Automatic LMM Calibration: Gremlin Solution

Hello Ralf,

If I remember correctly, you wrote something in several posts about monitoring the MAP (Manifold Absolute Pressure) voltage as an after-run check, in order to keep the measured values (or their interpretation by the engine computer) within the tolerances throughout the MAP's operating life.

However, general experience shows that LMMs (Linear Motor Modules) often need to be replaced due to performance degradation (especially if previous cleaning attempts have been unsuccessful).

Apparently, the aging of the mass airflow sensor is not completely compensated for, and the sensor value seems to decrease over the years.

I would like to develop an electronic solution to extend the operating time of the LMM (likely referring to a specific component) when the engine is running at full power.
I would need more detailed information about the calibration procedure.
What and how is being measured during the current inventory assessment, and what changes are made to the engine's operation if the results are altered?

oops, I completely overlooked that .


The EDC does the following:

The mass airflow sensor (MAF) is checked for drift in the after-run phase (i.e., after the engine has stopped). Normally, the measured MAF value should remain constant at this point.
A drift would indicate that, for example, the measurement area is unevenly heated, or that the Wheatstone bridge has a short circuit.

BUT: if the sensor surface is uniformly soiled (which is generally assumed), the measured value will not drift, or will drift only very slightly.


Unfortunately, in Europe, the actual monitoring of the LMM for dynamic drift is currently not possible.
because this monitoring is only required for OBD-II systems, specifically those in California.

This monitoring is achieved by constantly comparing the LMM (air mass flow meter) values with those of the AGR (air-fuel ratio) characteristic curve (expected and measured air mass). Based on the intake air and charge air temperatures, this comparison can be done within relatively narrow limits.
'Those who have thought about it will notice that with this monitoring system, it would no longer be possible to 'damage' the AGR (turbocharger). There would immediately be error messages and a fail-safe mode activated (which our colleagues in the USA have reported impressively). On the other hand, this could also be used to detect leaks in the charge air system (at some point, the measured air mass will become too large, long before the charge pressure collapses).'

Unfortunately, because this function is hidden in the software, the EDC now has no way to check the LMM values for plausibility, as it has no comparison possibilities other than the AGR map. Simply saving values is not enough without a second influencing factor.


Otherwise, only drastic deviations will be recognized as implausible. Since the HFM5 has a reverse flow detection, the output voltage is approximately 1V when the air mass is 0. Therefore, the mass airflow sensor (MAF) value must have a certain value in idle that cannot be significantly undershot. The same applies to excessively large MAF values, which can theoretically be calculated at any engine speed. However, this does not cover the case of a dirty MAF sensor, because even in idle, they can easily reach the normal values .

CU Gremlin.

CU Gremlin.

Gremlin wrote:

Oops, I completely missed that

.
Hi Ralf,
don't worry, it's not urgent .

Quote:

The mass airflow sensor (MAF) is checked for drift in the after-run (i.e., after the engine has stopped). Normally, the measured MAF value should remain constant at that point. A drift would mean that, for example, the measurement area is unevenly heated, or that the Wheatstone bridge has a short circuit. HOWEVER: if the sensor surface is uniformly soiled (which is generally assumed), the measured value will not drift or will only drift very slightly.

So, a defect is suspected only when there is a significant increase or decrease in the LMM voltage, despite the engine being stationary!?

Quote:

This monitoring is achieved by constantly relating the LMM values to those of the AGR characteristic curve (expected and measured air mass). Based on the outside air temperature and intake air temperature, this can be done within quite narrow limits. Those who have thought about it will notice that with this monitoring, you could no longer "break" the AGR (Agricultural Tractor Gearbox) because there would immediately be an error message and emergency mode. (which our colleagues from the USA have impressively reported).

It is my opinion that the (low-pass filtered) control signal for the EGR valve could also be used to modulate the MAF sensor signal "in phase."
Then, the engine computer receives the expected electrical response to actuating the EGR valve, namely a decreasing MAF sensor voltage - even though the "actual" EGR system might be disabled .
A corresponding electronic circuit could probably be designed relatively simply; at the very least, with a potentiometer to adjust the sensitivity, such a device should be functional.

Quote:

but unfortunately, that doesn't cover the dirty MAF sensor, because in idle, they easily reach the normal values

This is one of the mysteries: why do LMMs seem to spin up first at high mass flow rates ?

This is one of the mysteries: why do LMMs seem to spin up first at high mass flow rates ?


Why wouldn't you get burned by a soldering iron if you only touch it for a tenth of a second?

The sensor surface is heated to a constant temperature. The airflow passing over it cools the surface, and the degree of cooling is used to determine the air mass.

The layer of dirt now acts as a thermal resistance. If a small amount of air flows past, the amount of heat dissipated is also small. The temperature difference is low, and the thermal resistance of the dirt layer has only a minor effect.

When there is a large air mass, the sensor field loses a lot of heat. However, the thermal resistance of the dirt layer becomes proportionally much more significant, resulting in a large deviation.

It's similar to an electrical resistor (as an analogy). With a low current, there's hardly any voltage drop. With a high current, there's a significant drop.

Is everything alright?

CU Gremlin.

Gremlin wrote:

. . .when there is a large air mass, a lot of heat is drawn away from the sensor field. However, the thermal resistance of the dirt layer becomes proportionally much more significant, and the deviation becomes large. It's similar to an electrical resistor (as an analogy). With a low current, there's hardly any voltage drop. With a high current, there's a significant drop. Is everything clear?

Hi Ralf,

yes and no... my actual puzzle is why even cleaned LMMs eventually fail to reach the normal initial voltage - and this seems to happen first in the upper measurement range.

I was already aware of the "simple" effects of dirt layers, but you couldn't have known that.

yes/no... my actual puzzle is why even calibrated LMMs eventually fail to reach the normal initial voltage - and this seems to happen first in the upper measurement range.



Well... the sensor surface is only just under a micrometer thick... so any dirt diffuses right in, and brake cleaner sprays do more damage than anything else.

CU Gremlin.

Gremlin wrote:

Well... the sensor surface is only just under a micrometer thick... so any dirt diffuses right in, and brake cleaner sprays cause more damage than everything else combined...

Hi Ralf,

Wow, so it turns out that when cleaning with brake cleaner, you're essentially exorcising the devil with Beelzebub .

Are there any cleaning methods that are completely harmless for LMM ?