TDI Timing Revisited

aNUT

Vendor , w/Business number
Joined
Nov 29, 2006
Location
Boulder, Colorado
TDI
'01 TT (ALH-ish), B7 Audi gasser, '05 Golf
Sometimes it's good to have a refresher. I frequently get questions about optimizing mileage, or see threads about reduction in mileage after a timing belt change. This post addresses timing subtleties on VE cars, '96-'03 in North America.

Firstly, here's a screen by screen, step by step, pictorial on checking static timing with VCDS:

Connect the Vagcom to the OBD2 port located
a) above your left shin if you have an A4
b) behind the windshield wiper switch if you have a B4
c) next to the ash tray if you have an A3

Open VCDS.

Click "select"

Click "engine"

Wait for VCDS to establish a connection with the ECU.

Once in a blue moon, VCDS won't do this with the car running. If this is the case, turn the car off, then the ignition to on, (but not start) make the engine controller connection, start the car, and continue.

Click "measuring blocks"



Select group 000. Click "go". Click "switch to basic settings".



Click "TDI Timing"



The following graph will appear.



From the drop down menu in the lower right corner, select your engine.



The apex of the horizontal and vertical yellow lines is where the timing is set. If there are no yellow lines, the timing is very retarded. If there is only a vertical yellow line, the timing is very advanced. Sometimes on A3s and B4s VCDS gets confused when the timing is very retarded, and plots it as very advanced. Advance the timing, and the graph will 'come around.'

Adjusting the timing will not be covered in this how to. Please refer to the timing belt procedure for your car regarding how to do so. A3/B4 A4

Note that timing should be checked with the vehicle at operating temperature. The timing graph is graded by fuel temp, but advance rises slightly (~.5*) after the motor warms up.

OK, so that's the basic stuff out of the way. Before you're bored and move onto the next thread on oil banter, here's where the subtleties come in. Technically, with the apex between the red and green diagonal lines, timing is within spec. TDIclub dogma dictates that the driver wants the timing as close to the green as possible. The following photo is often referenced. The reason people claim this is optimal is to reduce EGTs. This is not true. More on that later.



Let's have a look at measuring blocks -> group 004. This shows actual and specified start of ignition.



Note that the actual start of injection cannot match specified with advanced timing. The Advance Solenoid/Cold Start Valve/N108 duty cycle is pegged at 2.8%. That means the minimum amount of advance is 2.6*. So what you say? Here's the rub. At light load, say 70mph on level ground, where one will get the best milage, specified advance will be .5*-1.5*BTDC. If the ECU cannot achieve requested (optimal) advance, economy suffers to the tune of ~5%. For us TDI nerds, 2-4 mpg matters.

So here's how I time cars:

I put the timing in the graph, then I look at group 004. I set the timing as advanced as I can get it, but retarded enough that actual can meet specified start of injection at idle.



And here's the graph with optimal timing:



Notice that I'm not even above the blue! On a cold day with not so stellar fuel in the tank, that apex can be a dotted white line or so above the red. This is very counter to the theory that's been propogated on these forums...but the car is more efficient.

Now, about that whole EGT thing. Shorter injection duration and earlier start of injection does reduce EGTs, but setting the static timing advanced doesn't achieve this. Higher injection pressures coupled with tuning that's optimized for the fueling hardware on the car does, but not advancing the timing.

Below is a log taken from my car at WOT and stratospheric (for a diesel) engine speed. These are the conditions that require maximum advance I've cropped off the speed, but this section of the log was taken between 3600 and 4600 RPM. It's clear that the car is only requesting a bit more the 13* advance, and the car can easily achieve over 15*. The Duty cycle (also not shown) is between 70-85%, so more can be had if necessary. Essentially, setting the timing for optimal economy does not sacrifice advance during high load situations as once thought.



The above graph was generated using JollyGreenGiant's superb plotting software.
 
Last edited:

runonbeer

Maintenance EnthusiastVendor
Joined
Apr 15, 2002
Location
Austin, TX/Chapel Hill, NC
TDI
'00 Golf 02M, '10 Golf 02E, '02 UTE 02M
Below is a log taken from my car at WOT and stratospheric (for a diesel) engine speed. These are the conditions that require maximum advance I've cropped off the speed, but this section of the log was taken between 3600 and 4600 RPM.
below what? below where?

Its nice to see that you're putting that physics degree to good use aNUT.
 

Growler

Got Soot Vendor
Joined
Nov 24, 2003
Location
Millersport, Ohio
TDI
Schmutz, 2015 Golf Sportwagen DSG & Schnurren, 2001 Golf GL 2 door 5M
5 stars for you Anut!

I didnt know about the specified & actual start.. I will have to check that on my car now...
I assume this is done with the car running, and if its not right like in your first picture, adjustment at the pump bolts is needed? (after turning the car off of course)
 

ymz

Top Post Dawg
Joined
May 12, 2003
Location
Between Toronto & Montreal
TDI
2003 Jetta TDI Wagon, 2003 Jetta TDI Wagon
Nice contribution !!

I thought that the reason folks wanted the timing set on the advanced side is to help with (extreme) cold starting - before the ECU takes over the compensation of the timing...

Is this not a consideration, especially for us folks who live in the Great Frozen North ???

Yuri.
 

greengeeker

Vendor
Joined
Feb 8, 2006
Location
Cambridge, MN
TDI
2002 Jetta GLS
Thanks for the info aNUT.

Building on what yuri's comment...do you not see any value in advancing the timing towards the top end? If there still is a benefit in your eyes to advancing the timing, is there a good compromise between advancing it and allowing it to match the specified and actual?

I look forward to seeing more discussion on this.
 

whitedog

Veteran Member
Joined
Jul 12, 2004
Location
Bend, Oregon
TDI
2004 Jetta that I fill by myself
How about using adaptation to get the spec to match the actual once the basic setting is set at the top of the graph?
 

Keith_J

Veteran Member
Joined
Apr 12, 2007
Location
West
TDI
2000 Jetta MT
Wow, the timing is really all over the map. I take it that is because the rotating ring doesn't exactly move all that fast?
 

aNUT

Vendor , w/Business number
Joined
Nov 29, 2006
Location
Boulder, Colorado
TDI
'01 TT (ALH-ish), B7 Audi gasser, '05 Golf
runonbeer said:
below what? below where?
Below the last paragraph; though I forgot to post the final screen shot. hehe. Thanks for proofreading

Growler said:
I assume this is done with the car running, and if its not right like in your first picture, adjustment at the pump bolts is needed? (after turning the car off of course)
That is correct. Engine warm and running. Standard timing adjustment rules apply. Twist the pump (A3/B4) or loosen the sprocket and rotate the hub for an ALH.

ymz said:
I thought that the reason folks wanted the timing set on the advanced side is to help with (extreme) cold starting - before the ECU takes over the compensation of the timing...

Is this not a consideration, especially for us folks who live in the Great Frozen North ???

Yuri.
Definitely agree with you there. Starting conditions are N108 at 100% duty cycle; full advance. I can't claim great white north climate; but it does get pretty cold here in the winter; and the altitude doesn't help. Less air = less heat on the compression stroke. I haven't had a problem in that regard. Coldest i've started here has been -20F.

greengeeker said:
Building on what yuri's comment...do you not see any value in advancing the timing towards the top end? If there still is a benefit in your eyes to advancing the timing, is there a good compromise between advancing it and allowing it to match the specified and actual?
I think the best compromise is advanced as much as possible w/ actual matching request at idle. Certainly actual can match requested a lower static advance, but hard starting does result if N108 duty cycle is >60% at warm idle.

whitedog said:
How about using adaptation to get the spec to match the actual once the basic setting is set at the top of the graph?
I think this would be a bad move IMHO. If advanced to the top of the graph, and adapted such that specified is brought up actual, timing will be advanced ~2* across the board. Efficiency will likely decline as will piston crown and head temperatures at light to moderate load. In most cases, the computer knows best...

Keith_J said:
Wow, the timing is really all over the map. I take it that is because the rotating ring doesn't exactly move all that fast?
Maybe. At 4K RPM, the pump has to make injections a ~35Hz. The pulse signal to the N108 is 20Hz, I believe. The sample rate on VCDS even less than that on my old laptop. Since the advance mechanism is a hydraulic device, the position of the roller ring could also be influenced by the forces applied to it via the cam plate and rollers. Inevitable air in the pump case and behind the advance piston could also have an effect.
 
Last edited:

MOGolf

Top Post Dawg
Joined
Jun 27, 2001
Location
underneath something
TDI
2001 Golf GLS TDI Reflex silver, rough road suspension and steel skid plate, 2004 Passat Variant, Candy White, rough road suspension and geared balanced shaft module, and much, much more. 2016 LR RR HSE TD6, 2019 Jaguar I-PACE
It's a little quicker to click "Basic Settings - 04" on the first window.
Then the 'GO' on the Group 000 line. Skip the bit about going to measuring blocks first.
 

Keith_J

Veteran Member
Joined
Apr 12, 2007
Location
West
TDI
2000 Jetta MT
So N108 is a PWM signal, just like N75. Not exactly Nyquist limitation but still that graph of actual/requested advance shows an underdamped condition.

But timing is based on the CKP and measured once every 2 revolutions from the needle lift (on VE engines) so I can see where it becomes only an issue with steady state conditions. Much like the open loop operation on gasoline engines at WOT.

So how precise is timing under steady state operation?
 

aNUT

Vendor , w/Business number
Joined
Nov 29, 2006
Location
Boulder, Colorado
TDI
'01 TT (ALH-ish), B7 Audi gasser, '05 Golf
MOGolf said:
It's a little quicker to click "Basic Settings - 04" on the first window.
Then the 'GO' on the Group 000 line. Skip the bit about going to measuring blocks first.
Touche. 004 needs to be viewed in measuring blocks though in order to get meaningful data. Otherwise the ECU just cycles the N108.

Keith_J said:
So N108 is a PWM signal, just like N75. Not exactly Nyquist limitation but still that graph of actual/requested advance shows an underdamped condition.

Indeed PWM on the N75, N108, N18, etc.

Keith_J said:
So how precise is timing under steady state operation?
It's a lot more controlled steady state at a more reasonable engine speed. I'll grab a log the next time I'm playing on my computer while driving.
 

tasdrouille

Veteran Member
Joined
Sep 17, 2007
Location
Quebec
TDI
2000 A4
Where do you take your information as to what timing if optimal for efficiency?

I understand the following is not exactly a VE TDI engine, but it's relatively close.

For this specific engine and operating parameters:


I have found the following:
 

aNUT

Vendor , w/Business number
Joined
Nov 29, 2006
Location
Boulder, Colorado
TDI
'01 TT (ALH-ish), B7 Audi gasser, '05 Golf
Michael - You need to tell me to send you an ECU so we can finally cure that freekin' hot start problem!

-Tasdrouille -

I'm not specifying a dynamic timing map. That was done some time ago, and I don't have the ability to change it.

My intentions with this post are as follows:

To address complaints about poor mileage.

To dispel an often echoed (and who knows where it started) dogma about setting the timing on the green for better efficiency and more advance at full load.

It does appear that optimal efficiency in the engine for which you have data for @ at speed of 1500 RPM is achieved with a timing between -4 to 1* BTDC...'bout what a TDI will request.

The timing map in the ECU is set without considerable effort to change it. Therefore, the ECU will adjust the timing to meet it whenever possible. From my experience, better mileage is achieved by setting the hardware such that the ECU can achieve specified advance at all times.
 

tasdrouille

Veteran Member
Joined
Sep 17, 2007
Location
Quebec
TDI
2000 A4
The last picture is of most interest. It shows fuel conversion efficiency, which translates to fuel economy if you will at various start of injection timing and EGR rates.

So best fuel economy for this engine under stated operating parameters would be near 30% EGR rate and SOI timing around 4 BDTC.

If this engine had an EGR delete, best fuel conversion efficiency would happen at roughly 2 ATDC, again at 1500 rpm.

I am one to believe the timing map in the ECU at low and medium loads is retarted from optimal fuel conversion efficiency on a stock car. You can't change the map itself without a tune, but you can easily offset it across the board through adaptation and/or add roughly 1.5 degrees of advance over the baseline at low and medium loads only through an IAT mod.

If you go in adaptation and set the specified advance at idle to be 2.6 degrees BTDC (to take your example) you'll have matching specified and actual.
 
Last edited:

paramedick

TDIClub Enthusiast, Vendor
Joined
Jul 29, 2001
Location
Versailles, Kentucky
TDI
2015 Audi Q5 TDI, 2000 New Beetle
aNUT said:
I've been following this with interest. In the last 3 days, I have had an A3 and a B4 in here for timing belts. Set the timing like normal in basic settings, and all is good. Timing set a mid-graph. Cars run like a champ.

Switch to channel 4 in basic settings, and I get a notation of "late" in one of the boxes. See the example below from another thread.



Comments, please, as to what if anything I'm doing wrong or why the late notation pops up.
 
Last edited:

eb2143

Top Post Dawg
Joined
Dec 26, 2005
Location
Rhode Island
TDI
None
Very interesting thread aNUT. The fundamental difference I see here is your understanding that the ECU timing map is set for optimal efficiency. There was (is?) a strong contingent of people (see post 17 and "And the Results are...." thread of a few years ago), who claimed that VW sacrificed efficiency for longevity (lower peak cylinder pressures). What is your opinion on this? If I understand, you have gotten data from advanced and matched cars and found that the computer (requested & actual matched) achieves better efficiency.

I don't think it would be very difficult at all for more people to get some hard numbers in terms of FE with these two methods of setting timing. It only takes a few minutes to adjust back and forth.

Can you also clarify a bit your opinion for someone living in a cold climate? You mentioned N108 requests full advance for cold starts and that hard starting does result if N108 duty cycle is >60% at warm idle. This doesn't mean very much to me.
Is it still best for me to use adjustments to static timing (via 3 IP bolts) to get actual to meet requested in group 004? Or will cold starting be helped by setting near the green (which is the only reason I have ever heard for doing such; EGT stuff is new to me)?

I always suspected the belief that setting static timing near the top was essential for good cold starts because I had it near the red and cold starts were excellent.

In any case, we do know who propagated these ideas about timing in basic settings, and it'd be nice to hear from him.

Thanks again; I really hope that we come to a consensus on ALH timing.
 
Last edited:

jcrews

Top Post Dawg
Joined
Jul 28, 2007
Location
Round Rock, TX - VCDS
TDI
All gone
paramedick said:
I've been following this with interest. In the last 3 days, I have had an A3 and a B4 in here for timing belts. Set the timing like normal in basic settings, and all is good. Timing set a mid-graph. Cars run like a champ.

Switch to channel 4 in basic settings, and I get a notation of "late" in one of the boxes. See the example below from another thread.

<image removed>

Comments, please, as to what if anything I'm doing wrong or why the late notation pops up.
It is by design. The display will switch between early and late, so you can evaluate the advance solenoid function. Use measuring blocks 004 instead.
 

jcrews

Top Post Dawg
Joined
Jul 28, 2007
Location
Round Rock, TX - VCDS
TDI
All gone
I've noticed the timing control above about 1500 RPM can pull back to <0.7 deg BTDC. At 70, it should be fine. I may check this out later for more conclusive evidence.

The lines on the graph are intended to allow the engine to operate correctly as long as the basic setting is in between them. Therefore, I disagree with the statement that an appreciable amount of fuel economy can be lost as a result.

Anecdote: 56.1 MPG at 70-75 avg 65 for ~14 hrs. Repeated several times.
Average fuel econ is as good or better than before the timing belt change.

I would like to see, and may (eventually) check out, is if the specified start changes with the basic setting. If it does, it can affect the operation, otherwise, the control range is wide enough that it won't make a difference except when starting.
 

eb2143

Top Post Dawg
Joined
Dec 26, 2005
Location
Rhode Island
TDI
None
jcrews said:
I would like to see, and may (eventually) check out, is if the specified start changes with the basic setting. If it does, it can affect the operation, otherwise, the control range is wide enough that it won't make a difference except when starting.
Explain to me why it only matters if the basic setting (which I take to mean the intersection of the yellow lines) can affect specified start.

If it affects actual start of ignition across the entire operating range in 004...well then isn't it affecting operation?

I don't understand the "control range" and how you know its extent.
 
Last edited:

jcrews

Top Post Dawg
Joined
Jul 28, 2007
Location
Round Rock, TX - VCDS
TDI
All gone
eb2143 said:
1. Explain to me why it only matters if the basic setting (which I take to mean the intersection of the yellow lines) can affect specified start.

2. If it affects actual start of ignition across the entire operating range in 004...well then isn't it affecting operation?

3. I don't understand the "control range" and how you know its extent.
1. If the basic setting has no effect on specified start, then as long as it is set within the design limits (chart), then it makes no overall difference, since the ecm can place the start where it needs to in closed loop mode.

2. Yes. But I haven't proven if it affects SOI across the entire range. Illustrated was a difference at idle, however, I've not observed the same difference at nearly all other operating conditions, given a small data set (1300 or so points if I recall correctly). More information is required to draw a conclusion.

3. By control range, I was referring to the start of injection basic setting chart. Outside of that range could cause a control deviation from pushing the physical limits beyond what the hardware can adjust through ECM control.
 

paramedick

TDIClub Enthusiast, Vendor
Joined
Jul 29, 2001
Location
Versailles, Kentucky
TDI
2015 Audi Q5 TDI, 2000 New Beetle
jcrews said:
It is by design. The display will switch between early and late, so you can evaluate the advance solenoid function. Use measuring blocks 004 instead.
OK, I'll buy this. Just notice that his tutorial shows block 04 in basic settings. Obviously, this is an error, or does that message show up in measuring blocks?

I'm gonna play with my car in the morning and see.
 

jcrews

Top Post Dawg
Joined
Jul 28, 2007
Location
Round Rock, TX - VCDS
TDI
All gone
If you're talking about aNut's picture, it is in measuring blocks (button labeled switch to basic settings).
 
Last edited:

tasdrouille

Veteran Member
Joined
Sep 17, 2007
Location
Quebec
TDI
2000 A4
At any time, the ECU knows the actual injection timing from the needle lift sensor.

There is a range the ECU can adjust the SOI timing. The start of this range is defined by the basic timing, the actual physical timing.

So if you set the basic timing (physically) at the top of the graph, and at idle the ecu is requesting 0 degrees, chances are it will not be able to adapt since the requested value is out of its adaptable range due to physical limitations.

However, as soon as the requested value goes over the lowest limit of the adaptation range, say cruising down the freeway at 60 mph, the requested and actual SOI timing will match, regardless where your basic timing stands. Go out and see for yourself in vcds.

Basic timing does not affect operation timing unless requested value is lower than basic value for a given operating condition. You need to go in adaptation to offset the whole timing map across the range.
 

greenskeeper

Top Post Dawg
Joined
Mar 10, 2003
Location
USA
TDI
1998 Jetta TDI
This is interesting.

I have had my timing at every part of the graph and not a difference in requested vs actual start. Maybe in your case the pressure in the pump isn't sufficent for advance at idle?

As far as fuel economy, I run the timing just above the upper line with max advance via adaptation (33268) which gives 5 deg advance across the operating range.

It is interesting to note that with stock timing the ECU can request 2 deg ATDC while crusing...probably to reduce Nox at the expense of economy.
 

tasdrouille

Veteran Member
Joined
Sep 17, 2007
Location
Quebec
TDI
2000 A4
Yup, when slowly cruising, say at 60, timing is advanced at light loads (on the flat), pulled back at medium loads (slight grade) and then advanced again at high loads (6% grade).
 

Bob_Fout

Oil Wanker
Joined
Sep 5, 2004
Location
Indiana
TDI
2003 Jetta - Alaska Green (sold) / 2015 GTI 2.0T
This past weekend I asked Paramedick to retard my timing some. He moved it from halfway between the middle and upper line to just below the middle line. This is where specified and actual seemed to match.

Differences I've noticed:
  • Peak EGTs went way the heck up. I was able to reach 1650*F doing 3rd gear WOT from 2K to 5K. In 5th, going from 60 MPH to 80 MPH or 70 MPH to 90 MPH gets up to 1500*F. This is 100 to 200 degrees more than before in each situation.
  • Feels smoother on acceleration.
  • There is little to no shudder when you suddenly let off the go-pedal.
  • Could be my imagination, but it also feels quicker.
I like the changes except the EGTs.:eek:.

I understand the trade-off of EGTs is cylinder pressure when it comes to timing advance. Methinks the block is more robust than my current fragile hairdryer turbo.

I'm going to bump timing in vag-com and see if EGTs can be kept in check, until I get a bigger turbo.
 
Top