Pages: [1] 2 3 ... 5
Author Topic: Adding turbo to VR6 3.2, my experience, QA  (Read 26482 times)
larppaxyz
Jr. Member
**

Karma: +4/-0
Offline Offline

Posts: 45


« on: November 30, 2019, 02:20:41 AM »

Because i got good help here, i would like to share my personal experience of adding turbo to my Porsche Cayenne VR6 3.2 in form of QA's. It's not a how-to guide, just some questions that i had and it was hard to find simple answers.

Q: What is VR6?
A: In this context, i talk about 3.2 VR6 engine that has variable intake and exhaust cam. It's used it Cayenne, Touareg, Phaeton, Golf (R32). They all share exact same engine, there are some differences in ECU software (torque limiting). Cayenne has different looking intake, but otherwise it's the same. They all have WIDEBAND front O2 sensors and that is very important.

Q: Can i just slap random turbo kit from ebay to my VR6 and go?
A: Yes and no. That is how i started. See QA down below about O2 sensors.

Q: What to expect with stock engine?
A: I had 400Nm peak and around 290HP on stock ECU with turbo kit installed. Torque limiting made it hard to use that power on streets. (This with 85% ethanol and larger injectors, no modifications to ECU, larger injectors balanced out the consumption needed by ethanol)

Q: What to expect with stock engine and tuned ECU?
A: Depends, i would say 330HP, 410Nm when stock is 250HP, 310Nm. (This with ethanol and larger injectors). With compression lowering kit, some people say, 400HP is safe limit.

Q: What to do before anything?
A: Make sure your engine and sensors are in good working order. Fuel trims should be good and no codes stored. Any sensor failure will damage your engine when on boost.

Q: What is BARE MINIMUM to install turbo and actually be able to drive it?
A: Install random VW turbo kit, install wideband o2 sensors for bank 1 and 2 to exhaust manifold. Connect all sensors like they are supposed to be, including MAF.  You car will now have around 20% more torque and power, but is torque limited and drives badly. Move brake booster hose and PCV hose to somewhere else, they can't stay on intake manifold anymore because there is boost in there now.

Q: So, i can use turbo with stock ECU?
A: Yes. ECU's are torque limited, so it will max out at +20%. It's drivable.

Q: Why my throttle response is jumpy and i get random stalls and power peaks?
A: There are many reasons, but you should disable variable intake system. It has two positions, torque and power. With turbo installed, it only causes issues. Simpy disconnect it mechanically.

Q: How much boost? Is compression reduction kit needed?
A: I'm running 0.65bar with stock compression ratio. However, i'm using 50% ethanol and still have torque limiting on my ECU (on purpose, i made KFMIRL values max out at 150 for safety)

Q: How much boost is safe on stock compression ratio and 'normal' gasoline?
A: I can only tell you, your engine will crack a piston at 0.65bar. That is what happened to me when running with failed sensor and there was no torque limiting. 0.5bar should be safe still? (UPDATE: I had piston failure, but it's very much possible that it was not caused by havin too much boost)

Q: MAF? Do i need to change that?
A: No, maybe if you are looking maximum power on high rpm.

Q: Why you keep talking about ethanol?
A: It's available here for a cheap price but most importanly, it reduces engine compression rate. You can give it more boost without need for compression lowering kit. Or to but it other way, it's more safe to boost it when using ethanol. If you can't use it, i would go for 0.5bar max.

Q: Can i just fit my wideband O2 sensors to exhaust pipe after turbo?
A: No you can't. Your car will not run at all. Exhaust gasses cant be mixed before they go to O2 sensors. You also can't run only single O2 sensor or anything else like that. You simply need to fit them properly so that sensor 1 gets exhaust from engine bank 1 and sensor 2 get exhaust from bank 2. Anything else, and your engine will pop.

Q: Rear O2 sensors? What about them?
A: They need to be electrically connected. You can fit them to exhaust on same pipe. They will give engine code every now and then without catalytic converters. If you leave them disconnected, front O2 sensors wont work AT ALL.

Q: Injectors?
A: There are differences between engines, some have fuel pressure set to 4bar, some cars to 3bars. Do your calculations. Anyway, i got around 30% larger injectors from Audi TT 1.8. They are almost plug&play, connector is bit different, no soldering required or any modifications. (022 906 031 J VW/PORSCHE3.2 -> 0 280 157 012, 06A 906 031 J AUDI TT 1.8 QUATTRO (AMU) -> 0 280 155 892)

Q: Where can i find ECU tune for turbo installation?
A: I don't think you can, most people i asked and offered money just said it's stupid and can't be done and blablaa... this is not on this forum, but when i asked "tune shops" here locally.

Q: Can i tune my ECU?
A: Yes. It's Bosch ME7.1.1 unit. Very common, used everywhere.

Q: What are keywords to ECU tuning?
A: Kess v2, WinOls Demo, ProTuner, ME7Sum.exe, Bosch Me7.3 documentation for Alfa romeo, S4 Wiki, this forum. That is all you need.

Q: What modifications to make for ECU to make car run well enough for daily driving?
A: To keep it simple, all you need to do, is make torque (limiting) curve match your new engine. Because wideband o2 sensors are used, you don't really have to worry about fueling at all. With narrowband O2 sensors i would say that turning NA engine to FI is close to impossible and dangerous.

Q: What exact modification to make for your ECU to get past torque limits?
A: Well, find 100% load caps, edit KFMIRL to suit your car. It takes forever to make it good, but simply multiplying last two rows and smoothing it out a little should make it "ok". After that you start learning and tweaking more, there are tens of maps that you can (and need to) modify.

Q: They talk about MAF scaling, why?
A: Basically you fake that there is less air coming to engine, and that way you can trick ECU to thing that there is less load and torque. It's method to get past torque limits that are programmed to ECU. I would not recommend this because internal load values control other things too (like engine timing). (when swapping different MAF scaling is needed too)

... ok this is getting long. These are just what my experience is. There can be errors, but i hope this helps out someone.

Highlights from below :

Q: I have DSG transmission, does it matter?
A: Yes, DSG transmission have their own torque limiting and basically that means you need to tune transmission ECU too.

Q: Can i use any ME7.1.1 ECU? Like, get a spare and test with that?
A: Yes and no, ME7.1.1 ECUs use different hardware and they generally are not compatible unless it's exact same hardware number.

Q: Do i need to modify ignition timing tables?
A: Personally i didn't touch them and i trusted knock sensors to keep me safe. However, people on this forum keep telling that they absolutely need to be modified. For optimal performance they absolutely need to be modified, but for safety, i'm not sure. You decide if you trust your knock sensors Smiley


 


« Last Edit: December 01, 2019, 01:34:59 AM by larppaxyz » Logged
Blazius
Hero Member
*****

Karma: +89/-40
Offline Offline

Posts: 1277



« Reply #1 on: November 30, 2019, 04:40:22 AM »

Nice writeup, few things I'd add.

You can use checkvalves on PCV and brakebooster so that they only open when engine is vacuum, this way you can keep it connected to manifold.

The variable geometry manifold has to be completely disabled, even if you unplug the connector the module is activated and there are multiple load changing variables that will reflect on your load.

You can tune a fi conversion with nbo2 too but you have to watch knock limit and sensor voltage during boost but highly recommended to fit external gauge, a bosch lsu sensor is cheap and a DIY 14.7 controller or etc too.

You NEED to fit a DV , else you'll stall everytime you let off low load / high rpm and on gear -> neutral switch.

If you move / change the torque limiters and/or you changed your torque control/calculating system, you might it hit a level 2 intervention cap, this comes from the "UM" modules. There are maps for torque allowed / pedal position % etc., might have to rescale this too to avoid the engine control limit 2 dtc during changing weather or conditions where your current torque exceeds the max allowed.

You can disable the rear o2 too.
« Last Edit: November 30, 2019, 04:41:54 AM by Blazius » Logged
larppaxyz
Jr. Member
**

Karma: +4/-0
Offline Offline

Posts: 45


« Reply #2 on: November 30, 2019, 05:55:07 AM »

The variable geometry manifold has to be completely disabled, even if you unplug the connector the module is activated and there are multiple load changing variables that will reflect on your load.

Yes, when tuning, identical values can be used for both positions (power/torque) if variable geometry manifold is mechanically disabled. It's bit more advanced topic, so i left it out. It's however important thing to know when tuning, so thank you for mentioning it.

Logged
fluke9
Full Member
***

Karma: +26/-1
Offline Offline

Posts: 113


« Reply #3 on: November 30, 2019, 06:21:36 AM »

Q: How much boost is safe on stock compression ratio and 'normal' gasoline?
A: I can only tell you, your engine will crack a piston at 0.65bar. That is what happened to me when running with failed sensor and there was no torque limiting. 0.5bar should be safe still?
A cracked piston is more a sign of some super heavy knocking going on, no engine will last for long under that conditions.
Some R32 here run around 0.6 bar without Ethanol, i dont know the compression ratio for the VR6 but on my Alfa V6 i can run 0.65 with 10:1.
Although most of them are supercharged so the boost comes on way smoother and possibly cooler that with a turbo setup.
Logged
larppaxyz
Jr. Member
**

Karma: +4/-0
Offline Offline

Posts: 45


« Reply #4 on: November 30, 2019, 06:35:27 AM »

A cracked piston is more a sign of some super heavy knocking going on, no engine will last for long under that conditions.
Some R32 here run around 0.6 bar without Ethanol, i dont know the compression ratio for the VR6 but on my Alfa V6 i can run 0.65 with 10:1.
Although most of them are supercharged so the boost comes on way smoother and possibly cooler that with a turbo setup.

Whatever causes it, generally failure mode for VR6 engines seems to be piston cracking between oil and compression ring. There is no damage on top or any burning marks. I personally experienced this.  I was on normal gasoline, redlined, without any torque limiting, and second wideband o2 sensor was not working correctly (slow to respond) and probably on around 0.7bar boost so i really asked for it.. Smiley

Logged
IamwhoIam
Hero Member
*****

Karma: +43/-99
Offline Offline

Posts: 1030


« Reply #5 on: November 30, 2019, 06:40:54 AM »

0.65 bar and a cracked piston is knock, nothing else. Probably because of running stock N/A ignition timing while blowing 0.65 bar at it. Duh, that's the most retarded thing one could do. If you had run the appropriate amount of timing, you wouldn't have cracked a piston at 0.65 bar even with 98RON/93 AKI init.
Logged

I have no logs because I have a boost gauge (makes things easier)
_nameless
Hero Member
*****

Karma: +320/-445
Offline Offline

Posts: 2654



« Reply #6 on: November 30, 2019, 06:43:27 AM »

i ran 1 bar stock compression on my 12v with a 60% / 40% e mix. Occasionally hit 1.3 bar with full e with no knock . props to op for documenting their findings
Logged

If you are in the market for a tune and would like the ease of downloading and flashing a dyno tested tune for a fair price check out https://instatune.sellfy.store/
larppaxyz
Jr. Member
**

Karma: +4/-0
Offline Offline

Posts: 45


« Reply #7 on: November 30, 2019, 07:34:20 AM »

0.65 bar and a cracked piston is knock, nothing else. Probably because of running stock N/A ignition timing while blowing 0.65 bar at it. Duh, that's the most retarded thing one could do. If you had run the appropriate amount of timing, you wouldn't have cracked a piston at 0.65 bar even with 98RON/93 AKI init.

Ok, since you seem to know enough of this to tell what is retarded, maybe you can tell us what kind of ignition maps we should be using when going past >100% cylinder fill and if cam overlapping has anything to do with it? Keep it simple, maybe we could add your answer to QA section.


Logged
fluke9
Full Member
***

Karma: +26/-1
Offline Offline

Posts: 113


« Reply #8 on: November 30, 2019, 07:58:18 AM »

Ok, since you seem to know enough of this to tell what is retarded, maybe you can tell us what kind of ignition maps we should be using when going past >100% cylinder fill and if cam overlapping has anything to do with it? Keep it simple, maybe we could add your answer to QA section.
You need to retard the ignition table too, otherwise it is retarded to run it this way. ;-)
Simple hacks like making the ECU see less load will do this, i used to add a diode on the MAF with a 0.5v forward voltage on my GTV which removed some timing,
but i ran fuel on a megasquirt.

Expand igniton table to cover your new load cap and retard there.
Best way would be live on a dyno, watch your knock and slowly find the spot where it sees slight knock, retard with a safety margin then.

Engines differ quite a bit in how much advance they can handle or not before knocking, also fuel plays in a lot and.
IAT is also a big factor, decent intercooling will help you a lot to run more advance.

Logged
Blazius
Hero Member
*****

Karma: +89/-40
Offline Offline

Posts: 1277



« Reply #9 on: November 30, 2019, 09:07:46 AM »

You need to retard the ignition table too, otherwise it is retarded to run it this way. ;-)
Simple hacks like making the ECU see less load will do this, i used to add a diode on the MAF with a 0.5v forward voltage on my GTV which removed some timing,
but i ran fuel on a megasquirt.

Expand igniton table to cover your new load cap and retard there.
Best way would be live on a dyno, watch your knock and slowly find the spot where it sees slight knock, retard with a safety margin then.

Engines differ quite a bit in how much advance they can handle or not before knocking, also fuel plays in a lot and.
IAT is also a big factor, decent intercooling will help you a lot to run more advance.




Exactly extend KFZW load tables(and the related maps that use the axis) and redo the timing tables with common sense, I'd say aim for 0 cf maybe a max of 3 avg, but even that personally I wouldnt like on such high compression.
Logged
larppaxyz
Jr. Member
**

Karma: +4/-0
Offline Offline

Posts: 45


« Reply #10 on: November 30, 2019, 09:29:55 AM »

Exactly extend KFZW load tables(and the related maps that use the axis) and redo the timing tables with common sense, I'd say aim for 0 cf maybe a max of 3 avg, but even that personally I wouldnt like on such high compression.

Actually i have already extended few axises in such way that second last row goes to 100 and last row jumps to 150. Currently values i'm using at 150 are (basically) identical to 100 row. Primary ignition A and B is one of those maps where modification is done. I haven't had time to check if i have found all places where axises are used, but it should not matter too much. Anyway, for example last row at max load and 5000 rpm is now 12.72 (for example). You mean this should be retared close to zero? That sounds like huge change, only values close to zero are below 3000 rpm. Picture added for reference.

... i quess we are slowly drifting from "how to get bare minimum to work" to "how to tune your turbo vr6", but i quess it's ok. Lets just stay on VR6 and known facts so it's still helpful. I also would like to add, that in this context, by load i mean cylinder fill percentage.



« Last Edit: November 30, 2019, 09:32:16 AM by larppaxyz » Logged
IamwhoIam
Hero Member
*****

Karma: +43/-99
Offline Offline

Posts: 1030


« Reply #11 on: November 30, 2019, 09:46:14 AM »

Ok, since you seem to know enough of this to tell what is retarded, maybe you can tell us what kind of ignition maps we should be using when going past >100% cylinder fill and if cam overlapping has anything to do with it? Keep it simple, maybe we could add your answer to QA section.

First off you extend the range of your load axises on KFZW, then you significantly drop the values in there starting with 100% load, you might have to do some rescaling and move columns/rows around, then you start with as little timing as possible and you go from there. Ideally on a dyno too, so you can see where MBT is at.
Logged

I have no logs because I have a boost gauge (makes things easier)
fluke9
Full Member
***

Karma: +26/-1
Offline Offline

Posts: 113


« Reply #12 on: November 30, 2019, 10:01:45 AM »

You should extend the whole axis, not just one row,
with your map you end up in the same row betwenn 100 and 150% load so your 0.0 to 0.5 bar boost has the same iginition angle.
Logged
larppaxyz
Jr. Member
**

Karma: +4/-0
Offline Offline

Posts: 45


« Reply #13 on: November 30, 2019, 10:05:40 AM »

You should extend the whole axis, not just one row,
with your map you end up in the same row betwenn 100 and 150% load so your 0.0 to 0.5 bar boost has the same iginition angle.

Sure about that? My understanding is that it should interpolate between two rows instead of using one or the other?

EDIT: I moved original "last row" one row up. Then i modified axis so that last row is 150 and one before that is 100, like you see in picture. Note that in picture, 100 and 150 row are intentionally same as i haven't modified them. If ECU now looks for row let's say 120, it will fetch rows 100 and 150 and interpolate between values in those. So in theory, whole map could work with just two rows.
« Last Edit: November 30, 2019, 10:11:14 AM by larppaxyz » Logged
fluke9
Full Member
***

Karma: +26/-1
Offline Offline

Posts: 113


« Reply #14 on: November 30, 2019, 10:10:00 AM »

You mean this should be retared close to zero? That sounds like huge change, only values close to zero are below 3000 rpm. Picture added for reference.

You need to learn a bit about the basics first.
The faster the engine runs the more advance it needs, as you can see in the table.
This is because after igniting the mixture it takes some time for the flamefront to travel to the piston.
The faster the engine is turning the earlier the spark must be ignited for the explosion to happen right AFTER turningpoint of the piston on OT.

Knock is if the pistion still travels up when the explosion happens, therefore the rotational energy of the engine and power of the other pistons made before
will push the piston into the explosion.
Thats what breaks your piston.
Logged
Pages: [1] 2 3 ... 5
  Print  
 
Jump to:  

Powered by SMF 1.1.21 | SMF © 2015, Simple Machines Page created in 0.026 seconds with 16 queries. (Pretty URLs adds 0s, 0q)