Verbrauchsoptimierung via Spritzbeginn: Soft- oder Hardware? | Posts 32+

Hi everyone! So far, the thread has provided some interesting insights, such as Ulf's mechanical late-adjustment mechanism with EDC's mercy, or Rainer's characteristic curves, which I had to further process with PaintShop in order to be able to recognize anything.

Allerdings ist das Thema auch ein wenig in Richtung AGR abgeglitten, dennoch bin ich Wolfgang synchro noch eine Antwort schuldig. Here it is:

The assumption that the exhaust gas cannot participate in a chemical reaction a second time is incorrect.

As a mathematically precise statement, it is certainly correct. That's why I wrote 'practically, the entire exhaust gas is no longer available as a reactant.' So only a negligible portion reacts again usefully.


The conclusion mentioned .

Doch, hat er. Here it is:

If the EGR rate is 50%, the amount of exhaust gases leaving the engine is only half as much.

This is a clear statement. I have explained in detail in my response to 'bafische' why I believe it to be incorrect.

But there's some truth to it; the 50% of exhaust gases aren't simply blown out, but instead return to the cylinder, and that's where things get interesting. There's a significant chance that pollutants can be converted back into 'good' substances.

You are phrasing it correctly: There is a chance for each molecule of the combustible gases in the exhaust that, upon its second pass through the working cycle, it will be completely oxidized. I have already explained in my answer to bafische why I consider this chance to be extremely low (although the reaction is not impossible, but rather occurs on a very small scale) – namely, due to the homogeneous mixture and its flammability.

If you want to contradict this now, you would have to provide plausible evidence that the total combustion products with EGR contain twice as much or at least significantly more pollutants than without EGR.

I don't need to make this plausible because numerous well-known measurement results prove it. This obviously doesn't apply to NOx, but rather to the pollutants produced from the reactants: fuel + oxygen.

With EGR: CO, HC, Particles, NOx values.
Without EGR: CO very low, HC very low, particulate matter very low, NOx increased.

You are making the distinction here between the reaction of fuel and oxygen, and 'unaffected' gas. That's not how it works (see above).

No, I have made the distinction between the useful reaction of fuel + oxygen, the incomplete and therefore harmful reaction of fuel + oxygen, and the harmful reaction of nitrogen + oxygen, and an uninvolved gas.

Das geht sehr wohl und ist auch höchst sinnvoll. Because that's what it is:

The only gas that is relatively uninvolved is argon, because it is a noble gas.

You probably don't mean that entirely seriously. With a nitrogen content of 78% in the atmospheric air, the proportion that reacts with oxygen in the engine represents an extremely small amount. Therefore, one can very well consider nitrogen as a practically inert gas. Furthermore, CO2 and water vapor are also included, whose proportions in the atmosphere, to my knowledge, are even higher than those of argon.

Therefore, it is a fallacy to consider the absolute mass flow of exhaust gases as being correlated with their harmfulness. The EGR reduces the mass flow of intake air by the percentage corresponding to the EGR rate and (conditionally) also the absolute exhaust gas mass flow. However, what is decisive for the harmfulness is the mass flow of reaction products from the fuel+oxygen reaction and side reactions, which as a whole is not reduced at all by an EGR (only the balance of harmful substances is shifted).

Simply put: If you're blowing extra air into the exhaust system, you're also increasing the mass flow of exhaust gases. Does this make the exhaust more harmful?

Okay, and now we're back to the legendary afterburning of CO and HC:

The 1% do not need to ignite independently; they participate in the ongoing combustion process, as I described above. Furthermore, they do not necessarily have to be located directly adjacent to the injection jets, because the combustion takes place throughout the entire cylinder.

To ignite a gas mixture, an initial activation energy is required. An exothermic reaction occurs when the reaction energy is greater than the activation energy. The combustion can propagate spontaneously if the excess reaction energy is not consumed by heating a large proportion of unreacted gas that contains the gas mixture.

That's why gas mixtures must have a certain minimum proportion of fuel and oxygen relative to the inert gas in order to be flammable. If this condition is not met, the 'total activation energy' must be higher than the subsequent reaction energy, because you have to heat up so much of the inert 'interfering gas' with your activation energy before the ignition temperature is reached throughout the mixture.

It is certainly the dream of engine developers to keep the combustion process away from the cylinder walls, but as far as I know, this has not yet been achieved.

I think you're mixing something up. What you mean is the combustion zone in the combustion chamber, which are the areas where the reaction takes place. Are you familiar with the impressive images of the glass diesel engine taken with a high-speed camera? Unfortunately, I don't have the link handy right now. You can clearly see how localized the actual combustion is within the combustion chamber of the diesel engine. You can see yellow-white flames shooting out from the injector nozzle.

What the engine developer dislikes is the cooling of the newly formed hot reaction products, which contain energy due to their high temperature and resulting high pressure, and which should ideally be transferred to the crankshaft before they cool down.

Under no circumstances should the reaction zone (and only the reaction zone is capable of igniting the locally homogeneous mixture of 1% flammable gas) come into contact with the walls of the combustion chamber! Otherwise, you will see what happens in improperly tuned or over-tuned TDI engines: piston melting.

To summarize the issue of Diesel Particulate Filters (DPFs): We are all certainly interested in reducing NOx emissions while simultaneously minimizing other pollutants and achieving high long-term engine efficiency. These are two demands that unfortunately contradict each other with the current technical solutions being implemented.

Regards,
Holger.

Hello,

I don't want to delve into this matter any further, especially because I get the impression that we are not really understanding each other. Also, the discussion isn't yielding any truly new insights; AGR is an established technology, and if it weren't beneficial, manufacturers wouldn't bother with it. So our discussion is essentially just about semantics. Therefore, I will only provide a brief response:


To ignite a gas mixture, an activation energy is initially required.
...
This is why gas mixtures must have a certain minimum percentage of fuel and oxygen compared to the inert gas in order to be flammable.

Ich muß mich jetzt wiederholen. Was du schreibst, gilt, wenn die Reaktion bei Umgebungstemperatur startet und 'gezündet' werden muß. Die zahlreichen Schadstoffreaktionen, die im Zylinder ablaufen, finden jedoch bereits eine hohe Temperatur vor, so daß die Vorstellung mit der Aktivierungsenergie überflüssig ist. Auch die Vorstellung, daß Reaktionen immer in einer Richtung ablaufen, ist falsch. Beispiel sei etwa das Boudouard Gleichgewicht 2 CO <C> CO + H2O + pollutants is incorrect.



What you mean is the combustion zone within the combustion chamber, which refers to the areas where the reaction takes place.
...
The reaction zone itself (and only the reaction zone is capable of locally igniting the 1% flammable, homogeneously mixed gas) must, for heaven's sake, not come into contact with the walls of the combustion chamber!

'Of course, there are hotter and cooler areas within the combustion chamber, but I'm not entirely clear on what criteria you're using to differentiate them. In TDI engines, the intake air is swirled, and 5- or 6-hole injectors are used, all in an effort to distribute the reaction zone as evenly as possible throughout the cylinder.'


Have you seen the impressive images of the glass diesel engine recorded with a high-speed camera? Unfortunately, I don't have the link to hand right now. } It vividly shows how the actual combustion process takes place locally within the combustion chamber of the diesel engine. You can see yellow-white flames shooting out from the injector.}

No, I'm not familiar with that. However, it's important to note that with cameras like these, you can typically assign arbitrary color values to different temperatures. They are not necessarily calibrated to match the visual perception of the human eye. Why would they be? If this is a demo video for the press, and the goal is to demonstrate what you're describing, then the colors have likely been adjusted to make that particular aspect stand out. In that case, the information about the actual temperature distribution within the cylinder may not be entirely accurate.

Hello.
Wolfgang.

I also don't want to expand on the discussion about the 'Arbeitsgruppe Rechtsextremismus' (AGR), especially since this thread is actually about a different topic. I just want to comment on one general statement made by Wolfgang:

The EGR system is established, and if it didn't provide any benefits, manufacturers wouldn't implement it.

This is a typical 'knockout' argument that can be used to justify almost anything and effectively stifles any potential objections.

Yes, the EGR system does indeed achieve something: it's a simple and inexpensive way to shift the balance between different types of pollutants so that they fit within the template created by bureaucrats, which is called the 'emission standard.'
It is questionable whether it actually reduces the actual damage caused by vehicles through their exhaust emissions - especially when considering the secondary effects of the EGR system: constricted intake passages, stuck EGR valves, and consequently, catastrophic emission levels. When assessing environmental relevance, these real-world consequences and the periods during which cars continue to operate with this defect must be statistically included.

For those who are interested: A comprehensive treatise on the exhaust emission issues of automotive diesel engines, sometimes also in comparison to gasoline engines: http://fvkma.tu-graz.ac.at/Skripten/Wachter/VO-Wachter_2004.pdf

***************************************************************************


Okay, now let's talk about the start of injection and consumption. After much consideration, I came to the conclusion that both an adaptation and a speed-dependent electronic delay have the disadvantage that they do not eliminate false peaks in the injection start characteristic curve.
'Alibi humps' are two short peaks in the injection start field that correspond to two standardized constant speeds in the part-load range of the ECE driving cycle (90 and 120 km/h). These peaks shift the injection start precisely at these speeds into the region of optimal efficiency, but at the expense of NOx emissions. The result is advertising-worthy fuel consumption figures for brochures, while simultaneously complying with NOx emission limits.
http://www6.1mb.at/uploads/06.02.2005_16:42:02_kf-spritzbeginn-org02.jpg
Unfortunately, I cannot specify which vehicles have these 'alibi humps' or what their exact values are. However, they must be considered when there is a global map shift. They have consequences: At the peaks, the spark advance is greater than intended and can, in total, exceed the efficiency maximum.

The only way to adjust the humps is Ulf's late-impact shifting method. Since this involves a relatively complex mechanical adjustment, one can unfortunately only approach the tolerance limit of the EDC in a very uncomfortable manner, and especially not while driving.

Therefore, I devised a method that achieves the same effect as Ulf's method, but is much more convenient and, above all, can be adjusted from the dashboard.

The SBV valve is closed in the de-energized state, which causes an increase in the pump internal pressure. This pressure acts on the non-spring-loaded side of the SBV piston and shifts it towards the 'early' position. Increasing opening allows fuel to bypass into the return line, reducing the pump internal pressure, and the spring then moves the SBV piston towards the 'late' position.

Since the SBV valve is a proportional valve, the maximum control current can be limited so that the maximum valve opening is no longer reached, which prevents the maximum bypass effect from being achieved. The minimum internal pressure of the pump is increased. This is equivalent to having a throttling valve in series with the SBV valve.

Instead of compensating for a mechanical late injection, the system now compensates for a hydraulic late injection. The internal pump pressure can no longer be reduced sufficiently to achieve the maximum allowable delay angle as defined by the characteristic curve. Using a 5W potentiometer on the dashboard, it is possible to approach the limit that the EDC has not yet triggered as an emergency mode, and in the long term, also optimize for minimum fuel consumption.

The fuel injection map is not utilized at higher engine speeds and loads because the start of injection is adjusted so far in advance that the solenoid valve (SBV) is already relatively closed.

What do you think of my idea (which is admittedly a bit experimental)? Have I overlooked anything?

Best regards,
Holger.

Okay everyone, now it's getting serious. A 4-turn potentiometer (10 ohms) should be sufficient to dissipate excess control power in series with the SBV valve (SBV valve: 15 ohms, maximum control voltage from the EDC: 12V, measured with an analog multimeter).

To increase the adjustment range, a 47-ohm resistor is added in parallel with the SBV valve. This resistor slightly increases the load on the switching stage of the EDC when the potentiometer is set to 0 ohms. The main thing is that the EDC doesn't complain when the resistance of the SBV valve (with the potentiometer) suddenly increases to over 20 ohms.

Okay, I've been thinking about the most elegant and 'invisible' way to route the SBV potentiometer cable to the dashboard. In my opinion, it makes sense to tap directly into the EDC (Electronic Differential Control) unit for the SBV cable. (The EDC is located quite close to the dashboard.) I took a picture of it six months ago:

http://www6.1mb.at/uploads/06.02.2005_07:48:15_edc.jpg

'Are the black parts where the wires disappear detachable connectors? What's the quickest way to identify the two SBV (signal bus voltage) wires? Does anyone have a pinout diagram for the connectors?'

I tried to disconnect the plugs back then. However, I couldn't manage it, and I didn't want to break anything. The plastic retaining clip for the EDC had already broken at that time. You can really get very close to it.

And once again, I ask the most crucial question to all the present experts and specialists: How much degree of deviation in the start of injection can the EDC still tolerate without triggering a fault mode ?

Thank you in advance for any comments you may have.
Best regards,
Holger[/b].

Hi,

Quote:

Are the black parts where the wires disappear the detachable connectors?

At least on the Ibiza (where it's a bit easier to access), there are sliding locking mechanisms on both connectors at the MSG (Multi-Sensor Gateway). You slide these aside, and then you can unplug the connector. I can't imagine that it would be different on the SDI in the Arosa.

Unfortunately, I can't help you with finding a suitable accommodation. I assume you have already searched and checked whether any of the listed accommodations might be a good fit for you?

Best regards,

January.

Hi Jan, Okay, I'll go ahead and look for availability schedules.

Here's another consideration regarding my idea: Unfortunately, the internal pump pressure generated is speed-dependent. If the internal pump pressure increases more progressively with increasing speed while the SBV valve remains open, and this increase exceeds the injection start point indicated by the characteristic curve, then my solution would not work. Therefore, if there are operating ranges where the SBV valve needs to open even further as the speed increases, in order to counteract premature timing advance, I have a problem. This could potentially occur in the blue characteristic curve (see above), for example, at 3mg/stroke @ 3200 RPM. This would represent a very late injection start at a relatively high speed. It is quite possible that with a 'throttled' SBV valve, the injection start point could then rise to the level of full load (40mg/stroke) because the generated internal pump pressure is so high. This would temporarily trigger an emergency mode.

Well, let's start by experimenting. Let's see what comes out .
hello

SeatArosa1.7SDI wrote:

..... "Alibi humps" are two short peaks in the injection start field that correspond to two standardized constant speeds in the part-load range of the ECE driving cycle (90 and 120 km/h). These peaks shift the injection start precisely at these speeds into the region of optimal efficiency, but at the expense of NOx emissions. The result is advertising-worthy fuel consumption figures for brochures, while simultaneously complying with NOx emission limits. Unfortunately, I cannot provide information on which vehicles have these "Alibi humps" or what specific values they represent. Best regards, Holger

.

Hello,
I've always wondered why the instantaneous fuel consumption suddenly becomes noticeably lower when the speedometer exceeds 2000 RPM (approximately 90 km/h in the SDI). Just above and below this exact speed, the fuel consumption always increases by about 0.3 liters per 100 kilometers, which is why I always set the cruise control to that specific speed.
Now I know why.
What surprises me is that the engine becomes slightly quieter at exactly this peak RPM, and runs smoother.

The SDI (ASY) cannot be adjusted to earlier software versions via diagnostics, and adaptation is not possible, which may be the reason why.

EDIT: Da durch den Kühlmitteltemperaturgeber G62 die Krafstoffmenge erhöht und der Spritzbeginn vorverlegt wird je niedriger die KM-Temperatur ist, wäre es naheliegend das Kühlmittelsignal zur EDC zu verfälschen, oder? Wenn Yes, HOW?

Best regards,
Pappl

Hello,
Do all VP37 TDI/SDI engines have a 4-pin G62 connector?
"Two wires for the EDC (Electronic Diesel Control) and two wires for the instrument cluster. It would be great if the instrument cluster display still showed the correct information, especially to prevent overheating of the coolant. As far as I know, the radiator fan is not controlled by the G62."
Furthermore, the question arises as to which two poles lead to the electrostatic discharge (ESD) and where the resistor should be connected.

ulf wrote:

the lower the temperature is reduced, but it is also limited to a maximum value, even if the engine overheats

.
I don't quite understand. Are you saying that you assume a faked -10°C coolant temperature at a real temperature of 80°C is the maximum value for the sensor, regardless of how much the resistance could be artificially lowered?
It wouldn't really matter anyway, but it would be nice if the EDC (Engine Control Unit) believed that the coolant temperature is 60°C instead of 80°C when warm. I think too much "farmer tuning" isn't good either, because the injection timing is also advanced at high RPMs, even if only slightly.
But I believe it would be better and safer than the "10-cent NHG capacitor" and the "10-cent pump connector with a 4+7 resistor."
Chip tuning probably wouldn't be worth it for the SDI engine because the entire system relies on turbo boost, which is why I find it interesting.

PAPPL wrote:

Are you saying that if a faked -10°C coolant temperature is assumed at a real temperature of 80°C, that's the maximum value for the sensor, regardless of how much the resistance could be artificially lowered?

Suppose that an additional series resistor of 470 ohms would correspond to 40°C with a functioning sensor (values chosen arbitrarily). Then, the faked temperature can never exceed this 40°C: Even if the sensor only had 20 ohms at a real temperature of 150°C - because the engine control unit (ECU) would then see the sensor resistance as 470 + 20 = 490 ohms, which would likely correspond to 39°C.

Hi,

Oh my goodness, all the things people come up with. a) The advance of the injection timing varies depending on the engine and can cause other unwanted "side effects," especially at low coolant temperatures. Reducing fuel consumption is more likely to be coincidental in individual cases.

b) When adapting by adding a fixed value earlier or later, there isn't really any "optimization" because you might perform better in one operating point, but worse in others.

c) The timing chain (ZKD) doesn't like it when there's a noticeable change in valve lift (SB) at partial load, but it really hates it at full load. What do you actually save if you can tear out the timing chain after a bit of full throttle and 40,000 km? The chickens would laugh. I just say, being cheap is great.

In my opinion, the only meaningful cost-saving measure that can be implemented with DIY methods is to simply rotate the injection pump so that, for example, in the case of an AFN engine, it injects fuel at 3° BTDC (Before Top Dead Center) when idling. While this may sound suboptimal, nothing should break as a result.
"This first compensates for the delayed partial load adjustment, and second, nothing can break down under full load because the specified values are maintained in that operating mode."

Best regards, Rainer.

I agree with Rainer on this point, but with the caveat that the inspector will notice the increased performance, and without any record of it, the higher power output might need to be temporarily switchable in order to pass the technical inspection (TÜV).
But that's a different matter.

Hello,

On request, we can also provide zero additional performance at any operating point.

Tampering with the wiring harness and deliberately altering temperature readings are just as prohibited as manipulating a few bits in the control unit. In both cases, you're driving illegally because the vehicle no longer conforms to its original specifications.

Best regards, Rainer.

ulf wrote:

Suppose that an additional series resistor of 470 ohms corresponds to a temperature of 40°C when the sensor is functioning correctly (values chosen arbitrarily). Then, the faked temperature can never be higher than this 40°C: Even if the sensor only had 20 ohms at a real temperature of 150°C - because the MSG would then see the sensor resistance as 470 + 20 = 490 ohms, which would likely correspond to 39°C.

Regarding the theory:
I looked at the G62 characteristic curve, and with a real temperature of 100°C, it would then be 55°C at +500 Ohms.
At a real temperature of 20°C, the perceived or "fake" temperature would only be 15°C.
So, the resistance would only gradually increase its effect when the engine is warm.
So, if the engine overheated, the EDC would still believe that the engine is running at approximately 60°C. The question is: Does the EDC do anything when the engine overheats? Does it continue to operate the diesel pump, or does the engine go into a limp-home mode? If not, then it wouldn't matter what the EDC thinks when the coolant is boiling; in that case, the important things would be the temperature gauge and the warning light.
According to the T4 wiki, if the G62 fails, the EDC (Engine Control Unit) continues operation with a pre-set value of approximately 5°C for the coolant temperature. If it actually does that, then the advance in injection timing cannot be very significant, otherwise it would damage the engine.

Rainer K. wrote:

Hi, Oh my goodness, all the things people come up with. a) The advance of the injection timing varies depending on the engine and can cause other unwanted "side effects," especially at low coolant temperatures. Reducing fuel consumption is more likely to be coincidental in individual cases.

"However, in this forum, some users describe pre-injecting the fuel at the beginning of the injection cycle using an NHG (presumably a specific type of injector) with a capacitor, and they are happy with it. A slightly higher power output is described here, not a huge amount, but still noticeable. The logical conclusion would be reduced fuel consumption, as you might need to press the accelerator pedal a little less to achieve the same driving experience."

Rainer K. wrote:

b) When adapting by adding a fixed value earlier or later, there isn't really any "optimization" because you might perform better in one operating point, but worse in others.

I think you're referring to the full load range. Yes, that makes sense to me as well.
The reasoning is as follows:
On average, I drive 95% highway miles using cruise control (partial load), and about 5% at full throttle.
Is it beneficial to compromise the performance at full load in order to improve performance at partial load?
Computationally, the answer is "yes" with a 95% probability.

Rainer K. wrote:

c) The timing chain (ZKD) doesn't like it when a noticeable change in valve lift (SB) at partial load results in something completely different at full load. What do you actually save if you can tear out the timing chain after a bit of full throttle and 40,000 km? That's just ridiculous. I'm just saying, being cheap is great.

That, in turn, completely reverses the situation to a resounding "NO" .
Do you think the SDI will still be able to handle the SB modification, considering it's actually designed for higher cylinder pressures? The SDI has the same cylinder head as some TDIs, but without turbocharging.
Theoretically, this lack of load could be compensated for in the SDI system by adjusting the setpoint, which would have a positive impact on consumption/performance.

Rainer K. wrote:

In my opinion, the only meaningful cost-saving measure that can be implemented with DIY methods is to simply rotate the injection pump so that, for example, in the case of an AFN engine, it injects 3° BTDC (Before Top Dead Center) at idle. While this may sound suboptimal, nothing should break as a result. "This first compensates for the delayed partial load adjustment, and second, nothing can break down under full load because the specified values are maintained in that operating mode."

Okay, I thought the NHG (National Hot Gas) system would compensate for this. What's the difference between reversing the polarity of the diesel pump's KWvOT (Kaltwasser-Warmwasser-Verhältnis-Optimierer) and inserting an NHG capacitor in between? Wouldn't reversing the polarity of the diesel pump be much simpler?

Hi,

Quote:

Ah, I thought the NHG would compensate for this again. What is the difference between reversing the polarity of the diesel pump (KWvOT) and inserting an NHG capacitor in between? Wouldn't reversing the polarity of the diesel pump be much simpler?

With an NHG capacitor, the entire frequency range is shifted.

The control unit regulates the output depending on the load, from 14 kW input to 2 kW output.
Now, it can be advanced by up to 5 weeks if you push it aggressively, meaning from 19 weeks gestational age (wGA) to 3 wGA.

Pump speed adjustment shifts the possible control range, for example, from 14° KW above opening to 2° KW below opening, to a new range, such as 19° KW above opening to 3° KW below opening.
The control unit then adjusts the timing, not controls it, from a maximum of late timing (currently 3° KWvOT) to the standard target value: from 14° KWvOT to 3° KWvOT.

Best regards, Rainer.

Also, egal ob man Pumpe verdreht oder NHG-Kondensator nimmt, es entsteht in unserem Beispiel immer 19°KWvOT bis 3°KWvOT, nur bei der verdrehten Pumpe kann auf 14°KWvOT bis 3°KWvOT im oberen Lastbereich gegengeregelt werden, beim Kondensator nicht. Stimmts?
So, the variable displacement pump is gentler in the upper load range and still provides benefits at partial loads.

I assume that the coolant temperature sensor signal also globally adjusts the injection start timing across the entire operating range, which is not ideal.

Best regards,
Pappl