N249 delete: Why you should not do it.

[Weehe]

The check valve will only let air out of the reservoir, creating vacuum. There will be nothing actuating the DV. Stock set up is the only way that works.

The N249 functions the same regardless of platform.

Good point, not sure how I didn't think of that

[rtl5009]

N249 reinstalled will 034 fuel rails.

Is this sucker right? I went off memory and self suck blow testing.

[topquarkpc]

looks about right according to the pic i got from MAX when i was doing the delete with 034 fuel rail upgrade...see below pic for reference...

now i'll have to reference your pic to restore mine to semi OEM config... thanks for the pic.

[Lupe-81]

N249 reinstalled will 034 fuel rails.

Is this sucker right? I went off memory and self suck blow testing.

Any testing yet with N249 installed back? Btw, is it ok to use green check valve instead of that black check valve in oem N249 setup. I've lost that black one and I have few spare green valves..

[bigern45]

food for thought, i removed mine a few years ago trying to chase down a boost leak. i never noticed any difference with it off, so i left it unhooked. ill have to find a diagram as i finish up my build here in the next few months and make sure i get it back on correctly. probably do a set up similar to rtl3000 as ill be running fuel rails as well.

[rtl5009]

Any testing yet with N249 installed back? Btw, is it ok to use green check valve instead of that black check valve in oem N249 setup. I've lost that black one and I have few spare green valves..

Ill report back once I have more samples but I got about 1.4 mpg better on the last tank of gas compared to my avg economy over this summer.

Which is a 13% improvement in fuel economy on e85. lol

I boost above MAP limit so I would never pick up on the boost spike. My gauge is tied into the intake manifold, which wont see the spike either.

The car sounds slightly different when shifting at WOT and feels that it has boost slightly faster than before directly after shifts. That is not real data tho.

[wbrunner23]

Ill report back once I have more samples but I got about 1.4 mpg better on the last tank of gas compared to my avg economy over this summer.

Which is a 13% improvement in fuel economy on e85. lol

Wow, that's a really good improvement so far! Hopefully that continues.

I have been meaning to hook mine back up for a while now following a few conversations with Phila_dot. Have you noticed any difference during partial throttle (the reason most of us bypassed the valve in the first place)?

[rtl5009]

Wow, that's a really good improvement so far! Hopefully that continues.

I have been meaning to hook mine back up for a while now following a few conversations with Phila_dot. Have you noticed any difference during partial throttle (the reason most of us bypassed the valve in the first place)?

Nope. Partial throttle is the same.

I deleted mine when I installed 034 rails cause it was kinda just flopping around in there.

[wbrunner23]

Hooked my n249 back up this morning and I did not feel the partial throttle jerkyness that I felt before... weird cause I swear it was jerky before w/ my n249 and the bypass alleviated it...

Anyway, my lying in-dash average MPG was up 2.3mpg w/ 50 miles since last reset (25mi w/ bypass and 25mi w/ n249) - so maybe that will pan out to 1 or 1.5mpg in real life, hopefully.

I'm running 93 unleaded so it may be interesting to compare to rtl5009's e85 mpg variance once I see some more miles

[somebody5788]

My own car with N249 delete felt very weak in that area with actual bucking when cold started and immediately light accelerated. With N249 is a peppy happy NA like car in the light load situations.

My car does this with out it being modified. So did my previous 2.8l A4. I've come to the conclusion that it's just because we flow a ton of air through these 30v engines and when cold they just can't do a proper burn with all of that air and fuel. Ever tried to drive a cammed LS1+ while it's still cold? It's even worse.

[julex]

My car does this with out it being modified. So did my previous 2.8l A4. I've come to the conclusion that it's just because we flow a ton of air through these 30v engines and when cold they just can't do a proper burn with all of that air and fuel. Ever tried to drive a cammed LS1+ while it's still cold? It's even worse.

I am not sure about 2.8 but the N249 is known to go out on people without any symptoms as far as CEL goes. It is worth checking if yours still works. It is not possible to do without disconnecting it and run output tests via vag-com while blowing it with your mouth or something to verify flow though.

[jeffreyk]

.... Coincidentally, these boost spikes were popping off my TBB pretty reliably so they were pretty severe.

Does the N294 open the diverters to reduce the typical "boost spike" overshoot when the turbos first hit? My car tops out at 20psi and tapers to 18 (K04s)

I started have an issue with my tbb popping off after a track day. It happens under boost about when the boost spike to 20PSI happens, not as I lift. It was perfect before. The boot itself is perfect. Nothing has changed (unless something has melted somewhere - like a n249 or a check valve).

Jeff

[rtl5009]

Does the N294 open the diverters to reduce the typical "boost spike" overshoot when the turbos first hit? My car tops out at 20psi and tapers to 18 (K04s)

I started have an issue with my tbb popping off after a track day. It happens under boost about when the boost spike to 20PSI happens, not as I lift. It was perfect before. The boot itself is perfect. Nothing has changed (unless something has melted somewhere - like a n249 or a check valve).

Jeff

No the n75 can meter the wastegates boost onset. The n249 does not operate in this capacity

[S4James]

No the n75 can meter the wastegates boost onset. The n249 does not operate in this capacity

I believe it does operate like this in a last ditch save the motor type situation.

[rtl5009]

I believe it does operate like this in a last ditch save the motor type situation.

The n249 never opens to control boost spike. Never.

[S4James]

Thats not what I mean, its my understanding that the n249 will be engaged in a run-away out of spec situation to save the motor.

[killerkali]

n249 missing here correct ?

[Timtheguru]

Yup, and you're missing the bracket too.

[killerkali]

Yup, and you're missing the bracket too.

Perfect, thanks

[jibberjive]

The n249 never opens to control boost spike. Never.

It allegedly is part of limp mode to control overboost (allegedly because I haven't looked the disassebled code myself).

[rtl5009]

It allegedly is part of limp mode to control overboost (allegedly because I haven't looked the disassebled code myself).

For limp yes, OP was questioning if it was ever similar to n75.

[jibberjive]

I believe it does operate like this in a last ditch save the motor type situation.

The n249 never opens to control boost spike. Never.

Here is the conversation string I was replying to, not the OP

[phila_dot]

IIRC the N249 never opens with positive target boost pressure unless pedal angle is decreasing.

Not for any kind of limp ever.

[Meow]

2) When coming out of heavy load/WOT pull, where there was positive boost developed by turbos, the N249 pre-opens DVs just before throttle closes to route compressed air back into inlets and letting the turbos to freely spin down.

#2:
When coming off-WOT where there is large amount of pressurized air between turbos and throttle, the car without N249 is about to experience what is technically called a turbo surge phenomenon. As throttle closes to practically nothing, the air has nowhere to go as DVs are still closed. It takes about 0.2 - 0.3s for the manifold to develop deep vaccum, evacuate the air from DVs and then for DVs to meanigfully open. That air goes the only route it can, which is back into the turbo which at this point has no forced propelling it anymore. The compressor wheel, shaft and turbine then experiences severe g forces as the assembly brakes from about 150k rpms to nothing within fraction of second.

To illustrate this point, here is a graph showing the boost spike at throttle boot MAP sensor (this would NOT be visible on normal boost gauge hooked up to intake manifold). It shows baseline 32-33psi run then a foot off the pedal with associated instant boost spike to 40psi which takes a good moment to fizzle out as it goes back into the turbo and then the DVs open sometime in the middle of boost fall line. Coincidentally, these boost spikes were popping off my TBB pretty reliably so they were pretty severe.

One sample = 0.1s.

Is there any way to speed up the throttle plate closing action to build vacuum faster in the intake manifold?

[julex]

Is there any way to speed up the throttle plate closing action to build vacuum faster in the intake manifold?

Maybe... but it closes within 0.1s - 0.2s to an aperture that is its minimum opening at a given RPMs to maintain minimal cylinder filling. The engine always take air in even in fuel cut off decel mode, remember that. Throttle will never close to 0%.

When I look at my logs, the throttle goes from 100% to 20 odd % to single digit throttle % within two log frames which is super quick and most likely limited to physical properties of step motor within throttle body.

[Meow]

Maybe... but it closes within 0.1s - 0.2s to an aperture that is its minimum opening at a given RPMs to maintain minimal cylinder filling. The engine always take air in even in fuel cut off decel mode, remember that. Throttle will never close to 0%.

I know, but it seems to go back to 20% rather than 5 to 10% which it really should. You want LESS cylinder filling than you are currently getting because you want more vacuum so therefore you want to close the throttle plate more than 20%. That lag could be the issue.

When I look at my logs, the throttle goes from 100% to 20 odd % to single digit throttle % within two log frames which is super quick and most likely limited to physical properties of step motor within throttle body.

If it can go from 100% to 20% in two frames why doesn't it go all the way back to 5% that fast? That little hitch in the throttle seems like the problem to me.


A solution could be to adjust the tune to have a lower minimum opening at a given RPM to cause more vacuum faster in the manifold. :)

[julex]

I know, but it seems to go back to 20% rather than 5 to 10% which it really should. You want LESS cylinder filling than you are currently getting because you want more vacuum so therefore you want to close the throttle plate more than 20%. That lag could be the issue.



If it can go from 100% to 20% in two frames why doesn't it go all the way back to 5% that fast? That little hitch in the throttle seems like the problem to me.


A solution could be to adjust the tune to have a lower minimum opening at a given RPM to cause more vacuum faster in the manifold. :)

That's because minimum filling is I believe 15% in decel and to attain such filling at 7k rpms requires the throttle to open MUCH more than at 1k rpms... as the engine requires something around 7x more air (don't kill me, I know about engine VE, just approximating) to get into the engine for the same cylinder filling. Either way the vacuum drops to over -20inhg (in my car -24inhg) in both instances

If you're interested in that I can give you the table names you can modify... don't go too crazy with it as if you close the throttle completely, you will cause the engine to start sucking stuff from ehxaust and bombard turbos with rapidly changing air pulses going back and forth (changing directions) instead of just slowing the flow to minimal value as stock set up is meant to do.

[Meow]

That's because minimum filling is I believe 15% in decel and to attain such filling at 7k rpms requires the throttle to open MUCH more than at 1k rpms... as the engine requires something around 7x more air (don't kill me, I know about engine VE, just approximating) to get into the engine for the same cylinder filling. Either way the vacuum drops to over -20inhg (in my car -24inhg) in both instances

If you're interested in that I can give you the table names you can modify... don't go too crazy with it as if you close the throttle completely, you will cause the engine to start sucking stuff from ehxaust and bombard turbos with rapidly changing air pulses going back and forth (changing directions) instead of just slowing the flow to minimal value as stock set up is meant to do.

From my logs I get a 4psi spike when the throttle lets off.

It takes mine about .2 seconds to go from 100% throttle to 3% throttle. This gives me about a 300mbar (4psi) spike in boost pressure.

[Meow]

Do you have a "let off" log of with the n249? I am curious what it looks like with the n249 in place. Also you should switch to me7 logger with higher data collection rate.

[julex]

I have tons of logs I can send your way... I run my logging at 10 samples / second as I log a lot of stuff and overloading ECU with too many samples is easy to do.

[Meow]

I have tons of logs I can send your way... I run my logging at 10 samples / second as I log a lot of stuff and overloading ECU with too many samples is easy to do.

Ahh, true that. Im just curious as to what it looks like now with the n249 reinstalled. I am probably going to hook mine back up. Hopefully my reservoir still works ok.

[Meow]

Food for thought:

Whats the sample rate from the throttle body to the ecu for throttle plate angle?

[Meow]

So i think these spikes are just because of the map sensor location. It happens on all turbo cars but this is a weird car with the map sensor before the throttle body. Most cars never see this because of their sensor location. I think that the throttle actually closes fast enough.

[jibberjive]

So i think these spikes are just because of the map sensor location. It happens on all turbo cars but this is a weird car with the map sensor before the throttle body.

If the MAP sensor was post throttle body, you wouldn't ever 'see' overrun compressor surge via the MAP sensor.

By definition, that spike is what the N249 is designed to mitigate.

[Meow]

If the MAP sensor was post throttle body, you wouldn't ever 'see' overrun compressor surge via the MAP sensor.

By definition, that spike is what the N249 is designed to mitigate.

And I don't think that spike is really that abnormal. I think if you had a sensor on most drive my cable cars you would see the same spike, its not very large and its not enough to cause compressor surge. If the engine isn't firing, nothing is driving the turbo, and all you are seeing is the inertia from the turbo creating more boost against a severly restricted flow path once the throttle has closed. The turbo is slowing down rapidly from the second the throttle is lifted. It would be quite difficult to determine what airflow is actually occurring here as the flow area is severely restricted. The turbo likely will not surge as Rpm's are falling extremely fast and you wouldn't have high shaft speed reverse direction type stuff which really is an issue.

[julex]

you wouldn't have high shaft speed reverse direction type stuff which really is an issue.

I could hear compressor surge when my car was without N249 for a short while. It was a very distinct sound lasting about 1s when blades were chopping through stalled air (due to no flow through either throttle or DVs) until it slowed down enough to stop surging.

[Meow]

I could hear compressor surge when my car was without N249 for a short while. It was a very distinct sound lasting about 1s when blades were chopping through stalled air (due to no flow through either throttle or DVs) until it slowed down enough to stop surging.

Your spike is 8psi and lasts .2 seconds, mine is only 4psi and lasts .2 seconds as well. I do not hear any surge noise at all. I do not think my turbos are surging, however yours might have been. Got any video of the noise?

What diverter valves are you using? I use 710n valves.

[jibberjive]

I think if you had a sensor on most drive my cable cars you would see the same spike,

I agree, most drive-by-cable cars would likely show the same spike, as they don't have an N249 valve and vacuum canister. The N249's sole purpose is to mitigate this occurance.

If the engine isn't firing, nothing is driving the turbo, and all you are seeing is the inertia from the turbo creating more boost against a severly restricted flow path once the throttle has closed. The turbo is slowing down rapidly from the second the throttle is lifted.

You pretty much defined compressor surge/stall. Independent of what causes the pressure in the intake or the lack of pressure in the exhaust, compressor stall/surge happens by a mismatch of intake pressure being too much and exhaust pressure being too little. One symptom is a momentary spike in boost pressure.

If you're looking for someone to justify taking out the N249, I doubt you're going to find it here. If you're not hearing any surge and aren't too worried about the effect it might cumulatively have on your turbos, just keep it off.

[sparky17]

This is great discussion. There is more discussion on this topic on VWVortex. http://forums.vwvortex.com/showthrea...-keep-the-N249

[Meow]

You pretty much defined compressor surge/stall. Independent of what causes the pressure in the intake or the lack of pressure in the exhaust, compressor stall/surge happens by a mismatch of intake pressure being too much and exhaust pressure being too little. One symptom is a momentary spike in boost pressure.

If you're looking for someone to justify taking out the N249, I doubt you're going to find it here. If you're not hearing any surge and aren't too worried about the effect it might cumulatively have on your turbos, just keep it off.

There is a difference between a small pressure spike right at the throttle body (where the map sensor on our cars is) and full on surge. How much of a spike is too much? Too hard to answer that right now without a lot of testing, though im sure some engineer somewhere knows the answer. (might have to make some phone calls to find out)


One thing ive seen pop up a lot if running divertervalves backwords for quicker response. Im curious if this is true or not...