Finally got logs, now questions!

[lswing]

Well, third laptop worked with the RT cable, windows 7 on this one. Car still "bumps" a bit when under full throttle at top boost, and you can see the boost numbers look good. But what about the timing, seems steady to me? I did not log the knock detect, maybe tomorrow, ideas? Here's a screenshot of some grouped logs...

Also, if boost is at 1 bar sitting, then 2 bar running open would be the 14.5 pounds I'm guessing? Car is wicked fast right now, just a rocket, and dialed in on the corners. The Koni's are just a touch soft for hard cornering, but are very stable, and this is on soft setting which is a bit bumpy around town.

We put the w/m kit on next Monday, so just trying to get some good baseline numbers. I still think the car is pulling timing, what do I need to log to see it? Thanks!

12548

[4everRS]

Need a good 3rd gear pull to redline. This is not really enough data. Timing is block 3.

[lswing]

Yep, pulls are in third, got up to about 80, redline is 105...thanks for the input, I'll find a bit more space. Looks like I had block 2 for some reason. Headed out tomorrow for another set, I think i need to increase the logging frequency also to get more details...so 3, 31, 115, will read up a bit here too. Just amazed that two laptops would not work, third did, odd.

I'll find a better way to post the logs also...

What are you running the w/m on for settings, and what has been the primary impact? I've got two d05's drilled into the y-pipe, kit installed, finishing boost lines on Monday. Got the tank tucked into the battery spot in back, making a custom cover, will post pics.

[lswing]

Ok, found the logs I thought i missed earlier today, I did get bank three, take a look. Boost goes up, timing goes way down. Please correct me on this, but if the car sits idle at 1 bar boost, then registers 2.3 bar boost, I'm running about 18-19 pounds of boost?, which is the maxed out Revo tune. This is my timing being pulled causing the "bumps", correct?

12549

[4everRS]

Timing is going too low. Sign of an aggressive tune. Maybe you need to hit turbo mode on VagCom. That's not enough data points.

[ttboost]

How are you running that much boost without going into limp mode after a while?

[hahnmgh63]

As mentioned, you need to hit the Turbo mode button on Vagcom so the sampling rate increases, and you are running too much boost. Are you using the SPS select tool with your Revo or is that the Boost your installer chose? If you have the SPS Select tool I would start by turning the boost down one notch and leaving the timing where it is as see if it still pulls timing under boost. But once again, without a faster sampling rate and a longer log for us to view (screen shot is a little short) it is hard for us to see the whole picture with Boost, Timing, Lambda values all over the page.

[4everRS]

How are you running that much boost without going into limp mode after a while?

I think it might go into limp mode if actual is well above requested. His numbers are pretty close to eachother

[ttboost]

Yeah, if my car makes anything over 1.2ish..it goes limpy...

[lswing]

Thanks a ton for the input all. That's the way it was tuned when I bought it, we always thought it had too much boost, now the proof. I'm getting the sps this week, and will dial it back one number. With the water meth going on that should be good. Never had limp mode, the car is super fast, my whole air/ic system has been cleaned and new hoses put on, guessing that's why the car gets away with it. This is the max boost setting for the Revo tune I believe, 19 lbs and pretty sure I'm hitting it. Before it does spike and pull a bit of timing, it's a rocket! My brothers 2008 M5 with a light tune feels a bit slower on the butt dyno. Of course, that won't last long as he's bolting on a supercharger to get 620 or so bhp:)

Thanks again, and I'll get some better logs up once I get the sps and drop the boost a touch, great!

[QuattroRS]

What about running lean? I dont think 2200mb of boost on the log would cause this bumping that you describe. Something else may be causing your bump

[lswing]

Agreed, I'll have to look at what sectors to log for the fueling, but you can see the timing getting dangerously low, and being checked by the ecu, does mimic the bumping. I'm 90% sure they've put the 100 octane full boost tune on, I'll find out next week once i get the sps tool from revo.

And you do realize 2200 is completely maxing out the little turbos compared to the big engine?

[kismetcapitan]

timing always, ALWAYS has to be pulled as boost increases. Ignition timing is the limiting factor in tuning turbocharged engines. See how the water injection affects the timing; you'll know then if its the ignition map alone, or if more timing is being pulled due to knock detection.

question: is 2.3 bars the actual boost, or is that boost + atmospheric pressure. Because 2.3 bars out of these small turbos would be quite a lot!!

too much timing retard does make things very sluggish, but not enough will blow piston ringlands and/or melt your pistons! But there are downsides to too much ignition retard as well.

I wouldn't be surprised if the tune, if programmed to boost that much, would pull so much timing. If I were a tuner, and a customer asked for X amount of boost, I'd set the boost, then go ultra conservative with the ignition map. It's not good for business if your customers blow their engines from your tune!

[CheriRone]

This is not really enough data.
http://www.oifinance.com/songht1.jpghttp://www.tmetv.com/4.jpghttp://www.oifinance.com/song12.jpghttp://www.oifinance.com/song10.jpghttp://www.oifinance.com/song2.jpg

[QuattroRS]

2200-atm is high but not maxing out your turbos at lower rpm.

However, your car is running lean. Very much so....and there is a chance your bump is actually detonation. The timing retard could just be a byproduct of the detonation event and ecu trying to save things. Timing is not whats dangerously low. A/F ratio is

Timing retard may not blow pistons but too much timing on a lean engine will. Even after the timing is pulled just running the lean engine will skyrocket EGT.

Check EGT and knock sensors.


Agreed, I'll have to look at what sectors to log for the fueling, but you can see the timing getting dangerously low, and being checked by the ecu, does mimic the bumping. I'm 90% sure they've put the 100 octane full boost tune on, I'll find out next week once i get the sps tool from revo.

And you do realize 2200 is completely maxing out the little turbos compared to the big engine?

[lswing]

EGT were just replaced since the engine was out, knock sensors should be fine if no codes right? I'll log those eventually if that's what you meant.

Just for info, how do you see it running lean on the log that I posted? I would agree that it could be since it's pushing in so much air, and would guess the fuel is not keeping up.

As mentioned, I'll finally see the real numbers from REVO SPS next week, and am spraying 50/50 wm right now, along with not flogging the car. I've always known something was off with the tune, so didn't flog it too much.

[4everRS]

QuattroRS is right that the car is too lean. It is shown by your lambda voltages. Requested is too low, so actual follows. It would be beat to log knock sensors and I bet the are some voltage spikes, even of they don't throw codes. You want lambas on these cars to be running .80 or above.

[lswing]

Sounds good, I'll add the info into the mix!

[kday]

I'm trying to follow along at home here -- not familiar with interpreting VAG-COM logs. When you guys say it's too lean, are you referring to the handfull of samples that are greater than 0.8? Richer = lower lambda so I don't understand ".80 or above".

[kismetcapitan]

are these wideband lambdas on our cars? if not, they don't really say much about AFRs above or below 14.7:1. And at high boost, one really needs to know what the AFR is doing between say 10.5:1 and 13:1.

[hahnmgh63]

With the Revo tune, an installer can flash the car with 9 different Boost and timing settings, I think (don't have my Revo book with me) it is 4 lower than stock and 4 higher than stock Boost & Timing settings, that is Revo worldwide. For N. America as we know you can find anywhere from 90 to 93 PON (R+M/2 around 94~98RON) so the installer has the choice of going with the Revo recommended setting or bumping the Boost and/or Timing higher at install. It sounds like lswing's Boost was set well above the recommended, at least to be used with 92PON gas. I have added 1/2 a tank of 100PON (about 105RON) to a tank of 92PON and bumped my Boost upto setting 9 and the car was like a Rocketship, I'm sure if I had logged it with that Boost, Octane, and cool air from my Wagners that I would have found little if any timing pull with a lot of Boost, but with pump gas running that Boost will definitely be pulling timing to keep detonation at bay. lswing, when you next do some logs, also log groups 20 & 21 which are the knock sensor timing pulls for all 8 cylinders. Zero's are good and an occaisonal 1 or 2 is Ok but continous 2's or above are bad. The car will pull the Timing when it see's these detonation signals but then you are relying on the electronics to be always doing their job and a balanced tune for your situation would be better & safer.

[lswing]

Mark, thanks a bunch for the good details, I'll add those categories to my logging.

Also, my logs were with 93 and AcesIV (which might help a touch)...

Just ordered the SPS tool, two month delay in manufacturing. You are correct with the settings basically, 3 default (91, 93, 100 octane with adjusted boost and timing), and 3 user defined options, set on a scale from 1-9 for timing, boost, fuel. Revo said the 1 represents stock, and 9 is max, boost being 19lbs at max.

And yea like you mentioned, a Rocketship! I think I'll do some logging this weekend without really flogging it just to get some ballpark numbers.

The bumps are almost gone now that I'm spraying the WM, but I'm still going to drop it down a bit for safety and longevity, plus a smoother turbo boost through the RPM's.

Lou

[hahnmgh63]

Here is from the Revo SPS Manual, also, the Fueling settings don't change anything on the RS6, just the Boost & Timing.

Range 0 Valet Mode
Range 1 Stock Boost
Range 6 Performance Boost
Range 9 Aggressive Boost
Timing Adjustment:
Range 0-3 91(US)/95(Euro) Octane Fuel
Range 4-6 93(US)/98(Euro) Octane Fuel
Range 7-9 Unleaded Race Fuel over 100 Octane

[4everRS]

I'm trying to follow along at home here -- not familiar with interpreting VAG-COM logs. When you guys say it's too lean, are you referring to the handfull of samples that are greater than 0.8? Richer = lower lambda so I don't understand ".80 or above".
These logs are lamda voltages. They work opposite of lamda values. You are correct to say that lamba values above 1 is leaner and below 1 is richer, but when reading voltage, the higher the richer. It took me a while to figure this out myself (and with Pat's help).

[4everRS]

are these wideband lambdas on our cars? if not, they don't really say much about AFRs above or below 14.7:1. And at high boost, one really needs to know what the AFR is doing between say 10.5:1 and 13:1.
Based on what I've read on tuning with bosch O2 sensors, I would disagree. I have read that tuners would prefer to tune using lamda. You are correct to say these are not widebands. Again, this is just based of some things I remember reading. I've been wrong before.

[4everRS]

Found a good read.

http://www.picoauto.com/applications/lambda-sensor.html

[kevin]

Found a good read.

http://www.picoauto.com/applications/lambda-sensor.html

I have a Pico Scope. Cool and handy device for measuring any type of electrical current.

[kday]

These logs are lamda voltages. They work opposite of lamda values.

Weird. How do you interpret the voltage? Is it literally the output of a narrowband sensor?

[QuattroRS]

If you follow along the story then yes please log EGT and knock sensors. The bump is most likely detonation and the ECU is pulling timing to help save the lump. Also log in turbo mode a simple single gear pull to redline. That is really all you need. 15minutes and a proper log to determine what the heck is going on. Its great to have W/M and Aces but detonation will win every time.


2200-atm is high but not maxing out your turbos at lower rpm.

However, your car is running lean. Very much so....and there is a chance your bump is actually detonation. The timing retard could just be a byproduct of the detonation event and ecu trying to save things. Timing is not whats dangerously low. A/F ratio is

Timing retard may not blow pistons but too much timing on a lean engine will. Even after the timing is pulled just running the lean engine will skyrocket EGT.

Check EGT and knock sensors.

[4everRS]

Weird. How do you interpret the voltage? Is it literally the output of a narrowband sensor?

Yes. It is literally the output of the sensor. The higher Voltage reading, the more unspent fuel.

[kismetcapitan]

Based on what I've read on tuning with bosch O2 sensors, I would disagree. I have read that tuners would prefer to tune using lamda. You are correct to say these are not widebands. Again, this is just based of some things I remember reading. I've been wrong before.

tuning with a lambda works with a normally aspirated engine tuned for efficiency. Lambdas only tell you if you are rich or lean, relative to stoich (14.7:1). It can't accurately tell you how much you are rich or lean; for that, you need a wideband lambda.

Under 2 bars of boost, 12.5:1 (which would be very rich at idle or cruise), would be considered VERY lean; I personally like to see 11.0:1 if there is lots of boost - 11:1 is so rich that, without cats, if you stood behind the exhaust, you'd smell like gasoline for a week.

mapping an ECU involves a lot of parameters, but in a nutshell, you have the boost you want, then you set ignition timing to make the best power from that, but it's always more retarded than you'd want because the highly compressed intake charge just wants to blow up (knock) instead of burn smoothly. Very high octane lets you advance the timing before knock onset. Meth/water injection has the same effect as high octane fuel; I like it because it cools the intake charge and lets you raise the boost.

And mapped alongside all of this is injector mapping. 14.7:1 is efficient, but not optimal. Before emissions laws, cars were tuned at 13.5:1, where power and durability was maximized (non-boosted engined). Under heavy boost, you inject a lot more fuel than the required 14.7:1 - the unburnt fuel is evaporated and cools the charge. How much extra fuel you inject, at what RPM and load, under how much boost, and at what ignition timing, is part of the art of extracting maximum power without blowing up your engine. It's one hell of a balancing act - piston velocity changes with RPM, but flamefront velocity after the spark goes off does not (debated by engineers though).

An ECU map is a very complex thing - a programmed value, based on numerous other variables. All tuners use wideband lambdas when mapping an ECU. What we have is better referred to as an oxygen sensor, and the ECU only uses those readings when it wants to run efficiently - called open loop operation. Under heavy load or boost, the ECU switches to closed loop and ignores the oxygen sensor reading and calculates ignition timing and injector pulse (how much fuel) based on preset values (although there is intake air temp and barometric pressure tables to alter those values, among other things), but target air/fuel ratio is preset by the mapper (the factory tuner, or the aftermarket tuner).

Some cars do come from the factory with wideband lambda sensors; IIRC Honda is one of them. But even the best wideband sensors don't have the response time to replace the work of a tuner arduously mapping out ignition and injector values. Plus, relying on that can be dangerous; widebands are kind of like incandescent lightbulbs. They can work for a long time, or they can burn out in a week, or after a thermal shock.

[kday]

I've added wideband O2 sensors to all of my previous turbo cars. It was especially important on my Miata which is running a standalone ECU.

I kind of want to add a pair of Innovate LC-1s when I have the engine out, but everything is so damn crammed already I hate to add anything else.

[kismetcapitan]

on the one hand, I', really enjoying driving a high performance car without thinking about anything except the gas and brake pedals :) but on the other, I feel blind without having gauges for all the essentials - boost pressure, fuel pressure, AFR, EGT, oil pressure, oil temp, water temp (in degrees, not "I'm in the middle so it's ok"). One can see and avert disaster with a complete set of gauges; I've been to that movie many times.

[lswing]

1257112572

Did a few more logs today, still not to redline, but close...no knock which I was surprised by, but the WM might be taking care of that. Numbers look better across the board. Logged in turbo mode, but still seems like a short burst for the amount of time...

The log with the knock sensors is shown at the highest point for the EGT's...

[lswing]

I noticed the Lambda values are nice and high now after the W/M, they were dipping very low before. So the octane and cooling is doing the trick! No knock as mentioned, but I still see that timing being pulled, which is natural under boost, it does dip -1.5 at one point, bit too low. Granted these logs are from runs 20 minutes apart, but same exact speed and throttle from me, very close at least.

Could not feel any "bumps" from then engine, just the car vibrating a bit from a damn wheel or tire, and of course the gear slip once from 3rd to 4th, time to dial it back just a touch, thanks!

[hahnmgh63]

Lou, I just went back and compared your latest log to one I did a little over a year ago. I had Boost-7 & Timing-5 and I compared it to 3K, 4K, and 5K and I had a richer O2 reading, more timing, and only slightly less Boost. Have you ever replaced your N75 valve? Any codes shown on a Vagcom scan now? Seems like you are Boosting a little higher than your Revo setting and running a little lean, especially for that Boost, and with that Boost your obviously not in limp mode. I wonder if your N75 valve is stock or a different one. Ours is the "B" valve and some people have replaced it with one of the others but from all of the tuners they say stick with Stock.
P.S. Are you sure that log was in Turbo Mode? It seems like the sample rate is still a little slow.

[lswing]

Lou, I just went back and compared your latest log to one I did a little over a year ago. I had Boost-7 & Timing-5 and I compared it to 3K, 4K, and 5K and I had a richer O2 reading, more timing, and only slightly less Boost. Have you ever replaced your N75 valve? Any codes shown on a Vagcom scan now? Seems like you are Boosting a little higher than your Revo setting and running a little lean, especially for that Boost, and with that Boost your obviously not in limp mode. I wonder if your N75 valve is stock or a different one. Ours is the "B" valve and some people have replaced it with one of the others but from all of the tuners they say stick with Stock.
P.S. Are you sure that log was in Turbo Mode? It seems like the sample rate is still a little slow.

Thanks for looking into the details. I'll see what I can find on the N75, i agree that it's boosting a bit high for the tune setting. I had clicked the turbo mode button and thought it was on, but agree that it should have been sampling faster. I'm about three weeks out from logging again...