Aftermarket VS motronic management

[pnanassy]

And what if my aftermarket standalone can handle the über mega hyper accurate MAF?

[Billman]

i think vems can also run with maf but almost noone uses it.

[A80Avant]

i think so too, which is something i may consider doing in the future as well

[prj]

MAF is nice to use on a speed-density ECU as a secondary input for the ECU to learn VE values.
You run the ECU in Mass-flow mode and let it learn the VE, because you can calculate it from the MAF, MAP and IAT signals.
Once the ECU has learned VE, the Speed-density calculation will be (with a few possible minor issues) match the MAF reading.
After this the MAF can be removed and the system switched to Speed-density.

This approach takes away the guesswork about engine VE.
If the the engine VE is changed, one can again temporarily attach the MAF and let it re-learn it, so one can again precisely determine cylinder filling.

[Ducky]

Also, I tune my friends cars for free....

Can I be your friend?

[A80Avant]

hahaha!! depends! if you pay for travel i assume!

[quattro84]

Nice read.

I work as an calibration engineer on a new Euro6 engine and i can confirm PRJ's findings.
We used speed-density method in the past and that clearly isn't accurate or redundant enough to predict airflow for passing future emission legislation.
In stead we use a reference AMF sensor in the EGR system and by mapping the volumetric efficiency the level of EGR is controlled close loop by an EGR valve based on NOx model corrected by NOx sensor. (i must clarify that delta p over the engine is controlled by a VGT turbocharger to increase EGR flow. VGT position is pressure regulated as well)
Before using the AMF sensor triple p+T measurements were used to predict the EGR flow but weren't exact enough...

I agree that MAF/AMF is the way to go for current emission legislation or the odd DIY control freak. But i reckon it isn't used on aftermarket systems such as VEMS due to complexity of the system, needing a lot of inputs which need to be calibrated before used in an closed loop, which most of us here don't want because they drive a lot of transcient in stead of steady state...

[prj]

Very interesting read, thanks.

[S2central.net]

Now this thread is getting somewhere. The MAF does a great job of measuring how much air is being drawn in and so gives automatic adjustment for engine mods such as exhaust, cams etc whereas on speed density the VE data would need to be changed. Dmitris idea of using maf as a means to measure VE is sound too - all that is needed is a reference MAF. Until fairly recently with devices such as the Bosch HFM5 that has publically available data it has been very difficult to get such info. I know that some of the aftermarket US V8 folk have spent time and money producing their own calibration data for MAF devices - needs a hefty flowbench.

Anyway - IMHO the only advantage for speed density systems is the lack of a restriction in the air path - a hybrid ecu that can be calibrated with maf in order to create the right VE table to use in normal operation is the best of both worlds.

[quattro84]

Yes, big advantage is that the MAF really measures, as where the v/ρ method uses an VE from a look-up table, big inaccuracy! VE's aren't stable over engine lifespan so you can see where i'm going at.
Agreed, the MAF is the biggest restriction, but incorporate a second one to lower the pressure drop over the intake piping.


Biggest problem of the MAF i've seen is under transcient conditions, the system delay due to the constantly changing airflow. There's where the v/ρ method comes in action...
And AMF/MAF sensors need a very descent cleaning strategy to stay accurate over time. Early audi MAFS for the B5/C5 platform are well known to getting an offset over time, especially when used with K&N preoiled filters...

[prj]

Yeah, the IAT sensor can accumulate some amount of oil and grease on it as well.
This throws off fuel enrichment and in case of speed-density the cylinder charge calculation as well, as the MAF usually is temperature compensated.

[S2central.net]

I think the hot-wire MAF has inherent correction for air temp and pressure as the system is measuring how much current is needed to keep the wire at the same temperature. I'd need to hit the books to remember if the hotfilm MAF has the same automatic behaviour - I suspect not as I've seen some HFM5 using an auxiliary thermistor to get a view of iat being inhaled. I was reminded if this a couple days ago when servicing the V6TDi.

Motronic uses separate sensors for iat and altitude to make other ambient related corrections though - it doesnt rely purely on the Volts coming from MAF.

I would say that WBO2 would be a nice upgrade on MAF/hfm systems to compensate for component tolerances, the effects of contamination and ageing drift.

WBO2 is a must have aspect on speed density due to the fact that all calcs are based on an assumption of the engines VE and that changing a camshaft could significantly change it... WB feedback against target AFR would correct this on with some limitations but only in relation to fuel.

[Thuppu]

Also, I tune my friends cars for free. Always have.

Damn, I assume only adding you to the friend list doesn't fill the criterions of a friend..?

[A80Avant]

Then how can you get the best of both worlds using both MAF and speed density systems, IF:

1)MAF although more accurate, it is worse on transient states as it is not quick enough to follow the change in airflow
2)Speed density is not as good in steady states as it is inaccurate in keeping the correct values, since VE is calculated, not measured?

Here is what i found written over at the vems site regarding MAF resolution. It may be wrong, but its getting too complicated in airflow dynamics and electrical wizardry to understand.

'Resolution is quite obvious:

with Speed-Density the injector pulsewidth is almost proportional to MAP (not completely, because of VE, lambda and injopen).

With MAF you divide measured airflow with RPM (than consider lambda and injopen), so you need very precise path for the analog signals to maintain a given resolution (or will have inferior resolution with similar technology). The box-box signals are often nonlinear (but close to logartihmic), so not much precision is lost there (that's just the signal transfer!), but still much precision is lost during the measurement (very precise and sensitive circuits come up with a relatively unprecise output because of the nature of the measurement). The division by RPM means that it is common to have unprecise result at lowRPM and rail out before max power.

Reliability is simply the result of complexity'

As it stands as far as i know, you can use MAP, or TPS, or MAF with VEMS so long as the latter is calibrated as mentioned previously. However i cannot see whether its allowed to be able to use both MAP and MAF at the same time.

[prj]

With MAF you divide measured airflow with RPM (than consider lambda and injopen), so you need very precise path for the analog signals to maintain a given resolution (or will have inferior resolution with similar technology). The box-box signals are often nonlinear (but close to logartihmic), so not much precision is lost there (that's just the signal transfer!), but still much precision is lost during the measurement (very precise and sensitive circuits come up with a relatively unprecise output because of the nature of the measurement). The division by RPM means that it is common to have unprecise result at lowRPM and rail out before max power.

This is exactly the reason why Motronic uses a 16 bit counter running on a RPM interrupt and not an 8 bit ADC for the MAF. Due to this, you don't have to do expensive calculations to calculate LOAD.
Also, MAF's are more sensitive to airflow changes at the lower end and the middle and get less sensitive in the top end.
This is why a curve is fitted in a special circuit, which is inversely proportional to the MAF curve. That way a fairly linear response is achieved.

Most of that quote is only true for poorly engineered systems. The MAF is precise enough on Motronic to just put in target lambda, and as long as you don't max it out, you can alter engine VE greatly, and it'll still reliably get you to said lambda.
Of course, as computing power increases, so do the calculations - ME7.1 is taking MAP in account for calculating secondary load and sanity checks, etc.
However - where that quote is true, it does take complexity to make it work well.