Help me understand my engine map (1.9L ALH) (Or I intend to do stupid things)

[bhodgkiss]

if you have a 228mm stock clutch in decent condition it should handle 350Nm which is 'approx' 53mg of fuel. Just as a guide. 45mg is not that much! in fact my stock AFN map was 42mg max and that was 110bhp. You sure its only 45mg now?
Update: Oh hold on your car was 90bhp stock? That would make more sense.

My AFN (very similar) with 51mg and stock nozzles was nowhere near slipping stock clutch. Now I have approx 55-56mg and in 3rd at 2000rpm its a bit too much. R520 nozzles in now and its transformed the car! now feels as torquey as a PD but over a wider range.

 

[baja_brewer]

I'll be reading this later

 

[Sbeghan]

Yup, 90 hp ALH. Wheeeeee. I'd love to put in some of the T4 1019 DLC nozzles (mostly since it has 350k miles on it and I think the nozzles could use a refresh as part of maintenance, haha) which should flow more fuel more easily.

So I am in the process of checking the logs now but it looks like there in some minor overboosting, after which it holds the boost at specified. I plan on dropping boost max to around 2100-2150 mbar from 2200-2300 mbar just because while fun I'm not that comfortable with where it is. I'll also have to edit the N75 map to slow down the boosting which should, hopefully, reduce the overboost.
On the other hand, I have an issue. Max fueling reported is 38 mg/stroke even though requested goes up to 46mg/stroke. I believe that the cause of this is that the MAF max requested that I can see is 850 mg/stroke whereas the actual MAF readings go up to around 1050 mg/stroke. So even though it should be injecting more fuel based on actual MAF readings its pulling it back due to lower requested MAF values. I think. I'll have to look at the maps again and think about this.

One of the annoying things is that the vag-com logs I'm getting are only once every 1.5 seconds because I'm logging 3 fields at once.

 

[bhodgkiss]

Yes log one at a time and just do 3 runs!
I think my MAF is maxing out somehow like yours. Hmmm

 

[Sbeghan]

Just to review...
Driver wish maxes out at 50 mg/stroke at full throttle (no that I expected to actually get that). I can see that request in VCDS.
Torque Limiter caps that to about 43-45 mg/stroke across the 2-4k rpm range. I can see that too.
Fuel quantity based on MAF is limited to 38/mg stroke and specified MAF caps at 850 mg/stroke.
The smoke limiter has 850mg/stroke limiting fueling to 42.5 mg/stroke (and it should cap off at 1000mg/stroke).
I notice the old smoke limiter (the one tha's labeled as the automatic that I didn't change) is limiting fueling to 38mg/stroke at 850 mg/stroke air, same limit that I'm currently hitting.

The SVBL and boost target map are good and working as expected.
Unless there is a hidden map I'm guessing that the manual and automatic smoke limiter maps are mislabeled and I need to swap them. Will try and report on them tomorrow.

 

[Sbeghan]

I copied my modified smoke map (manual) to the smoke map (automatic) and uploaded that to the car. No change. Max requested MAF is still 850mg/stroke and max fueling is still 37.5-38 mg/stroke. I did some more logging and can confirm that though requested MAF can be lower (375-360 is the lowest that I saw) the actual MAF rarely drops below 500 mg/stroke and seems to like to hover around 520 mg/stroke when coasting in gear with 0 boost. I think it would be safe to say that no boost request is necessary when requested map is under 500 mg/stroke (with a FAR of 20 that is 25mg/stroke of fuel!). With boosting to 18 psi (2300 mbar actual) you will get around 950-1000mg/stroke air measured with the MAF.


Here's the graph of the fueling. This is not a bad example of how the various limiters work. You have driver's wish, the torque limiter (which varies slightly with RPM), and the smoke limiter (I think that's it, it is labeled as I.Q. via air mass in VCDS). The actual is always going to be the lesser of those 3 requests.

So either there's a hidden map that I'm not seeing, or what I'm editing as a smoke map is not the actual smoke map that I'm supposed to be using. Grr.

 

[Sbeghan]

I think I found the hidden map in winols (its also sort of in vag suite, but its misaligned and uneditable). In winols its not identified as a map either and I don't know how to edit it. It seems to be another copy of the smoke map but with the extra 3 columns repeated like in the other ECU binary that I was looking at earlier in the thread.

Sent a support email to the vagsuite guys and am going to post on ecuconnections too for help. For now it looks like my smoke map is uneditable and until it can be solved I'll just revert back to my original map.

 

[nicklockard]

I'm not a tuner, but this is a really cool thread. One way I've heard to build boost faster is to have the ECU retard timing briefly on transient throttle inputs. (Or, advance less rapidly on rapid increase of TPS). Late injection = later combustion which builds up EMP and EGT (more drive pressure on turbine wheel for given amount of IQ).

 

[Digital Corpus]

subscribed as well. Interesting stuff

 

[Sbeghan]

No word on the map editing issue. I might just get another ECU that works and edit that instead. I did bump up the torque limiter to 38 mg/stroke which should give me a 5% increase in torque (that's the smoke map limit).

In the meantime, editing my SOI graphs back to normal has made my last CEL go away (intermittent start of injection adjustment code). I think that when things are wrong for a long time the ECU will edit the maps itself.

Currently tweaking my boost map to hug the smoke map more closely. While this will make the car a bit less responsive (since it won't have extra air going into it ALL the time it can't instantly add more fuel before hitting the smoke limiter) it should increase my mpgs by wasting less energy producing excess boost.

 

[Sbeghan]

I know you guys love big pictures so I have more for you. I modified my boost target map based on taking the smoke limiter map and working out what boost level would give me about 10-15% more air than the smoke limiter required. Yes, I know that this could conceivably make my car smokier but its currently smoke free and the stock fuel air ratio is something like 20-23.

Here is a stockish car taking a drive around the neighborhood. Actual injected fuel will be the lowest of these 3 quantities. Orange is the torque limiter and is rarely reached. Blue is driver's wish, its what my pedal says to do. Yellow is the smoke limiter, its what the MAF says the limit should be. Notice how it hovers around 27-28mg/stroke sometimes? That's when the car is idling. Notice how its never actually flat? The car requests boost even with 0 driver's wish at high rpms, or it requests boost even with very low driver's wish.. just... because? I think it may have something to do with balancing the EGR, but I can't say.

So why not tell it to not produce boost until it needs it? And then, not significantly more than it needs? Ok, let's try that.

Here we go. See the flat idle in the smoke limiter now? It seems to drop below that sometimes, not entirely sure why. Usually when I clutch in and shift it will drop.

I am hoping that by reducing unnecessary boost we'll increase mpgs.
I'm going to take a highway trip today; we'll see how it works out when I get back. I've been getting 43 mpg regularly but have recently added some grill blocks and replaced my thermostat so...

Oh yeah, it feels a touch less responsive, but the top end feels the same when I mash the pedal and it can build the boost.

 

[Digital Corpus]

It feels less responsive because you have that delay for the spooling of the turbo, which you already kinda figured. If, previously, you had boost but were off the throttle, and now you have it setup as the above, then technically, you have no benefit in fuel consumption.

 

[Sbeghan]

There is still benefit even if I'm off throttle since the turbo acts as an engine brake. This way the car slows down less. Like said, we will see what the mpgs are when I refuel.

 

[robnitro]

If you want to do a dynamic idle, and turn on the coolant GP's earlier, I found out how to do both with VAGEDCsuite!!!
It's verified as I have a light on my dash to indicate coolant GP's on (I have a switch/thermostat wired up- will be removed- woohoo).

I wasn't sure which maps are used so I edited all 2 of the idle maps (manual)... the higher number set to 1210 (or whatever you want), and edit the Y axis.

For coolant GP, there's 3 maps, unknown which one corresponds to which (manual, auto, ???) so I did all 3 of them to be sure.

To set a specific idle while coolant gp's are on (help alternator and serp belt load)
open up in vagedcsuite. Click view file in hex. Click find, click hex. put in 32 00 E8 03
The next two numbers after that is the idle when GP coolant heaters are on. Stock seems to be 34 03 (hex 0334- 820 rpm). I am doing 1088 rpm- in file that is 40 04. Change it to whatever you want! Remember it is reverse hex, as you see in my example.

 

[Sbeghan]

And it works? That's pretty cool. I might do this, or not. I don't live in a bitterly cold area and winter's almost over here.

I've been investigating my smoke map issue. Looking at the map addresses the map is clearly labeled as being a 13x16 map in the file (eg, 0D 00 x 10 00 = 13 x 16). I still don't know why vagsuite isn't properly detecting it. Oddly, Winols doesn't automatically detect it either.

The boost editing seems to be having some results and I'm on track to deliver higher mpgs. I do need to up the boost on the top end when I mash the pedal as its now smoking when I do that, but on partial load (cruising on flats at 45-75 mph) it hangs around 0-2 psi boost now, which is less than it used to, and it doesn't smoke. I think as long as its combusting all the fuel the lack of boost is improving fuel economy.
Of course, that could also be the slight timing advance talking too. Once I fill up a new tank and report my mpgs I'll edit the boost map again.

 

[Sbeghan]

I shake my fist at the mpg gods. 42.5 mpg, which is pretty much what I always get.
I'm going to up my boost target map (not back to stock, but enough so that its peppy and doesn't smoke).

 

[robnitro]

The dynamic idle I had to do some adjustments to it.
Since it is a straight line between the points the idle was dropping quick, as soon as the coolant went past 1C.

So I made the lower number(y axis) now 30C (86F), so the high idle drops only after it hits that temperature (better for heating and the alternator load with the coolant GP on). It tapers off to regular 903 rpm at 60C coolant temp.

Surprisingly, the high idle feels smoother than regular idling right after a cold start.

The coolant glow plugs are based on IAT temp, and that will heat the coolant to a specific temperature. From what old threads speak of, it starts with using just 1 coolant GP, while afterglow is on, then the other 2 (total of 3) coolant GPs stay on.

My cruising boost at those speeds is 3-5 psi. What was your stock cruising boost?

 

[Sbeghan]

3-5 psi as well. It's not really necessary though.
I should note that this tank I also changed out my stock 15" alloys with 17" alloys.

 

[Dilemma]

Remove the bumps in the SOI map, smooth it out. If you have stockish nozzles, you can increase timing on the top end a little. You can run 19 degrees of timing on stockish nozzles on an ALH no problem. Seems like the units are a little off?

Hi guys,

Just to prevent the software from being incorrect...
PD SOI correction factor = 0.023437
VP SOI correction factor = 0.01

Is that correct? If so, i will implement this in the next release version.

Thanks
Guido

 

[Sbeghan]

Dilemma, I'll check the vag com logs at idle to try to figure out what vag-com thinks the SOI is and compare them to the maps. I don't know what the actual correction factor should be.

Thanks for the latest update (1.3.6), I see you're already added in that conversion factor and updated the map detection for the north american ALH roms.

BTW guys, Dilemma is the author of VAG EDC15P Suite. Thanks again! (With all the V support, maybe you should rename it to VAG EDC15P+V?)

I'm going to edit my smoke map now (the 16x13 one) and see if that works. 42mg/stroke here I come!

BTW, there are also a couple maps that are IDed in the Euro ALH binaries that aren't IDed in the NA binaries (the fuel pump voltage map for one). I'm going to do some digging and try to get the addresses for those maps and send them to you as one big block to make it easier for you.

 

[robnitro]

Also maybe add the maf/map switch,
map sensor diagnostic setting (vagcom read more than 2601),
iq diagnostic setting (vagcom read more than 51mg/s)
egr by temperature (dynamic egr)

Oh and the coolant glow plugs tables I posted a page back.

Great program!

 

[Sbeghan]

rob, it would probably help if you could find the map addresses, the x and y axis addresses, as well as units and conversions for each of those maps. It would make adding those to the program a lot easier and make it more likely that they'd get added (and maybe donating too, haha).

I saw talk about how to switch the map from MAF to MAP. I'm kind of surprised that's not a button you can click on. The dynamic EGR would be nice too, as I wouldn't mind reenabling it when cold to help with warm up.

Anyway, I got back from my test drive.
I am using the older version of vag-com and have noticed something funny. The actual start of injection reading is given as positive degrees ATDC or BTDC - the name of the reading changes but the number always stays positive. EX, 5* BTDC, 2* ATDC. When you log it, the name is *BTDC and all the numbers are positive, even though some should be negative.
LOL.
I compared the SOI map with the vag com logs and I can't really tell how they correlate. Idle is around .2-.4* BTDC which is about what it should be based on the map (but of course, given such a small value the conversion factor has a huge wiggle room). When I mash the pedal down the SOI jumps to 15* BTDC, with an actual fueling of 5 mg/stroke and requested at 43mg/stroke (my max in driver wish) at 3717 RPM (where it pegs itself due to launch control, I can't go over 3717 RPM at a stop). Based on the SOI map I have it should be 2-4* BTDC at that quantity of fuel, or 15* based on REQUESTED fueling (why the hell would you make a SOI map that is based on REQUESTED fuel and not ACTUAL???). At partial pedal load (8mg/stroke requested) holding at 3717 RPM it was still giving me 15* BTDC SOI but the map would be requesting 4-5* BTDC.
So really, I have no clue what the ECU is doing in this case or what the actual conversion factor should be.

And in other news, my butt dyno and logs indicate that I am getting the requested fueling with the new smoke map. Yay! 10% more power. The calculated torque is up from 220 to 250 and calculated hp is up from 90 to 114. I also don't have any clutch slippage either (so far).

I might tweak my file a bit but I don't think I'll be doing anything else unless I get new performance mods or I find out ways to increase fuel economy.

 

[Sbeghan]

Someone just asked me about why the car won't start when hot. There are two start IQ maps here. I have no idea why it uses the map on the right. As you can see it injects NO fuel until 250rpm when hot. This shouldn't really be a problem if your starter is working well and you have a good battery but if you want to fix it then simply edit the map to look like the one on the right so that you know, it starts when hot.

 

[Sbeghan]

If there's anyone in NC who can donate some dyno time I would love to make a table of torque as a dependent variable of RPM, SOI, and MAP (ie, what's the most fuel efficient regime to tune these engines to?).

 

[J.P.]

I am following this thread to watch and learn. My head hurts a bit though

I should note that this tank I also changed out my stock 15" alloys with 17" alloys.

I am sure you know this, but if the rotating mass of the wheel/tire combo or diameter of the tires is different than your previous setup it will likely impact your mileage.

 

[robnitro]

What boost level to smoke map level did you settle on?

For engine braking, I tweaked the 0mg n75 duty cycle. With lower numbers being vanes closed (higher boost):
1500/25% 1750- 26% 1911 27% 2050 29 2250 31 2500 36 3000 40, and scale up to around 52% for 5500.
It works surprisingly well, not gasser like, but cool when lifting off at 3k, seeing the boost rise to 8 psi, lol.

I also found out how to redo the boost for the VNT17, 1749vb PD150 turbo. Turns out that even the oem PD150 boost target/ boost limit map is wrong, especially at altitude! In doing the calculations based on the garret website's compressor map, I see that at higher altitudes , the turbo surge line becomes way more apparent. The turbo makes up for it by being able to hold boost to higher rpm, as flow reduces. Check out that thread to tune your vnt15 comp. map which is easy to find.

Also, launch control with around 20-25% wheelspin allowance after 20kp/h. This was a bit hard to do at first, the calculations I did with +20% would pull back at higher rpm, I think because the ecu is slow to react?

 

[Sbeghan]

rob, I'm running 2150 mbar max, with a ratio of 56 mbar : 1 mg/stroke in the requested boost map and running on the smoke map a ratio of 21-22 mg air: 1 mg fuel.

I shifted the lower end of the boost map over two columns to reduce the low end boost but still have it be a bit peppy, its working pretty well so far, much better than my previous ham fisted attempt. When I get back home I'll post pics of the graphs (I hope no one is stupid enough to use my maps in their own tune).

Interesting bit there with the wheel spin on the launch control. I didn't think that that was possible.

I have noticed that if I mash the pedal down it will spike to about 19.5 psi (2350 mbar absolute). At 2150 max mbar requested (16.5 psi) that's a significant boost spike that I'm concerned about, especially as I'm already pushing the turbo a bit. I have a boost valve in the garage that I can hook up, but I feel that that is an inelegant solution (which will still make its way on as an additional security layer) to the situation.

I do not understand the N75 duty cycle map or how to edit it. I am led to believe that I can use this to adjust for surging, boost spikes, etc. I do not understand what to do with it though. Can anyone help? I am guessing that 25% duty cycle means all open (lower end of adjustment), and 70 or 75% is max closed.

How do the boost target map and the N75 duty cycle map interact and how do they work to make the actual boost? I really wouldn't mind lowering the responsiveness of the turbo to high fueling requests so that it does not create spikes over 17.5 psi (which is the max spike that I'd like to put my VNT15 through).

 

[robnitro]

It's been hard for me to see what to do with part load boost. I'm running 11mm with Titan 764 nozzles (240?). So my fueling, though it says 51mg, is actually more. But I did trim down the requested boost a bit below 20mg, and no smoke!

I figured that with EGR, let's say maf 300- the boost will be higher because the turbo has to push air in against the EGR gasses that are already pressurized due to EMP (however low they are).

Remove that EGR, and now you need practically no boost to do even 420 mg! So, it seems to only really need the boost past mid range...

I used to have a VNT15 stock turbo, tuned to 18.5 psi, no problems at all. Spikes were crazy actually when I changed nozzles but didn't get retuned, lol.


N75 info:

According to vag com and easy view in vagedcsuite:
20 means max vacuum to actuator, closed vanes
80 means min vacuum to actuator open vanes

But the confusion is that the actual values are inverse.. So some mapping programs may say 80 when they mean 20. How you can tell is that at higher rpm/ higher fueling, the stock map numbers go towards opening the vanes.

Make sure you set the actuator to crack at 3in Hg and be fully at the stop around 19in Hg. Or you can do what I prefer, set the low boost position to have a little bit of preload pushing on the stop. That way I have more spring force against the flow of exhaust on the vanes.

To set the n75 map, you take vag com, set up ONLY group 011 for logging (as it cant do too many fields - slows down sample time), and run WOT 3rd gear from 2200 or so up to redline. Look at the CSV file in excel or open office. You will see where the boost matched requested and the n75 duty cycle at RPM. Those values would go into your n75 map for whatever fueling you run (I assume 51mg?). For the values that don't match close at all... I would disregard those numbers because the ecu does its own PID adjustments to get target boost. You pretty much will see something like for example
2500 70%
2700 70.2%
3000 70.6%
3250 71%
3500 71.3%
and so on. Those n75 values are what you put in your table.
Make sure you smooth it out. If you are running less than 51, lets say 46 now, then make sure the next column is a bit higher. Lets say you need 71 at 3500 rpm, and are using 46mg. Then set up 40 column to have 70.5 and 51 to have 71.5 so that in between (where 46 is ) gets the 71 you wanted!

For me, I was getting undershoot, because I purposely adjusted the vanes to be conservative like I explained above about preload. So I had numbers in vag com that were lower than what was in my n75 map. After adjusting that, I am spot on perfect!

 

[Sbeghan]

I'll take a look and start working on that. I too would prefer to be conservative.

You say that you can get 420mg of air without boost. I can get 500-550 mg air on a cold day without boost.

 

[TDIJetta99]

Have any of you guys found the values for hot idle? I need mine to idle at 1050-1100 while fully warmed up and haven't had much luck.. I changed all of the temp/idle speed maps to taper to 1100 at 90C, but it always drops to 903 after about 90-120 seconds of running.. Can't find MAP sensor calibration either.. Found most everything else, and found a few things in the hex view.. Still learning the OLS software but I'll eventually make a 6000rpm 4bar tune that passes an OBD2 readiness test..