Injection timing (advanced, closed-loop. etc) questions

This isn't bait. Seriously. I've read many, many timing threads and this has never been settled.

1) Since it seems the timing is closed-loop, what is the difference between being at one end of the timing band or the other, as long as the ECU can achieve the requested "Start of injection" value? Different 'authoritative' posts, including the FAQ, specify that it's better to be in a certain part of the graph.
I don't doubt there is a difference, but if the ECU wants 12° and gets it, does it matter how much it has to modulate the advance mechanism to get it? People claim to see operational differences with different timing settings but this has never been really explained. Could it be that adding 50% modulation to achieve the 12° isn't really the same as achieving the 12° with no mudulation?

2) Is the top of the graph really "advanced?" I was readjusting my timing last weekend because it was at the bottom of the graph. I slipped the belt one position forward on the pump sprocket (because I was at one end of the slot) and the timing value shot off the top of the graph. This should have retarded the timing, as I see it, because the engine rotates "forward" and this would have caused the pump to reach the same position later... i.e.; I retarded the timing. It's remotely possible that I observed this wrong, but I really don't think so.

- Anton
'98 TDI, 104k miles

1. In "Basic Settings" you can see the injection timing of the pump, without adjustments from the ECU. If it is in the middle of the graph, the pump can be adjusted internally a certain amount advance and retard. If you adjust the pump wheel so the basic setting is closer to the top, the ECU has the same amount of adjustment in the pump, but now it can give more advance and less retard.

2. I can't advise on question 2 as I was not there to see what you actually did.

Originally posted by Anton Largiader:
I was readjusting my timing last weekend because it was at the bottom of the graph. I slipped the belt one position forward on the pump sprocket (because I was at one end of the slot) and the timing value shot off the top of the graph.
- Anton
'98 TDI, 104k miles

Click to expand...

<font size="2" face="Verdana, Helvetica, sans-serif">What slot were you at the end of? Usually the A3 pump timing is adjusted by moving the pump (not the gear via the bolt slots).

Originally posted by Anton Largiader:
I don't doubt there is a difference, but if the ECU wants 12° and gets it, does it matter how much it has to modulate the advance mechanism to get it? People claim to see operational differences with different timing settings but this has never been really explained. Could it be that adding 50% modulation to achieve the 12° isn't really the same as achieving the 12° with no mudulation?

Click to expand...

<font size="2" face="Verdana, Helvetica, sans-serif">I think you're right, it's never been fully explained. Basically the timing is controlled by a feedback loop using various sensors on the engine. However, the feedback loop doesn't work instantaneously. Some time is required for the adjustment to occur, and during that time the injection is advanced/retarded, depending on how it's mechanically timed.

Seems to me that with the basic mechanical timing set more advanced the ECU 'can' run more advanced timing, but "won't" because it has a fixed map of where it is supposed to be running, and will use that setting if at all possible. If one were to adjust the pump position so far advanced that it could not retard it to compensate, then you would get more advanced timing, but I'm sure other bad side effects would creep in.
-Les

Originally posted by VW Vet:
[QB]1. In "Basic Settings" you can see the injection timing of the pump, without adjustments from the ECU. If it is in the middle of the graph, the pump can be adjusted internally a certain amount advance and retard. If you adjust the pump wheel so the basic setting is closer to the top, the ECU has the same amount of adjustment in the pump, but now it can give more advance and less retard.

Click to expand...

lesd's comments are exactly right.

the "basic setting" of timing sets the possible range of timing, and the ECU adjusts the timing within this range as appropriate according to internal maps... it uses the needle-lift sensor in injector three to detect the actual start of injection, and modulates the command to the timing solenoid in the pump (called the "cold-start solenoid" in all the VW lit).

tadc, i beg to differ... the method that the timing control works has been fully explained here in the forums, many times. the problem is in knowing which people to believe. there has been posting in the past by people who don't understand how the timing system really works, and unfortunately that leads to confusion.

another excellent source is the bosch "yellow book" on axial distributor pumps... it very clearly explains how the timing works.

some more info: there are two cases when timing is completely controlled by the basic setting of the pump. both of these cases share the common ground of not having feedback from the needle-lift sensor. The first case (which GoFaster has mentioned recently) is when starting the engine. At engine start, the NLS signal is ignored, as the ECU waits for a "good" signal out of it... so setting timing to the top half of the graph will advance timing during starts. The other case is when coming off of an overrun condition. This is when your foot is off the pedal, engine is in gear, and the car's inertia is driving the engine at over idle speed. in overrun, the ECU injects NO fuel... so there is no signal from the needle-lift sensor (there is no start-of-injection to detect!) when the engine nears idle or you put your foot down, the ECU starts injecting fuel again... and the basic setting of the pump determines the relative advance/retard of timing until a few revolutions of the engine have passed and the ECU gets a good signal from the NLS. then it goes back into closed-loop mode.

Actually, it is called the "solenoid valve for start of injection" inside the VE 37 pump.

Yeah, that's what it is and what Bosch calls it, but notice that I said in all the VW lit. Bentley insists on calling it the cold-start solenoid. I pointed that out so that people who read the Bentley don't get confused.

Thanks for the input, though, Slypoop.

Speaking of 'cold-start', does any one know what adjustments are made and how during low-temperature starts?

Control Loop diagram

[ December 18, 2002, 19:24: Message edited by: KERMA ]

Thanks for the inputs, but with one exception it's a rehash of the same old material. What I want to know is, what is the difference between setting at the top vs. the bottom of the graph, if the ECU adjusts it to the same nominal value anyway?

I realize that at extreme ranges of adjustment the timing can be changed as the pump runs out of internal adjustment, but is this the actual reason why people recommend setting "advanced"? Does anyone actually know that this is having any effect at all during operation? Will the requested "start of delivery" not match the actual at some point, and wouldn't the ECU throw a DTC if it couldn't get the desired value?

2. I can't advise on question 2 as I was not there to see what you actually did.

Click to expand...

<font size="2" face="Verdana, Helvetica, sans-serif">That's why I'd like to see what the group knows, independent of what I did or observed.

What slot were you at the end of? Usually the A3 pump timing is adjusted by moving the pump (not the gear via the bolt slots).

Click to expand...

<font size="2" face="Verdana, Helvetica, sans-serif">Yes, the pump was maxxed out in its mounting slot, so I jumped the belt one tooth on the sprocket. Turns out I probably moved it the wrong way, hence my question regarding which side of the graph is advanced. I think the only adjustment on the A3 is at the pump anyway, not at the pulley.

- Anton
'98 Jetta TDI

Originally posted by DavinATL:
tadc, i beg to differ... the method that the timing control works has been fully explained here in the forums, many times. the problem is in knowing which people to believe.

there has been posting in the past by people who don't understand how the timing system really works, and unfortunately that leads to confusion.

Click to expand...

<font size="2" face="Verdana, Helvetica, sans-serif">True, but the "full explaination" doesn't fully account for all the differences observed when running "advanced" or "retarded". Theoretically, there should be NO difference in operation except for the cases cited above(start and overrun, plus the extension of the allowable "range"), but in reality the entire character of the car is changed slightly, for example my fuel economy went up several MPG when going from 50% retarded to 50% advanced.

And you determined that you were 'advanced' how, exactly? Which direction did you move the pump to get this? My understanding is that you would rotate the pump towards the engine (or make the equivalent adjustment to the pulley on an A4).

-Anton
'98 A3 Jetta

As you say, there are many opinions on timing advance on the TDIs. Here's mine. Although many manufacturers track actual igntion timing most systems are not closed loop. The only time they are closed loop is when there are knock sensors(not a good idea on a Diesel). The TDI does not use closed loop timing. That's not to say it doesn't control advance, but the numbers don't change as a result of any actual timing input numbers. I've done some emission testing on the TDI. The changes in NOx emissions when changing pump timing is evidence enough for me that altough the processor does see the actual timing it doesn't do anything about it.
Joe Rappa

Joe, at what RPMs do you measure NOx?

BTW, if it's not closed loop, then what are the #3 injector and Commencement of Injection solenoid doing?

Perhaps the signal from #3 injector is used for a very slow timing compensation. The idea was to adjust for a stretched belt, or a slightly misadjusted pump position, so it may take a few minuites to let the ecu take a whole lot of samples and make a change. There is no need for this to happen fast. I suppose one can trigger a timing light with a signal derived from injector #3, and see where the timing really is.
-Les

From the Bosch "technical instruction" for the VE pumps:



Solenoid valve for start-of-injection control The pump interior pressure is depen- dent upon pump speed. Similar to the mechanical timing device, this pressure is applied to the timing-device piston (Fig. 4). This pressure on the timing- device pressure side is modulated by a clocked solenoid valve.

The start of injection has a decisive in fluence upon starting, noise, fuel con- sumption, and exhaust emissions. Start- of-injection maps programmed into the ECU take these interdependencies into account. A closed control loop is used to guarantee the high accuracy of the start-of-injection point. A needle-motion sensor (NBF) registers the actual start 0 injection directly at the nozzle and compares it with the programmed start of injection (Figs. 2 and 3). Deviations result in a change to the on/off ratio of the timing-device solenoid valve, which continues until deviation reaches zero.

[ December 21, 2002, 20:02: Message edited by: KERMA ]

What I've seen is that the closed-loop control used for injection timing is VERY tight and fast-responding. Provided that the actual timing is within the range that is mechanically possible, requested and actual timing will ALWAYS match. It stands to reason that under any operating condition where the requested timing is mechanically possible to achieve, it won't make a difference.

Adjusting the pump adjusts the range which is mechanically possible, and it adjusts the timing that will happen under conditions when the automatic advance mechanism is not operating (i.e. cold starting).

Cold starting, and full load / higher speed, are the operating conditions where mechanical timing adjustment will matter. The first, is because the automatic advance mechanism does not operate during cranking. The second, because those are the conditions where the ECU wants the most advance, and if your pump timing is late, it might not be able to get what it wants. Result, lower power, more smoke, higher exhaust temperature. And that's under the circumstances where those effects are the worst possible...

Boundless,
I measured NOx at lots of RPMs. Check out the following thread.
NOx testing
KERMA:
If the ECU corrects timing then why the changes? I understand that Bosch claims closed loop timing control but I can't verify it with testing. I'm not trying to be argumentative (this is one nobody is going to win anyway) just stating what I've witnessed myself.

Joe Rappa

To verify the range of injection timing control, use VAG-COM (or equivalent) as follows (assumes basic knowledge of how to use the tool):

Go to engine controller basic settings.
Go to group 004.
Zone 2 should show "advanced" and zone 3 should show the maximum advanced timing (should be 7.0 to 9.0 degrees BTDC).
After approximately 10 seconds zone 2 should say "retarded" and zone 3 should show the maximum retarded timing (should be 3.0 to 5.0 degrees BTDC).

Now if you change the physical setting at the pump, the numbers displayed according to the procedure above should change.

As for "closed-loop" vs. "open loop" I found this info in a description of the purpose of the needle lift sender:

"The needle lift sender signal is used to determine the commencement of injection. If it fails the commencement of injection switches to open-loop control (according to engine speed and engine load). During normal operation the commencement of injection is controlled by a closed-loop function (according to engine speed, engine load and temperature)."

So if your testing did not change the engine speed, load or temp, this would have influenced your findings.

Joe:

I dunno. Just plagiarizing the Bosch publication. You must have me mistaken for someone who knows something!!

Originally posted by Anton Largiader:
And you determined that you were 'advanced' how, exactly?

Click to expand...

<font size="2" face="Verdana, Helvetica, sans-serif">Vag-Com.

Injection Timing at Startup

The feed pump supplies the fuel pressure to the injection timing mechanism.

The Commencement of Injection (COI) solenoid valve is used to control the application of that pressure to the injection timing piston.

The force produced by the fuel pressure acting on the piston is opposed by a spring. The spring is always trying to retard the injection timing.

When the COI is "on", it is bleeding off fuel pressure, as in not allowing pressure to be applied to the injection timing advance piston. This will result in timing being retarded.

When the COI is "off", it is not bleeding off fuel pressure but is applying fuel pressure to the injection timing advance piston. This will result in timing being advanced. Having the COI off during start up makes sense.

Now the feedpump is a positive displacement pump. As soon as it turns, it is making pressure.

The pressure applied to the injection timing piston is modulated by turning the COI on & off really really fast.

On startup, the COI is off. This allows for maximum advance during startup. As soon as a feedback signal is available from the NLS, the ECU sends commands to the COI and brings in the timing. The ECU is always in control. It's a matter of when feedback signals are available to implement that control.

If the basic settings timing is set advanced, the COI will be on more to bleed more pressure off than if the pump basic timing were retarded.

If the the basic settings timing is retarded, the COI will be off more to get higher pressure to overcome the spring and get the timing up to the setpoint.

Now, when there are starting problems that are "fixed" by advancing the basic settings timing, well that doesn't mean it was a real timing issue. What it means is that there is probably a weak feedpump or a malfunctioning COI that isn't sealing off the bleed port properly, preventing the maximum available feedpump pressure from being applied to the injection timing piston.