value, and what I should try to set things to in my pursuit of better operation. The info I've seen now for the BRM seems to indicate that torsion values of 0-1° are usually best, so I guess this makes sense.
 
Last edited:

bmwman91

Member
Joined
Sep 17, 2019
Location
CA, USA
TDI
2006 Jetta (BRM) w/ DSG
OK, success! I got out there and made the adjustments, and while I have no idea what it will do with fuel economy, the car actually runs WELL now. It was pretty gutless before, which I just attributed to it being a ~100HP engine and me being used to a ~200HP 2600lb E30 BMW. The TDI is not "fast" now, but I think that I an actually merge in the way that I usually like to without needing a 300ft runway. Overall it is a lot more pleasant to drive, and it starts a lot faster too.

Here's how I approached it this afternoon. First, I drove it around to get it fully heated up. After getting the boost pipe and belt cover out, I noted that the cam was retarded in the slots at almost the maximum possible amount.




Looking in VCDS, the torsion value was stuck at zero until the engine was warmed up, at which point it read -5.9 at idle and -6.4 when driving. So basically, it on or a little past the range where it could be measured at all.




So, given that it was a good 6~7 degrees retarded, and the cam:crank speed ratio is 1:2, I figured I would need to advance the cam at least 3 degrees (which is 6 degrees on the crank, and I assume torsion value is measured in crank degrees). I have a nice digital level/protractor, so I held it to the wrench and zero'ed it to start.




I then advanced the camshaft ~3 degrees, which got the three 13mm bolt heads nearly centered in their slots.






Back in VCDS, things were improved, but still a bit retarded relative to where I think they needed to be. Still, the digital level really helped to ensure that I didn't grossly overshoot. 3 degrees is not a lot, as far as your eye and a breaker bar are concerned.




Thereafter, I advanced it another degree, which got me slightly past where I thought I wanted to be (0.5 is indicated on the belt cover).






I made a final tweak of ~0.25 degrees in the other direction with the level, and that got me right where I wanted to be. It was reading 0.5 at idle, and 0.0 when driving. I probably missed it somewhere, but is the torsion value on the belt cover supposed to be the value at idle, or some other operating condition?




From there, I put everything back together and drove around for 15 minutes. The difference in response and power was very noticeable, and while I hope that fuel economy will improve, the car is vastly more pleasant to drive. I'll just have to see how my economy calculations look over the next couple of tanks of fuel. Anyway, after the drive I re-checked the idle values and they seemed to hold, so I considered this little project done for the day.




My idle fuel consumption went up a little versus before, but most info I have read indicates that minimizing idle fuel use does not translate into minimizing driving fuel use. We'll see. How do my Injection Quantity Deviation values look?
 
Last edited:

bmwman91

Member
Joined
Sep 17, 2019
Location
CA, USA
TDI
2006 Jetta (BRM) w/ DSG
Oh, and lastly how do my EGR and boost values look? I logged them while driving, and here is what things looked like with a mix of WOT, deceleration and cruising. It seems like the EGR & MAF actual values are considerably higher than the EGR spec value during decel...is this normal, or a sticky valve? Also, actual boost tracks the spec pretty well, but seems to chronically read high at low boost values. Could the VNT actuator be a little sticky at the low end, or is this also normal?
Thanks again!
 

bmwman91

Member
Joined
Sep 17, 2019
Location
CA, USA
TDI
2006 Jetta (BRM) w/ DSG
Here's a small follow-up. Since fixing the cam timing / torsion value to be at the factory spec, not only is the car a LOT more lively to drive, but I am seeing a 5-10% improvement in fuel economy (measured with the odometer and fuel pump). Additionally, the econometer reading was previously 10-20% optimistic, and it now only reads 2-3% high. Although lots of people across forums report all sorts of "benefits" to purposely taking the torsion value to non-factory values, at least in my case going back to stock has made numerous improvements all around.
 
Top