Blended Alpha-n Question

[73Inka2002]

Hi everyone,

I'm having some trouble understanding the relationship between the two VE tables in a blended alpha-n setup. I guess I'm thinking that they sort of lay over one another and in the places where the values overlap (at idle for example) then you would put a value of "100" in one of the table so the multiplied value would be x*100.

Would someone mind posting their blended alpha-n MSQ with a brief description of what I'm looking at? I think this would be the quickest way for me to gain a better understanding.

Thanks,
Robert

[muythaibxr]

It depends on whether you have "multiply" or "additive" set up.

Basically, the way I've set it up in the past is to have a spot in the alpha-n table where I want the engine to run completely speed density... that spot is set to 100%, and the blend is set to multiplicitive.... then in the speed density table I have it similar, a spot that's at 100%.


Basically I end up tuning an alpha-n table and SD table separately, then above 85 kPa I make VE 100% in the SD table, and below about 5% in the alpha-n table I make VE 100% (meaning it'll just end up with the value in the other table).

This worked pretty well for me.

Ken

[shauer]

I'm running blended tables with my "prmary" table being SD and my second table being AN. I set the blending option to "multiply".

My SD table goes from idle up to 90% load.

I plotted my throttle position vs. RPM for 90% load and set all bins below the 90% load line on my AN table to 100% so that the AN table has not effect on mixture below 90% load. Above 90% load the AN table takes over multiplying the 90% load bins on the SD table by the value in the AN table.

I think I did it the "opposite" way that Ken did. I was running my ITBs with SD prior to the extra code so that was my starting point. The one thing I might change is setting the cross-over point a little lower, like maybe 85% load.

Here are my two tables. Table1 is SD and table2 is AN. These are a bit rough but they illustrate the concept. The values look a little odd because I'm running an AFR of about 16:1 at cruise and 12.5:1 at WOT.

[muythaibxr]

Actually, this is almost exactly what I was talking about doing, except I had a 100% bin in the SD table, and filled it across the board with 100's.

It also looks like you have "multiply MAP" turned off.

Ken

[73Inka2002]

Thanks guys,

This info is very helpful. I'm going to need to stare at it for a while to understand what's happening. For some reason, it's still not clicking in my head.

I plotted my throttle position vs. RPM for 90% load and set all bins below the 90% load line on my AN table to 100% so that the AN table has not effect on mixture below 90% load. Above 90% load the AN table takes over multiplying the 90% load bins on the SD table by the value in the AN table.

Shauer, how did you know where to put the 100% values? What does "I plotted my throttle position vs. RPM for 90% load..." mean? How did you complete that process? Did you look for something in the logs?

Thanks again guys and sorry to be dense.
Robert

Edit: Removed a point that I was confused on.

[shauer]

Yes, I have "multiply map" turned off on my .msq That is why my SD values are so small.

To plot the 90% load point I used logs from several runs. I looked for places where my throttle and load were stable for at least a couple seconds and where the load was at 90%. I recorded the TP and the RPM for that point. I located TP values for as many RPM values as I could find in the logs. This provided me with about a half dozen TP vs. RPM points where the load was 90%. I used Excel to create a simple plot of this using TP on the Y axis and RPM on the X. This gave me a line which I can use to look up the TP for any RPM where the load will be 90%.

If you look at my AN table you see a more-or-less triangular area where the bins are all 100%. The boundry between the 100% bins and the bins greater than 100% is where this line ended up for my engine.

I actually repeated this process for several load points so that I have TP plotted against RPM for 70%, 80%, 90% and 100% load.

The 100% load line is also important for AN. If you look at my table2 you will see that after the throttle opens to a certain point the values in the bins stop increasing. This is the 100% load line.

I found it was easiest to tune AN if I first got a rough tune using SD only and then collected a lot of log files so that I could create the TP vs RPM load chart.

I hope this helps. It's kind of tough to describe in words. I'll see if I can post a picture of this chart.

[73Inka2002]

OK, I've taken a few log files and extracted all the 85 kpa values and then created a scatter plot with the results. I drew a quick line through the points on the graph. So now I assume anything to the upper left of the line will be my usable alpha-n values and anything to the lower right of the line will be 100s on my AN table. Does that sound about right?

I guess I also need to chop anything on my SD table above 85kpa and/or make those values 100s?

Thanks again,
Robert

[shauer]

That plot looks about right. Looks like you have pretty large throttle bodies on your engine? I'm assuming this is for a BMW M10 engine? Are you running 45mm ITBs?

Yes, everything to the right and below the line would be 100% on the AN table and you are tuning the region above and to the left of that line for AN. You should also plot a similar line for the 100% load points. This will give you a triangular region on the AN table where the real tuning will be done. Everything above the 100% load line should be set to the same values. Refer to my AN table to see what I'm talking about.

For example, at 2000 RPM you need 8% throttle to achieve 85% load. I'm guessing that by 12%-15% you are at 100% load and any additional throttle above that point should not provide any increase in fuel since your engine cannot draw any more air beyond the 100% load point. At higher RPMs you need more throttle between the 85% and 100% points. I'm guessing you will need something like 60% - 70% throttle at 6000 RPM based on your plot to get 100% load.

As for your SD table, I would set the top row at 85% load and do not define anything above this point. MS will then use the values in the 85% load row for all loads above 85%. The multiplier in the AN table will be applied to the SD value to calculate the fuel needed.

[73Inka2002]

Thank you again Steve, you and Ken have been a huge help.

Yes, it's an M10 with GSX-R throttle bodies which start at 48mm and then taper down to 40mm. They actually work surprisingly well. I just cant help being curious which is why I want to try AN.

I think I'm good for the moment. I've started creating my AN table and so I should be able to see if this all works sometime over the long weekend.

Oh, one final question... what are the advantages/disadvantages of the "multiply map" setting?

Thanks,
Robert

[shauer]

Oh, one final question... what are the advantages/disadvantages of the "multiply map" setting?

I'll Let Ken address the multiply map setting. I'm still a little fuzzy about it. If I remember correctly disabling this feature provides a more stable tune at low loads when using SD. If you disable multiply map you will need to retune your SD table. If I remember correctly I had to cut my SD values almost in half when I disabled it.

I think you will like the results of the blended tables on ITBs. I found it to be the single biggest improvement on my car. It takes a while to get used to thinking about two tables interacting with each other and MegaLogger gets a little confused on the blended tables and you cannot really use the autotune features of MegaLogger.

I'm using 45mm ITBs from TWM and I'm thinking of going down to 40mm ITBs if I can afford it this winter. I use my car as a good weather daily driver and the extra code has really helped the driveability but 45mm ITBs are just too big for a moderately modifed M10 engine unless you are racing it.

[6040solder]

I'm using 45mm ITBs from TWM and I'm thinking of going down to 40mm ITBs if I can afford it this winter. I use my car as a good weather daily driver and the extra code has really helped the driveability but 45mm ITBs are just too big for a moderately modifed M10 engine unless you are racing it.

according to this :

http://en.wikipedia.org/wiki/BMW_M10

the m10 is a 1766cc 4 cylinder

the toyota 4age successfully runs map sensor on the later model 20v and larger than 45mm itbs, and it is 1600cc.

perhaps you have more fiddling to do?

Silvertop 4A-GE ITB's have a 45mm opening tappered into a 42mm throttle plate.
Blacktop 4A-GE ITB's have a 48mm opening tappered into a 45mm throttle plate.

reference :

http://forums.vwvortex.com/zerothread?i ... d=36139650

i dont know how accurate it is, but i have measured them both myself some time ago, and forgotten.

[muythaibxr]

Oh, one final question... what are the advantages/disadvantages of the "multiply map" setting?

I'll Let Ken address the multiply map setting. I'm still a little fuzzy about it. If I remember correctly disabling this feature provides a more stable tune at low loads when using SD. If you disable multiply map you will need to retune your SD table. If I remember correctly I had to cut my SD values almost in half when I disabled it.

The main advantages I've seen from turning off multiply map are:

1) VE changes of 1 point make much finer changes in pulse-width
2) MAP jitter has a much smaller effect on pulse-width.

Ken

[shauer]

the toyota 4age successfully runs map sensor on the later model 20v and larger than 45mm itbs, and it is 1600cc.

perhaps you have more fiddling to do?

You are comparing engines with about 30 years of development separating them. The BMW M10 is a SOHC 2 valve per cylinder design based on 1960s technology. Very good in its day but it flows a lot less air than a modern 4 or 5 valve per cylinder engine. You need to look at other factors than just the engine displacement. My engine is just over 2.0L and its realistic limit with the mods I've done is around 150HP. (stock was about 110 HP). Anybody with a modern DOHC engine would laugh at these figures but it is considered pretty good for this engine.

If you only need 80% throttle to achieve 100% load all the way up to and beyond red-line it is a good indicator that your TBs are too big. I'm also not interested in running my engine at 6500-7000 RPM so I am willing to trade off some minor losses at redline to improve the low-end throttle response.

[6040solder]

You are correct, 40mm should be adequate for 150hp of engine, however that was not my point.

The toyota 1.6 spinning at 1000rpm at WOT can not be swallowing more air than your 2.0 m10 at 1000rpm, and still has perfectly crisp throttle response with the oem setup. At that point, if you are having drivability issues with it with 45mm throttles, then either the tune, or the ecu firmware leave something to be desired.

If you disagree for the 1000rpm benchmark, then compare the 1.6 at 1000rpm to the 2.0 at 1500rpm to make things more fair.

The earlier silvertop engine had an airflow meter, but the later one (which I have put many miles on) was MAP and tps based in some fashion. The only advantage at low rpm that the toyota could have is in the ecu. that ecu is now 11 years old in its design.

I'm sorry if I offended you in any way, that certainly was not my intent. If there are flaws in my thinking, then please point them out, as I am always happy to learn :-)

[shauer]

No offense taken.

The driveability issues I was trying to descibe are not related to throttle response or tune-ability of the large ITBs. In fact, this is the reason I converted over to the extra code. The work that has been put into the blended table support, the EAE, and the map delay allow Tbs that are too large for a given engine to still be tuned correctly. I have managed to get my engine running with scary-fast throttle response and I can run up through the gears at WOT and my AFR only drifts maybe +/- 1 from my setpoint. I was able to remove my vacuum reservior as well as my vacuum restrictor from my map line with the map delay feature as well.

The issues I'm talking about are related to the trade off associated with TB size. When I did the work on setting up my blended tables and plotted the effective dynamic range of my ITBs against the MAP signal (engine load) I found that after the TP hit 80% there was no benefit to opening them further. In fact, I reset my throttle linkage to max out at 80% open and re-calibrated my TPS to reset the 80% mechanical position to be 100% open on the TP reading. My engine performance improved noticeably. This is due to the increase in air velocity improving the fuel mixing with the TBs only 80% open at max throttle. This change did not cause any drop in the MAP signal at redline, it was still at 100%. If I keep the throttle at about 50% until I get over 4000RPM my mid-range torque is also significantly improved again with no drop in MAP signal.

With 45mm ITBs my idle MAP is at about 50% and my performance is helped by preventing the TBs from going more than 80% open. No matter how well I can tune them for outstanding throttle response I will not get around the issues that the TBs flow much more air than my engine can ever use. This hurts engine performance because the air velocity through the TB is reduced due to the large diameter which hurts the air/fuel mixing. EFI may not be as sensitive to air velocity through the intake as a carb is, but it is still a factor.

Another more practical problem is a very touchy gas pedal. You spend almost all your time driving in the first 10% of throttle position. This can be worked around somewhat with a progressive linkage and a gentle foot but if it can be improved with a smaller TB without hurting performance then why not do it?

So to try to summarize all this, I found that even though you can tune ITBs that are too large for a given engine and get outstanding response from them via the features of MSII-extra, you are leaving HP on the table, especially in the mid-range where the air velocity through the TB is too low.

Robert, sorry for hijacking your thread.

[73Inka2002]

No problem at all, Steve. My question has already been largely answered. Plus, I think the discussion is interesting.

I'll add one observation on my engine. I noticed that with these ITBs, my WOT MAP starts to falloff towards redline. (about 1 or 2 kpa at redline - it goes from 100 kpa to about 99 or 98 kpa) I assume that's because there's a pressure drop caused by the ITBs. This would lead me to believe that they may be slightly undersized for my engine in its current state of tune. I've got a 292 deg cam, mild head work, header, etc. I'm thinking that a straight 45mm ITB might be more appropriate for my engine. Of course then the issue with direability, MAP shooting straight to 100 kpa, etc only becomes more significant. Anyway, it's a fun problem to try to solve.

Robert

[6040solder]

Ok Steve, I see where you are coming from now.

The single valve and its throat size compared to the throttle and the tapered tube required to flare out to it makes the throttle too big for the max flow you are experiencing with your particular cams at your particular rpm. makes good sense. So in order to up the midrange torque you need to down the throat size.

I would imagine the throttles on the 4age are progressive, but i cant remember. Ken may be able to answer that as he has a set himself.

I have to say though, the amount of time I spent at or below 10% throttle was minimum :-) with itbs, its hard not to have a lead foot :-)

Thanks for taking the time to explain it all to me.

Robert, have a look at this :

http://www.msruns.com/viewtopic.php?t=10169

his dyno results show little loss at lower revs and a significant gain at higher revs. interesting anyway if not directly applicable.

[shauer]

I have to say though, the amount of time I spent at or below 10% throttle was minimum with itbs, its hard not to have a lead foot

I completely agree, unfortunately there always seems to be somebody in front of me and they are bigger than I am. It's when I'm in traffic that I curse the digital nature of the throttle with the ITBs but they sound sooo nice.

I also agree this is a fun problem to try to solve. I'm trying to achieve daily driver behavior with the throttle response (and sound) of ITBs. It's a tough set of compromises to balance.

[6040solder]

I have to admit that it wasn't the easiest car to drive smoothly, but I put most of it down to loose engine mounts and the factory light flywheel. I could control the revs in neutral pretty accurately and never had an issue cruising or start stop driving in it. It must have had progressive throttles I guess.

Those corollas have a pressed steel "plenum" around the factory plastic or rubber trumpets, it doesn't insulate sound very well at all. Even dead stock with all pipes etc in place it was never a quiet car either. Certainly wasn't quiet with the plenum off :-)

[RdSnake]

Ive been running MS-II/EDIS/ ITBs in my miata and I must say that Ive been having a lot of fun with it. I just got back from a track day over the weekend and the car was a blast to drive to 8000rpm in contrast to the factory set 7200rpm rev limiter.
I am looking forward to trying out hybrid alpha-n and maybe gain something. Im pretty happy with how the car drives using SD. The car was originally tuned in alpha n and that didnt work too well.
Im hoping to gain snappier throttle response with hybrid alpha n.
On the VE tables above, I noticed that one of them is VE 2. I thought the transition table was VE 3 because VE 2 is reserved for the 2nd cylinder bank in case youre running 2 tables. Also, how is spark controlled? Would spark still be Speed Density based or theres also some sort of blending between SD and AN for ignition? TIA