ECMlink DSMLink : MAFRaw vs. MAFComp

[oldman]

Hello children...

I posted this on Link but didn't really get an answer so let's see if I have better luck here. I'm in the process of setting my VE table and discovered that my MAFComp is a lot higher than my MAFRaw values with zero adjustment to my MAF table, is this normal? Shouldn't the two values be identical with zero adjustments?

 

[hopwoodp]

Hello children...

I posted this on Link but didn't really get an answer so let's see if I have better luck here. I'm in the process of setting my VE table and discovered that my MAFComp is a lot higher than my MAFRaw values with zero adjustment to my MAF table, is this normal? Shouldn't the two values be identical with zero adjustments?

Here is some stuff I found on dsmlink. It may help answer your question. I saw your post on there but, saw you didn't really get your answer. Let me know if this help.

http://www.dsmlink.com/forums/showthread.php?t=15938&highlight=mafcomp+mafraw

OK, a few things.
Fake MAF + 0 Hz Switchover point =

You need to understand how the Fake MAF function works. Once you check that box, the ECU will completely ignore MAFraw after the switchover point. Instead of using MAFraw, it will calculate the MAF Hz signal by using RPM and the last boost level that the ECU saw just before you past the switchover.

But in your case, your switchover was 0 Hz, so ECU had nothing to calculate Hz with. The way you pick your switchover point is by looking at your GM Bar and locating where you achieve a stable boost level. Then look at MAFraw at that point and that will be your switchover point.

In your other post, you were stating that the highest level you could place the switchover was 3188 Hz. I would use that and see how that affects your PLX data.

I still personally think that there is a fuel delivery problem somewhere there, since DSMlink calculated AFR goes down while PLX goes up.

EDIT : MAFraw will not change at all, even though you have Fake MAF checked. You will need to datalog the other MAF values like "MAF", "MAFcomp" so that you can see how Fake MAF alters your airflow.

e Isiaho

Join Date: Feb 2005
Location: Springfield, MO.
Posts: 2,328
diambo4life is on a distinguished road
Quote:
Originally Posted by NOSLO2PT0
Mike,
I'm starting to wonder if there might be a problem with the chip. Airflow was ok, if not just a hair low, but MAFRAW was way off. I will try the FAKE MAF option with a lower airflow reading. Now I just need to figure out why my estimated a/f ratio goes one way while my actual a/f ratio goes the other way, totally opposite.
You have to remember that "MAFRAW" is a raw value and isn't used directly by the ECU to calculate fueling. MAF value is and it's not being displayed in your logs. As JT mentioned, the ECU monitors the behavior of the MAF signal (MAFRAW) before the clamp point or the boost signal (if MAP is enabled and a boost sensor is logged) to extrapolate airflow beyond the switchover. That's why it's important to set a suitable switchover point where the boost/airflow is stable otherwise the extrapolation will be off.

Are you running a full -8AN fuel line?


Another update! I went out and did another pull tonight. Everything except all the knock I had looked promising, in certain ways. I used the FAKE MAF feature, had it switching over at 2900Hz. Airflow was actually up in the 60-62lb/min range, right where it should be. However, HP was WAY down, due to slightly uphill road and a fair amt. of knock. I did one pull to check AFR's, but didn't pull out any timing. I also logged MAF and MAFComp. Can someone explain what it all means.

39.11s
MAFRAW = 3152Hz
MAF = 4404Hz
MAFComp = 2856Hz

I selected the USE MAP feature, and by doing so, my log looks like what a normal log should look like. My est. A/F actually follows my true A/F!!! Jeff, here is where you come in. Take a look at what my fuel is doing. A/F still increases, but not as much. However, look at the damn IDC's!!!!!! I'm upwards of 121%, but A/F is still at a high of 11.5:1 (subtract .5 from logged values to get actual). I gotta be outta fuel pump Jeff.

Although I was knocking and the log didn't show me making much power, and the fact that my head is most likely warped (or I have a blown HG), the car overheating, I'm kinda happy that I finally am able to get a normal looking log and see what the hell is actually going on.

You fuel experts...You think I'm out of pump? Mike, I know you don't think so, but take a look at these logs man.


Here is another one:

They will not necessarily follow each other at all. MafComp is the MafRaw value after the airflow slider adjustments are applied. MafSD is the MafComp value after VE slider adjustments are applied.

Going from memory, at 100% VE, the sliders should overlap; at > 100% VE the MafSD value should be higher than the MafComp value; at < 100% VE the MafSD value will be lower than MafComp.
__________________

Here is one more:


Quote:
Originally Posted by flybyuin2.2
I have installed a Evo 8 MAF and now the MAFRaw seems to only read up to 1500hz where with the 2g mas it was reading 2000hz
Right, that's what's supposed to happen. For the same airflow, the EVO8 MAF will spit out a different frequency than the 2G MAF will. That's why you had to enter those compensation values into the MAF dialog. If you want to see what the MAF reading is after the compensation is applied, datalog MAFComp instead. MAFRaw will show you exactly what's coming into the ECU, straight from the MAF.

Thomas Dorris

 

[scottsee]

That is werid. I really don't have an answer for you..

Have you tired a diffrent MAF? I notice your airflow is off a bit even though your on the Stock MAF. click use MAP and Log the diffrences between MAF and MAFraw. Your log shows BoostEst at 15psi at 5500rpm's while your 3bar messured 17.4psi. Thats 8% airflow deviation.. (14.7 + MAP) / (14.7 + BoostEst) - 1
Your AirflowPerRev and MAFraw dosen't look right either. I wouldn't expect 2000hz at 2.4g/rev and 37lb/min airflow, your g/rev should be closer to 3. I would check for a leak.

WTF is going on with your Fuel Delivery? Your fuel compensation tables and A/F est are MASSIVLY rich while the Front 02 sensor shows that your engine is about to explode. Can you get a Wideband on there to see whats going on?

 

[oldman]

David, thanks for the link. I understand how everything work, I'm actually not overrunning my 2G MAS just yet. My question is very simple and I suspect Dave didn't answer my question on Link because he doesn't know the answer. MAFComp is suppose to be values after the compensation on the MAF is applied, if no compensation is made on the table, Comp should be identical to Raw but mine isn't. I want to know what's up before I start matching my MAFSD to MAFComp on the VE table. In the meantime I have disabled my fake MAF, we will see if I end up overrunning the MAF once I up the boost to 25+psi. Thanks again.

 

[scottsee]

Whats up with your Timing? There shouldn't be any reason you should be that flat with that kind of advance at those g/rev's...

Edit: what kind of g/rev airflow do you get at idle?

 

[oldman]

Have you tired a diffrent MAF? I notice your airflow is off a bit even though your on the Stock MAF. click use MAP and Log the diffrences between MAF and MAFraw. Your log shows BoostEst at 15psi at 5500rpm's while your 3bar messured 17.4psi. Thats 8% airflow deviation.. (14.7 + MAP) / (14.7 + BoostEst) - 1

Do you mean MAFComp vs MAFRaw? So you want me to click on use map although use fake maf is unchecked?

Your AirflowPerRev and MAFraw dosen't look right either. I wouldn't expect 2000hz at 2.4g/rev and 37lb/min airflow, your g/rev should be closer to 3. I would check for a leak.

The only leak I have is the BOV which doesn't start leaking until 17psi, that is why I set it to 17psi for building my VE table. I leave my bottom fitting (1G BOV DG modded) open except when I'm at the track because of the low throttle flutter.

WTF is going on with your Fuel Delivery? Your fuel compensation tables and A/F est are MASSIVLY rich while the Front 02 sensor shows that your engine is about to explode. Can you get a Wideband on there to see whats going on?

I have LC1 (can't log WB and map at the same time) and I'm running about 9.5 - 10.2 : 1 in that posted log, I have yet to lean it out but I just wanted to get my airflow side (VE) correct before I start the leaning process. I just got back from some tuning runs and I'm at about 10.5:1 across the board at about 22psi now, I will bring that down to about 11.2 : 1 @24ish psi tomorrow as long as I don't see knock.

 

[oldman]

g/rev = .25-.26. at idle.

Thanks for the help Scott.

 

[scottsee]

I'm at work so hold up 30mins k..

 

[oldman]

Take your time Scott, I'm done with test runs for the night anyway.

BTW do you think my MAFComp is much higher due to using a 2G mas on a 1G? In other words, is Link's 2G mas in a 1G compensation still present although it's hidden from the table itself?

 

[hopwoodp]

David, thanks for the link. I understand how everything work, I'm actually not overrunning my 2G MAS just yet. My question is very simple and I suspect Dave didn't answer my question on Link because he doesn't know the answer. MAFComp is suppose to be values after the compensation on the MAF is applied, if no compensation is made on the table, Comp should be identical to Raw but mine isn't. I want to know what's up before I start matching my MAFSD to MAFComp on the VE table. In the meantime I have disabled my fake MAF, we will see if I end up overrunning the MAF once I up the boost to 25+psi. Thanks again.

Oldman, I am glad I helped. I am kinda of familiar with this type of operation. But, unfortunatly it is on gas turbine engines. I know that with airflow and fuel maps you can relate the two. Let me know how it goes once you reach the 28 psi mark like you said on your other thread.

 

[scottsee]

Take your time Scott, I'm done with test runs for the night anyway.

BTW do you think my MAFComp is much higher due to using a 2G mas on a 1G? In other words, is Link's 1G mas compensation still present although it's hidden from the table itself?

Ohhhhhhh, I didn't catch that, absolutly! You need to calibrate your MAF using The methid decribed in the Wiki sections. Good call.

 

[oldman]

Ohhhhhhh, I didn't catch that, absolutly! You need to calibrate your MAF using The methid decribed in the Wiki sections. Good call.

Why? I would think that the 2G mas compensation is already done by link since 1G dsmlink has a 2G mas option. Are you saying the internal compensation done by Link could be off?

Oldman, I am glad I helped. I am kinda of familiar with this type of operation. But, unfortunatly it is on gas turbine engines. I know that with airflow and fuel maps you can relate the two. Let me know how it goes once you reach the 28 psi mark like you said on your other thread.

Will do David, thanks again.

 

[2gGSX]

Hello children...

That made me chuckle thinking of Chef

Why? I would think that the 2G mas compensation is already done by link since 1G dsmlink has a 2G mas option. Are you saying the internal compensation done by Link could be off?


Will do David, thanks again.

There are variabilities between MAS sensors, so it's not inconceivable that the default calibration might not work too well on your specific sensor.

 

[oldman]

There are variabilities between MAS sensors, so it's not inconceivable that the default calibration might not work too well on your specific sensor.

So I should still follow the MAFT calibration method on the Wiki site to get it fine tuned?

BTW do you know if my Comp and Raw difference is due to the default 2G mas compensation although no adjustments were made to the table itself?

 

[scottsee]

Why? I would think that the 2G mas compensation is already done by link since 1G dsmlink has a 2G mas option. Are you saying the internal compensation done by Link could be off?

Going back to the Boostest/MAP airflow inacuracy of 8% at 5500rpm i would say so. You have alot of stuff going on there with your timing, airflow, fuel. Have you compleatly tuned/set your closed loop idle/cruse operations?

 

[oldman]

Have you compleatly tuned/set your closed loop idle/cruse operations?

Yes, my ltft low and high is at about +.7% as of tonight although my global is about 4% lower than where it should be, maybe my fuel pressure gauge is off.

FIC 1000 @ 43.5psi
Sqrt (43.5/37) x 1000 = 1085cc
450/1085 - 1 = -58.5%

My current global is set to -54% with deadtime of 300, the injectors were already sent back to FIC for verification.

I will post a new log tomorrow after MAF calibration and getting my AFR to 11:1.

Thanks children.

 

[2gGSX]

Trivial since you have stable fuel trims, but FIC rates their injectors at 43.5 psi, so you don't have to do the fuel pressure calibration.

 

[oldman]

Trivial since you have stable fuel trims, but FIC rates their injectors at 43.5 psi, so you don't have to do the fuel pressure calibration.

With a 1G ECU you do, we'll see if things change after the maf calibration.

 

[oldman]

Ok just got done reading the calibration with map page. Considering the fact that my boostest at idle is lower (-23 inHG) than my actual (-16 inHG, map matches my boost gauge), should I change my global to -58% and start the entire process over OR should I keep my global where it is and just focus on the WOT part of the calibration? What are your MAFRaw's and gm/rev's at idle?

 

[2gGSX]

Ok just got done reading the calibration with map page. Considering the fact that my boostest at idle is lower (-23 inHG) than my actual (-16 inHG, map matches my boost gauge), should I change my global to -58% and start the entire process over OR should I keep my global where it is and just focus on the WOT part of the calibration? What are your MAFRaw's and gm/rev's at idle?

At idle, my old 2.0 was anywhere between .26 g/rev and .28 g/rev. I don't have any MAFRaw readings from then, but on my current 2.3 engine is seeing ~58-60 hz while idling at operating temps.

BoostEST shouldn't be used as a "gauge" per se--it's set up for certain variables, such as 100% VE, 70* intake temps (or something like that). Thus, it should only match boost at ~5500 RPM or wherever you have 100% VE.

If you have your fuel trims, throttle position, etc. all set up then there's no reason not to start WOT tuning.

 

[scottsee]

If you have your fuel trims, throttle position, etc. all set up then there's no reason not to start WOT tuning.

If your idle speed operations (g.rev,maf,TPSvolts,ISCPostion) are working correctly along with your closed loop fuel STFL and LTFT are within +/-5% at 50,150,250hz there isn't a need to change your global or dead times. Just work on calibrating your open loop fuel tables.

I know you don't seem to be concerned about this, but I still think there is somthing really wrong with your timing...

 

[oldman]

I'm going to do some experiementing after work today by changing my global to -58% and re-calibrate my 2G mas as if it's hacked, reason being I really think my MAFRaw (used 2G MAS) is way off even at idle (Raw = 38 Hz and Comp = 45 Hz) forcing me to go -54% (almost losing 100cc) in my global, I suspect the culprit is the mas to filter adapter. I have my current settings saved so I can always go back if need be.

BTW John (DSM90AWD) send me one of his logs and verified that his Comp is also much higher than his Raw with no adjustments on the maf table, this indicates that in fact the difference is definitely coming from the hidden default 2G mas in a 1G compensation.

 

[zippyshoe]

Yes, my ltft low and high is at about +.7% as of tonight although my global is about 4% lower than where it should be, maybe my fuel pressure gauge is off.

FIC 1000 @ 43.5psi
Sqrt (43.5/37) x 1000 = 1085cc
450/1085 - 1 = -58.5%

My current global is set to -54% with deadtime of 300, the injectors were already sent back to FIC for verification.

I will post a new log tomorrow after MAF calibration and getting my AFR to 11:1.

Thanks children.

I have an idea of what might be causing the global fuel discrepancy. Let's start with the factory 1G injectors which are rated at 450cc/min, which I am assuming is at a base pressure of 37psi (?). Now let's increase the base fuel pressure to 43.5psi - these factory injectors should now flow at:

450cc/min*Sqrt(43.5/37) = 487.93cc/min

Now let's swap-out these factory injectors for the FIC injectors rated at 1000cc/min at 43.5psi base fuel pressure. Calculating the global fuel value based on simply changing the injectors yields (Note: Both flow rates in the next formula are now at the same base fuel pressure):

(487.93/1000) - 1 = -51.2%

This would be the global fuel value that I arrived at for a 1G with an increased base fuel pressure of 43.5psi with 1000cc/min injectors installed. This appears to be value a bit closer to your actual global fuel setting. The one assumption I'm still questioning is whether or not the 1G factory injectors are actually flow-rated at 37psi base fuel pressure. Does my reasoning make sense to you, or am I making some invalid assumptions?

 

[eclipzed]

FYI your 450cc injectors are rated at 43.5 psi as 99% of all injectors are. Now since you are running a base pressure at 37psi, you are actually flowing LESS that 450's. Let me see if I can find that article that states this....

Also if my calculations are correct, 450cc injectors ran at 37psi (base pressure) are actually flowing at 415cc since the base pressure is lower.

 

[zippyshoe]

FYI your 450cc injectors are rated at 43.5 psi as 99% of all injectors are. Now since you are running a base pressure at 37psi, you are actually flowing LESS that 450's. Let me see if I can find that article that states this....

Ahh.. That's interesting. If that is the case, then the factory 1G injectors at 1G base fuel pressure actually flow at

450cc/min*Sqrt(37/43.5) = 415.02cc/min

I was wondering if this might be the case.

EDIT: I gotta learn to type faster.