Our Community is over 1 Million Strong. Join Us.

Carnivore Diet for Dogs

AIR DRIED BEEF DOG FOOD

Crazy knock in 3rd, o2 sensor or knock sensor bad?


nates6969
06-21-2006, 01:05 AM
Ok so I just finished converting my n/t car to turbo and running about 8 pounds of boost. Its great however I ran some logs the other day and noticed I was getting some rough knock in 3rd, 1st and second arent bad

heres my log...
This is a 3rd gear pull.
TIMA - FTRL - FTRM - FTRH - O2V - TPS - RPM - Knock
18° 139.6% 139.6% 113.9% 0.92V 100.0% 4531 16
14° 139.6% 139.6% 113.9% 0.92V 100.0% 4656 15
14° 139.6% 139.6% 113.9% 0.92V 100.0% 4688 17
14° 139.6% 139.6% 113.9% 0.92V 100.0% 4813 17
15° 139.6% 139.6% 113.9% 0.92V 100.0% 4844 16
15° 139.6% 139.6% 113.9% 0.92V 100.0% 4906 16
16° 139.6% 139.6% 113.9% 0.90V 100.0% 5000 16
15° 139.6% 139.6% 113.9% 0.92V 100.0% 5094 15
16° 139.6% 139.6% 113.9% 0.92V 100.0% 5156 15
17° 139.6% 139.6% 113.9% 0.90V 100.0% 5250 15
17° 139.6% 139.6% 113.9% 0.90V 100.0% 5313 14
17° 139.6% 139.6% 113.9% 0.90V 100.0% 5406 14
18° 139.6% 139.6% 113.9% 0.90V 100.0% 5438 14
18° 139.6% 139.6% 113.9% 0.90V 100.0% 5469 13
18° 139.6% 139.6% 113.9% 0.90V 100.0% 5594 13
17° 139.6% 139.6% 113.9% 0.90V 100.0% 5625 19
18° 139.6% 139.6% 113.9% 0.90V 100.0% 5750 22
17° 139.6% 139.6% 113.9% 0.90V 100.0% 5750 22
17° 139.6% 139.6% 113.9% 0.90V 100.0% 5813 22
17° 139.6% 139.6% 113.9% 0.90V 100.0% 5875 25
17° 139.6% 139.6% 113.9% 0.90V 100.0% 5938 25
17° 139.6% 139.6% 113.9% 0.90V 100.0% 6031 25
17° 139.6% 139.6% 113.9% 0.90V 100.0% 6063 24
17° 139.6% 139.6% 113.9% 0.90V 100.0% 6063 24
17° 139.6% 139.6% 113.9% 0.90V 100.0% 6156 24
17° 139.6% 139.6% 113.9% 0.90V 100.0% 6188 25
17° 139.6% 139.6% 113.9% 0.90V 100.0% 6250 24
17° 139.6% 139.6% 113.9% 0.90V 100.0% 6281 24
17° 139.6% 139.6% 113.9% 0.90V 100.0% 6344 24
15° 139.6% 139.6% 113.9% 0.90V 100.0% 6344 24
16° 139.6% 139.6% 113.9% 0.90V 100.0% 6438 23
17° 139.6% 139.6% 113.9% 0.90V 100.0% 6469 23
16° 139.6% 139.6% 113.9% 0.90V 100.0% 6500 23
16° 139.6% 139.6% 113.9% 0.90V 100.0% 6500 22
15° 139.6% 139.6% 113.9% 0.90V 100.0% 6563 22
16° 139.6% 139.6% 113.9% 0.90V 100.0% 6594 22
15° 139.6% 139.6% 113.9% 0.90V 100.0% 6625 22
14° 139.6% 139.6% 113.9% 0.94V 9.8% 6844 14
42° 139.6% 139.6% 113.9% 0.84V 9.8% 6250 0

I'm very new to tuning but my timing looks very high for having so much knock, is that right? I know the base timing is good, I just set it to 5 degrees. Now my question is could this be reading faulty knock values since my knock sensor is in rough shape (nice drops of sticky goo!) cuz I also have a bit of phantom knock off the line. But then I also wonder if it could be a shot o2 sensor? when I'm 'cruising', I was looking at a log of my o2s and half the time the graph bounces smoothly up and down but then sometimes all of a sudden for upwards of 8-10 seconds the voltages drops to 0.01 with no change in the TPS or load on the engine (flat road). Then at idle with engine fully warmed, the o2s drop to 0.01 again and then all of a sudden my car seems like it starts flooding itself, the rpms drops then all of a sudden it reads rich and the rpms kick up. it will continue to do this until the I accelerate away. My fuel trims are also way too high! I have a turbo pump and 450s, old fuel filter though. but still im only running max 8psi boost. also almost no boost leaks in the system, a very minor one through TB seals but stills holds a decent pressure. What do you guys think? knock sensor bad? o2 sensor bad? fuel filter bad? any other possibilities?

Thanks for your help
Nathan

crunchymilk55
06-21-2006, 01:33 AM
well I'm going to answer some of your questions, but keep in mind ALL my knowledge is from a 4g63 perspective so I'm not sure what directly translates over and what is different. First off, just looking at the log there's some huge problems, obviously the knock values are crazy high, but your timing isn't even really a curve. It shoots straight up then bounches up and down a bit. I'm not sure how it's supposed to behave on a NT to turbo conversion though.

With the knock values going up like that, it doesn't seem to me to be phantom knock. Since it gets worse as the rpms go up, it almost looks like lean knock. What are you using to tune? I would add fuel through the whole powerband and see what that does to timing and knock. What ecu are you using? If this was a regularr 4g63 I would say might as well replace the knock sensor just to make sure, the 02 sensor seems to be working fine since you said it normally cycles. The signal, however, you say is staying up or down sometimes which tells me it's either on its way out, (do you run lots of race gas?) -OR- it is being truthful and you really are having a/f issues.

Honestly, I think checking out another forum that deals with more NT to turbo conversions is your best bet. Hope that helped a little bit.

kjewer1
06-21-2006, 02:46 AM
The knock looks real, though it can be hard to tell. A really bad sensor can knock all the time, not giving the traditional knock decay that phantom or mechanical knock gives. The timing curve is all over the place because the knock retard is interfering with it. Overall though it looks like a lower version of the stock 1g turbo timing curve. Unlike the 2g, the 1g curve drops timing after 6000 rpm. The 2g curve is far superior, and whenever I tune a 1g I shoot for a 2g timing curve to get started, and maintain the shape of a 2g curve even when raising timing for race gas use. What ECU are you using?

Another thing that makes be believe the knock is most likely real is that it gets worse in the higher gears. Load is higher in higher gears, and the parts are also more heatsoaked. What IC are you running?

Your timing is roughly where you would like to be. What I would do is retard base timing to zero degrees and see what that does for knock at WOT. It will hurt cruise performance and maybe even fuel mileage slightly, but if the knock goes away you know it's real, and you need some way to control timing to avoid knock/damage. The trick to keeping this stuff fun is keeping it safe. So I'm glad to see you are datalogging. Now you just need some way to make changes based on what you see in the logs ;)

A leaking knock sensor should definitely be replaced. I woiuld do that ASAP, they are only 40-50 bucks. A bad O2 sensor will hurt mileage and emissions, but has nothing to do with WOT. .90 volts looks pretty normal for a 1g, for whatever that is worth. I don't put any weight at all in stock o2 sensor voltage, I've seen mine sit at .98 volts whether I'm at 10:1 or 13:1 ;) Your trims are maxed out high, which means the ECU thinks it's running lean in closed loop based on what it sees from the front O2 sensor. If there is no reason for it to be lean (injectors match the ECU, airflow metering is accurate, no vac leaks, etc) the O2 sensor is the most likely culprit.

nates6969
06-21-2006, 11:43 AM
Thanks a lot for the quick replys! just to answer some of your questions, the engine I have is the 4g63n/t. its fully converted to a turbo engine minus the 9:1 compression, so thats why I'm running low boost. as for the injectors, mafs, fuel pump, fuel regulator, intercooler, ecu...they are all out of a 92 tsi M/T AWD. (my car is a 91 by the way but has all the same basic sensors as the 92). the computer is a 6262. I know the normal tsi FWD use the 6260 computer in that year but appearantly doesn't really make any difference. Also as for the intercooler there was this nasty peice of sheet metal covering most of it in the front, looks like it hold the stock fog lights, so I cut the out and now I get some crazy airflow threw it. I did a pull on my way to work today in second gear and it was only reading 0-1 counts of knock at 6000. I'll have to try a 3rd gear later.

As for the knock sensor I totally agree and am going to order one today! its pretty nasty looking. What do you guys think about the fuel filter though? this could also cause a lean situation if im not mistaken. although in the high end my injectors are only running ~70% max duty cycle. how often have you guys seen a filter cause a lean burn?

crunchymilk55
06-21-2006, 01:05 PM
Personally I've never seen it first hand. It sounds like you were essentially running without an intercooler now. Show us another log now that you have some cooling going on and we'll see if everything is all set.

O yeah, when you do the log please start aroudn 3krpm. It makes reading the timing curve alot easier, at least for me.

nates6969
06-21-2006, 01:47 PM
Ok sounds good, ya sorry about the earlier log, I was logging from 1st thru 3rd and when I shifted 3rd I was at 4600 rpm. I'll do a 3000-redline log this afternoon, its pretty cool out today so that might help. I'll post it this evening

Thanks again

kjewer1
06-22-2006, 01:26 AM
Sounds like it was indeed heatsoak, which makes sense now. ;) I've also never seen a fuel filter cause any problems, even with 150k miles on them. You won't be able to gather anything from injector IDCs though, the ECU doesn't have any AFR feedback at WOT so IDCs will be the same even if you don't have a filter. Or a fuel pump. Or injectors for that matter. :D

crunchymilk55
06-22-2006, 09:42 AM
kevin, IDC's are just a guess??? That's news to me, I thought they could measure the impulse sent to them or something...

kjewer1
06-22-2006, 11:28 AM
They are what the ECU is commanding the injectors to do, but has nothing to do with what is actually happening. I see over and over again cases where people think something like a pump running out of capacity of a clogged filter will increase IDCs. The ECU doesn't know that these things are going on, IDC won't change. :) It's not a guess, as in it is a real measurement. Or more accurately, it's what should really be happening, since it's not a measurement at all. Another example is injectors with different dead times. The ECU still only displays what it TOLD the injectors to do. With different dead times the time that the injector is actually open will be different than what the ECU requested. Hopefully that makes some sense...

corning_d3
06-22-2006, 12:01 PM
Doesn't a high O2 voltage indicate a rich condition? I hope you installed some good double plat. plugs, new wires and coils. I think the knock is from too much compression, or carbon build-up. 8.5:1 is more reasonable for boost. You may want to remove the head, and clean the combustion chambers. This is also a good chance to install a high performance head gasket(since you are boosting...). Also, if you aren't running synthetic already, it may be a good time to switch. Turbo's are hell on conventional oils..

nates6969
06-22-2006, 12:56 PM
Hey guys just an update, couldnt get a good log last night, it was raining and everytime I'd step on it in 3rd gear the damn tires would break loose! haha. I will get one up tonight
to answer some of corning_d3 questions, I actually rebuilt the motor with new rings, new but OEM gasket, surfaced head, 8 new valves (bent) and then had all the valves reseated about 20, 000 miles ago so the motor is pretty strong so im hoping this head gasket, because it is fairly new, can handle the extra compression. if not, i'll throw a better one in and since the motor was recently rebuilt I'm thinking it should be pretty clean but who knows, I'll probably wind up running some of that seafoam combustion cleaner through, if i can find some, i hear thats really good. but I don't think thats the problem, my engine has alway run really clean and been great on gas so it shouldnt have much build up.
Also I was looking at a log today of just cruising and my o2 seems to be running perfect when cruising around 2200 but then when I drop to an idle my o2s also drop to zero which explains why my low fuel trims are at 139, and my mids have gone down to 104 which is fine by me. I wonder if their is a vacuum leak causing this. I do get a slight bit of idle surge for a few seconds sometimes when cold but that could be a whole other story. my mafs is also hacked by the previous owner so I'm going to try a buddies in my car sometime soon. is there anything else that could be causing this lean condition?

kjewer1
06-22-2006, 11:42 PM
Hackes MAS could do it if not compensated for properly. I won't even get into my fun with hacked 2G MASs back in the day. DSMlink makes it easier though... It has it's biggest effect at low airflow though, so trims will certainly be affected. I don't think 9:1 is a problem, the EVO is 8.8 stock and seems to work fine. Most of my built motors were 9-10:1.

nates6969
06-22-2006, 11:58 PM
Alright guys heres the new 3rd gear pull log from 3000-6500. This is the log with Intercooler air flow opened up but still with a knock sensor with a ton of goo dripping out. (replacement will be in tomorrow:) !)

let me know what you think, IMO the timing seems far to advanced for pulling that amount of knock

Timing - O2v - TPS - RPM - KNOCK - AIRTEMP
29° - 0.74V - 40.8% - 2906 - 0 - 88.7°F
28° - 0.47V - 43.9% - 2969 - 0 - 88.7°F
27° - 0.80V - 65.1% - 3031 - 0 - 88.7°F
25° - 0.86V - 90.2% - 3094 - 0 - 88.7°F
24° - 0.88V - 100.0% - 3188 - 0 - 88.7°F
23° - 0.90V - 100.0% - 3281 - 0 - 88.7°F
22° - 0.92V - 100.0% - 3406 - 0 - 88.7°F
19° - 0.92V - 100.0% - 3469 - 0 - 88.7°F
20° - 0.92V - 100.0% - 3594 - 2 - 89.6°F
20° - 0.92V - 100.0% - 3719 - 2 - 89.6°F
20° - 0.92V - 100.0% - 3813 - 1 - 89.6°F
21° - 0.92V - 100.0% - 3844 - 1 - 90.5°F
21° - 0.92V - 100.0% - 4000 - 0 - 90.5°F
21° - 0.92V - 100.0% - 4063 - 0 - 90.5°F
21° - 0.92V - 100.0% - 4156 - 0 - 91.4°F
21° - 0.92V - 100.0% - 4281 - 0 - 91.4°F
21° - 0.92V - 100.0% - 4375 - 0 - 92.3°F
21° - 0.92V - 100.0% - 4500 - 1 - 92.3°F
21° - 0.92V - 100.0% - 4594 - 1 - 93.2°F
20° - 0.92V - 100.0% - 4656 - 0 - 93.2°F
20° - 0.92V - 100.0% - 4781 - 0 - 93.2°F
20° - 0.92V - 100.0% - 4875 - 0 - 94.1°F
22° - 0.92V - 100.0% - 4938 - 0 - 94.1°F
21° - 0.92V - 100.0% - 5000 - 0 - 94.1°F
22° - 0.92V - 100.0% - 5094 - 0 - 95.0°F
22° - 0.92V - 100.0% - 5156 - 0 - 95.0°F
22° - 0.92V - 100.0% - 5250 - 0 - 95.0°F
23° - 0.92V - 100.0% - 5344 - 0 - 95.9°F
23° - 0.92V - 100.0% - 5375 - 0 - 95.9°F
23° - 0.92V - 100.0% - 5469 - 1 - 95.9°F
23° - 0.92V - 100.0% - 5500 - 1 - 95.9°F
23° - 0.92V - 100.0% - 5594 - 1 - 96.8°F
24° - 0.92V - 100.0% - 5656 - 0 - 96.8°F
26° - 0.92V - 100.0% - 5750 - 0 - 96.8°F
26° - 0.92V - 100.0% - 5750 - 0 - 96.8°F
27° - 0.92V - 100.0% - 5781 - 4 - 96.8°F
25° - 0.92V - 100.0% - 6031 - 11 - 96.8°F
24° - 0.92V - 100.0% - 6000 - 11 - 96.8°F
24° - 0.92V - 100.0% - 5938 - 10 - 96.8°F
24° - 0.92V - 100.0% - 6063 - 10 - 96.8°F
24° - 0.90V - 100.0% - 6125 - 10 - 96.8°F
23° - 0.92V - 100.0% - 6156 - 14 - 97.7°F
22° - 0.92V - 100.0% - 6156 - 14 - 97.7°F
22° - 0.92V - 100.0% - 6250 - 14 - 97.7°F
21° - 0.90V - 100.0% - 6344 - 13 - 97.7°F
21° - 0.92V - 100.0% - 6375 - 13 - 96.8°F
20° - 0.92V - 100.0% - 6406 - 13 - 96.8°F
20° - 0.90V - 100.0% - 6438 - 16 - 96.8°F
19° - 0.92V - 100.0% - 6469 - 16 - 96.8°F
18° - 0.92V - 100.0% - 6531 - 16 - 96.8°F
18° - 0.92V - 43.5% - 6594 - 15 - 96.8°F
35° - 0.90V - 9.8% - 6469 - 0 - 96.8°F


As for my o2 sensor, im going to hold off on that for now cuz that seems to be working ok. When my cars has only been running for about 15 minutes it idles around 1100 and the o2s fluctuate nicely, its not only till its full hot, the idle drops to 750 and then the o2s read 0, I think its because my mafs is hacked and the air is taking the path of easiest route.

but anyways, let me know what you guys think!
thanks a lot

kjewer1
06-23-2006, 12:26 AM
Holy shit, way too much timing! Any idea why it's so high? A couple frames after hitting 24-27 degrees the knock starts. That's real knock for sure at that level of timing.

crunchymilk55
06-23-2006, 12:33 AM
yeah no kidding. I know 1g's run higher timing but that's crazy. What is your base timing at? It's an obvious pattern, knock kills the timing up top. IMO it has to be an air metering problem, your car isn't acknowleding all the airflow it's getting. What are you using to tune? What is your injectors size?

please put your mods and car info somewhere in your sig or your profile, it makes things easier for us

nates6969
06-23-2006, 12:51 AM
Ya see I wondered about the air flow issue. the previous owner of the turbo mas hacked the hell out of it and now at idle it runs very lean, low fuel trims at 139 (maxed), mids at 120, and highs at 100.4 right now but computer was reset a day ago and hasnt had much time to adjust since im not constantly ripping it in the car. Right now I have nothing to tune with! and the main reason I don't is im trying to run a stock setup for now, im only running 8 psi which with my 9:1 compression works out to be just over the 'boosted compression' of a stock tsi. (boosted compression = cylinder compression * ((boost psi/14.7) + 1). Therefore I should be obtaining the same fuel trims as a stock tsi. correct me if I'm wrong. but since my mas is hacked I think its giving me a lean situation at all rpms. I'm hopefully getting my hands on a unhacked this weekend just for test. But as for that IC opening up, it did really seem to help! I'm going to get some sheet metal and build a little tunnel for it this weekend to help direct all air flow on it. Also to answer your other questions, I actually set the base timing a couple days ago dead on 5 and my injectors are stock 450s matched with a m/t ecu and m/t fpr

kjewer1
06-23-2006, 10:11 AM
I wouldn't go running around at WOT all the time until you get the timing figured out. Part of it could be the super low boost, but I would wait until you test with a stock sensor.

Add your comment to this topic!