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

robnitro said:

JASON, search for 200 in vagsuite to find map linearization. For some reason the higher value doesn't get found.

Click to expand...

I tried that yesterday and didn't find anything, but I open up edcsuite today and there was an update.. Found it bu searching for 200.. I looked through every single one of the "unknown" maps over the last few days to no avail.. with this most recent update, there are 400 more "unknown" maps in my file.. I guess the GN ecu is a weird one since I was able to find everything easily in an older ECU file..

*throws hands up in the air* ok, I started trying to finesse the PID maps, and then I just went at them with a butcher's knife and chopped the PID I and PID P maps in half which should have done something, and it seemed to, but it didn't have much of an effect.
The N75 lower limit attack was even worse and didn't seem to do a damn.
I'm going with Rob's original suggestion of adjusting the actuator to open at 5 inhg, or whatever it takes so that I can feel that I can safely mash down the pedal under ANY conditions (maybe 1st or 2nd in the rocky mountains on a cold day) without grenading my turbo. Turbo lag be damned. Lag > turbo explosion.

Did you try to change the n75 static map to be higher, like adding 5 or so on easy mode vagedcsuite for 40+ mg/s?

Lag is fine for me, if I roll into throttle its instant, I think the n75 like this is best set conservative cause at wot when the fueling is insane.

Comical as it is seeing you guys flail around like this it's kinda driving me a little nuts

you forget that the ecu has separate internal and external values.

The internal value is used within the ecu itself for its internal calaulations.
This gets translated into an external value for use by the diagnostic module (which is what you see in vcds)
Two separate things entirely. It's why the ecu can control boost to 29 psi with a 3 bar sensor, yet the boost readout in VCDS ends at 2601 unless your tuner knows how to change this to read the full 3 bar. (tuner limitation, not a vcds limitation)

If you want to limit the spiking or slow the rate of increase you are barking up the wrong tree by looking at the PID. This only affects the amplification of the control signal.

You are more or less looking to limit the magnitude of the actual control signal, not modifiy how quickly it changes. This means the N75 upper limit. Raise the limit = faster response and more spikes/ undershoot. Lower = slower max response and smaller spikes, and potential underboost.

ECU maps are for internal representation not external so don't get them confused with what you see in vcds.

re-reading what I just wrote I see a few points that can be nit-picked.

But for what you guys are trying to accomplish, look at the N75 UL. Of course the effectiveness of further increases is subject to the physical limits of the stop screw and so forth, unless you are trying to hysteresis lock the actuator at full travel. test and learn.

There's more that can be done with the PID but it doesn't come easy. For guys that are struggling to figure out how to find and understand the very basics like map sensor linearization I figure let's just keep any comments as straightforward as possible.

I was wondering when we'd get a pro in here
haha. REMEMBER THE TITLE OF THE THREAD.

Rob, took your advice and set the actuator to crack at 5 psi (or maybe a bit over). Now I can make it overboost to about 18.5 psi, which isn't much over the 2200 mbar I've got in my boost map. Much better than the 22 psi I would sometimes see (and then immediately let off the gas, lol).

Kerma, I was thinking that if I didn't touch the N75 upper limiter it would allow it to fully open during the initial rise, and then back off so that it wouldn't overshoot (if the PID were tuned correctly). Of course, we're talking differences of a fraction of a second to come to full boost from no boost which is not really something I care about in a daily driver. Anyway, I did (I think) try to change the N75 limiter to get that effect - I probably didn't do it right - and messing with the actuator worked perfectly.

Hey we all learned somewhere.. I'm sure even the best professional tuners struggled with some basics at first..

Awesome, that's great! It helps because sometimes the springs get weak too.

Now you should log again, and then set your static map to the values you see that stabilize the boost. After that you won't get any spikes!

BTW, I just did the calculations for the VNT15, aka 1749va turbo, and you can go higher on boost.

It's nice to see that yes, it can do 20-21 psi, but only at a certain range! (and of course to limit spikes!) I am reminded of a few people that did exactly that, 21 or so psi! Even my old vnt15 did 24 psi spikes with no probs. I upgraded when they had vnt17 cheap on ebay lol.

It explains what causes them to blow up (surge esp at altitude - not really choke), and they seem to be even less affected in high rpm than the VNT17, I suppose because they share the same housing, so it is flow limited (which pushes the right line- aka choke line).

Also, this map and the VNT17 map makes me see why people get dynos where the HP becomes flat, the boost drops, so less mg/s at higher rpm, BUT higher rpm makes up for the loss in torque (same hp). At higher altitudes the HP curve looks more like a gasser at sea level (thanks for turbos!).

If you want to use these values (or a percentage of them), use the absMBAR ones, in the boost limit map (redo the rpm axis to cover all rpm), and for the in between altitudes ex in vagsuite: hold shift select 1500rpm for 1000 and 1500rpm for 850, then right click smooth.
The 1000mbar values go in the normal boost map.

Sbeghan said:

I was wondering when we'd get a pro in here
haha. REMEMBER THE TITLE OF THE THREAD.

Rob, took your advice and set the actuator to crack at 5 psi (or maybe a bit over). Now I can make it overboost to about 18.5 psi, which isn't much over the 2200 mbar I've got in my boost map. Much better than the 22 psi I would sometimes see (and then immediately let off the gas, lol).

Kerma, I was thinking that if I didn't touch the N75 upper limiter it would allow it to fully open during the initial rise, and then back off so that it wouldn't overshoot (if the PID were tuned correctly). Of course, we're talking differences of a fraction of a second to come to full boost from no boost which is not really something I care about in a daily driver. Anyway, I did (I think) try to change the N75 limiter to get that effect - I probably didn't do it right - and messing with the actuator worked perfectly.

Click to expand...

I'm aware that I could hug the top of the map if I wanted, but because of the possibility of overboosting and pushing it into the surge region I decided to leave myself a larger buffer zone. I actually took some of the boost target values below 2000 rpm down as well as using another limiter by atmospheric pressure and rpm to limit the boost at lower altitudes.

Besides, I need larger injectors to use the air that I've got already!

I'll log the N75 DC later and do as you suggested.

Surge is really something that happens when flow is low. At higher speeds the turbine wheel is what spins too fast. The problem is when the dots are off the top line, that is the top safe wheel speed. On the vnt17 map I posted earlier, the map seems like it is missing something on top, cause those ovals (efficiency islands) are halved?!
So, in my case I followed what others did (pressure wise - w good results) on other forums, and went a bit above the top line, but hugged it. I was surprised that the vnt15 map can hold longer!

At higher altitudes, the boost can be held out longer, because there is less flow. Same w vnt17. Keep that in mind, it helps you regain the power that you lose down low due to surge.

What pisses me off is that even the oem map, at least for the PD150, calls for too much boost down low, to the left of the surge line! I could care less how much torque I have below 2000. In fact it makes people not rev the car!

If you're tuning for power, neither a VNT15 or a VNT17 are going to do any good really haha.. use a 17/22 if you still want the stump puller effect and a stout, responsive daily driver, 1756VK is a little softer down low but makes up for it throughout the midrange and up top.. 2260 if you want to go big..

I was playing around with a stock ALH over the last few days and I got it to run a LOT better only making minor changes.. still zero smoke, and no additional boost pressure anywhere in the rpm range.. I actually reduced it a little in certain areas.. I have a nice steep hill near my house that's good for doing before and after testing too.. I drive past my road at 50mph in 4th and apply full throttle as I pass, and record my speed at the one pole right before the top of the hill.. a stock ALH that runs right will usually only get to about 60.. PD about 65..

Yeah a stock file is really underfueled lol. The map I'm looking at now is 38.5 mg/s. With 14.5 psi boost thats around 850mg/s air. Divide by 16.5 afr, you can do 51 mg/s (actual- if you can reach that by the 10mm pump and stock .184 nozzles) 51/38.5= 1.33x , So a 90 hp TDI can end up doing 120hp if the 184 nozzles can push it! If you set the turbo to 18.5 psi, that's around 1.13x more than that, which gives you 136hp.. which is around what people do with vnt15/205... hmmmm.

I got my car smokeless... at 51mg/s max smoke map (and max 4.85 pump voltage in map) it was smoky with IQ idle of 4, I turned it up to 5.5 and there's no smoke at all. So, for 11mm with Titan 764 (240?) nozzles a VNT17 can give enough air at 1.6 bar (23-24 psi) if turned down by 1 - 2 IQ. Of course like we talked about, 51 is really more, estimating 16 afr to be smoky at that airflow (1118 mg/s) gives 67 mg/s which is around what the guy that calibrated the 11mm/.232 pump voltage map to max out at (65mg/s he has as last column). If we use a real 51mg/s @ 136hp (from above), 65mg/s is around 170hp, around what we see with VNT17.

For vnt17/22, I guess 27 psi is what you run, and thats 1.86 bar. (2.86/2.6= 1.1x the air flow) so bigger nozzles would be needed to fuel it. With the airflow around 1230 or so and afr of 16.5, would need 75mg/s! That would do around 200hp, which is what you got on the dyno right?

It's pretty cool to see how the boost/airflow/fueling can be used to calculate hp. Check out what vagsuite says in actions -> view performance. If IQ is calibrated, it is pretty much accurate!

I don't have a compressor map unfortunately.. I was running the 17/22 with R520's and it would smoke like a train at full pump voltage even at 30psi from 2000-5500rpm. I typically was running it at 27-29psi all the way out to max rpm.. The tune was asking for 28psi but it would creep up to 29 out on top with the excess fuel.. This was all with the 99.5 and a pretty aggressive RC5 tune..

I still have to swap map sensors but I'll find out in the next week or so just how much air I need to clean up fuel from nozzles that flow a little bit more than the R520's.. I'm going to start really conservative and turn it up a little at a time.. My engine hardware should handle more than the 2260 and current injectors can throw at it..

After I fixed the air in my fuel lines, the leaky/blocked EGR valve, and the screwed up quantity adjuster and SOI (lots of diesel purge) my car has been a smoke free wonder. WOT, 2-4k rpm, no smoke. Before and after tuning.

I'm interested in the VNT17/22 but I plan on going with the VNT17 for the cost and low end torque. My next upgrade whenever that will be will be the largest nozzles I can stick on it so that I can improve my fuel economy by closing the injection window. I really don't feel the need for mega power (I'm still pretty tickled with the 10% increase in torque and 20% increase in HP just from tuning).

tdijetta99, what car are you tuning? you said a stock ALH? So you're not tuning the monster?

Wave 4 ALH cars here at the house LOL.. I'm messing with 2 of them at the moment since #'s 3 and 4 don't run yet.. #4 will be the fuel economy car that my wife will drive.. It has to be at bare minimum as fast as a 150hp 1.8T, and return the kind of fuel economy my 99.5 did with RC5.. The 99.5 would get mid-high 50's on her commute (900 mile tanks) and had completely stock internals (1/2 clogged intake and all) with a 17/22, PP764's and a 10mm pump.. It had the RC5 ecu that was in it when I had the other engine in it (the one that had the camshaft debacle) but with a stock map sensor.. the tune was set up for a 3bar and a LOT more fuel (11mm+R520).. It ran about 19psi and was smooth as silk, no smoke at all unless you stood on the throttle from 1100rpm, and it was reasonably fast considering the conservative setup..

Save up the few extra $$ and go for the 17/22. Use the cash you're saving by tuning it yourself lol. It's more responsive than the 17 in almost every setup, and can make enough power down low to break pistons if you desire..

Do you have compressor maps for the 17 and 17/22?

I don't have any compressor maps for these cars.. I have a few for turbos that are MUCH larger than ours since I'm trying to spec out a turbo for a low boost 7.4L big block..

darn.
Seeing those others could be fun though.

Sbeghan said:

Do you have compressor maps for the 17 and 17/22?

Click to expand...

Check my other compressor map post, thats the VNT17.

Here's a little thing I did, after I noticed that I was making smoke past 3500 rpm or so.
After seeing the bosch and tech4tdi pages on how the maf works and what limit gets hit when it is saturated, I was able to do a work around. Instead of using smoke limiter , which the maf is fooled, saturated reading 1100+mg/s I put the limits in the torque limiter.

So, we see how the maf gets saturated with VNT17 and above. Most people with big turbos go with map based limiter or get the VR6 maf housing which reads lower for a certain flow.

Your compressor map picture is kinda hard to read...
And I wanted to compare it to the vnt17/22.

Ok, I did the data logging and by adjusting my actuator for less boost the duty cycle changed the point that it stabilizes at, previously where it was at 63% it went to 56% for instance. I changed the values in the N75 DC map to the values that it stabilized at when I logged them and... the boost spikes are back! I only tried a couple times but I easily got it to spike to 20.5 psi instead of the 18.5 psi it was spiking at with the stock N75 DC map.
I'm just going to flash it back to stock and leave it there.

It was 56 when it stabilized? If so, it may be that the stock ecu pid is designed to allow a small spike as explained in the faq. Spikes are fine btw, as long as the don't happen too low (surge line below 1900 on vnt15/17).

I think its designed to allow a small spike. I'm still paranoid about them so I'm going to prevent them. Its only a fraction off a second slower without the spike and that's not really important to me.
Next question. Have we confirmed how to change the idle permanently? I'm curious if I can use a TDI as a stationary generator and changing the idle to 1800 rpm seems like a good way to do this.

Just hook up a throttle pedal and set up the fueling to limit engine speed.. This way you can idle it for a minute before slamming a load on it.. I'm sure the engine won't be happy from a dead cold start to 1800rpm/heavy load instantly..

I figured I'd have it start at 900 and ramp up to 1800 when it got warm.

that'll work too haha..

Idea to change idle to be higher for Jason: like your fueling map idea, pick the driver wish for zero and fill in the rpm up to where you want to idle, like how we do launch control limit.

You may have to change the numbers for rpm axis..and tweak it to be smooth, ex: 1100 idle; trail off from 6 IQ in 1080 to 0 at 1150.

Give it a try and let me know how rough it idles LOL.. Look at all the maps like that, they start a 0 and something dumb like 21rpm..

robnitro said:

Idea to change idle to be higher for Jason: like your fueling map idea, pick the driver wish for zero and fill in the rpm up to where you want to idle, like how we do launch control limit.
You may have to change the numbers for rpm axis..and tweak it to be smooth, ex: 1100 idle; trail off from 6 IQ in 1080 to 0 at 1150.

Click to expand...

If I got my launch control smooth at 2600, the same can be done for 1200. See page 3 for my example.
It wont work for running a generator load though.

I was thinking that setting the idle to 1800 over 50C would be the best way to make a generator as it would run up to speed once it hit operating temp, but launch control should work too.

It might just be a bit bouncy...
I think the reason is that launch control is dumb, whereas cruise control and (probably) idle are controlled by PIDs.