1G...tune to knock?

[KeltonDSMer]

I've heard so many opinions on this and I was wondering what you all thought. Should 1Gs be tuned for no/very little knock, or should they be tuned to O2 values and/or EGTs? Timing advance is controlled by the knock sensor and a few other factors, so is it best to look at knock and timing? If there is "phantom knock", is there a way to calibrate the knock sensor to not pick up the false signal? Oh, and is timing advance ever too high to be safe if there is no knock; I ask this because people have said this can blow the HG even if there is no knock present.

Thanks guys,
Kelton

[Dallas J]

Those people who blew their head gaskets probably had a healthy tune and mid 20s boost. The stock studs and headgasket is pretty good up to about 25psi where it starts to stretch the head bolts/studs and then will blow out the head gasket.

Up here you can add a little extra base timing, I went with 7degree's on my old 1g and it worked fine. Since im betting you dont have any other timing control than the base you should tune for a low knock count. You cant really believe anything that the stock o2 sensor is telling you. Some cars will like .98v others will like .89 volts...If you want to tune by it you could find out where you start to knock at and use what the o2's are right their as a guide to tune the rest of the rpm band.

I hope this helps.

-Dallas J

[sbiggi]

o2 readings off a narrow band are shit....
I get more knock at .9v then at .88v on one log....
then more knock at .9v on another log....
to tune for o2... get a wideband

and in my case I had to tune for some knock.... with a severly hacked 1g maf, and 680cc injectors at 0 knock I was getting 27 deg timing advance

tuned for 0-5 knock.... timing went back to normal getting max of 23 deg but ending up at 20 by redline

-seth

[KeltonDSMer]

So are narrow band O2 sensors not reliable alltogether, or is it just the stock O2 value that you can't count on? I ask this because it seems like they would work just fine, it's just that the stock ones get old and aren't accurate anymore. Do new stock sensors give accurate numbers or is a wideband the only way to go?

I like the thing about looking at what O2 value things start knocking; that makes a lot of sense to me. Even if there is no knock, if you are able to know at what value things start to knock, you can lean things out safely without risking detonating the motor.


So as long as the ECU is advancing the timing a good amount, (18+ degrees), and there is no knock, I don't have a chance of screwing anything up, do I?

Thanks again fellas,
Kelton

[fusionsport]

narrowbands do not have enough resolution to be accurate in the range we need them to be extremely accurate. There are numerous articles on this-
A wideband has a much better resolution in the ranges we need, and is far more accurate in general.

Tuning to knock largely depends on the sensors used and the motor they are used on, and what the ECU does with the information

[KeltonDSMer]

I've heard of calibrating the knock sensor somehow to make it only pick up true audible knocking to make it more accurate, but I have no idea if this is real or how it is done. Can someone elaborate on this?

[ImolaFem]

Originally posted by KeltonDSMer
So are narrow band O2 sensors not reliable alltogether, or is it just the stock O2 value that you can't count on? I ask this because it seems like they would work just fine, it's just that the stock ones get old and aren't accurate anymore. Do new stock sensors give accurate numbers or is a wideband the only way to go?

For a difference of 1-12 hp you'll find people will get their cars on dynos and hook up the WB o2 and tune their happy hearts out. ;)

However, once you do start logging your car THEN tuning -- notice, get a GOOD baseline before messing with things. You'll notice what o2 voltage your car gives when it produces knock, so on and so forth. Your stock o2 is not very precise and it won't be the same as anyone elses, but once you get used to it, you'll know what's safe for your car.

Also, I've never heard of anyone calibrating the stock knock sensor on a stock ECU. I believe this is an option on the AEM. It might be in DSMLink v2, but I don't think so.

[DlandryTSI]

On a side note anything under 7 counts of knock and you'll be fine. Tuning for no knock will get you some pretty bad performance.

Another MAJOR knock contributor is heat soak and high IAT's.

On my old setup I was running under 7 and it was running nice and strong.

--Dave

[KeltonDSMer]

Yeah, I understand high intake temps making a lot more knock. I am definately going to be using a big FMIC core or at least a WI system to keep intake temps under control.

I thought the ECU stops advancing the timing if the knock count goes over five or something. I forget where I read this but it also said the ECU would continue advancing the timing if the knock sensor picks up less than 3 counts. I'm not going to have timing control over the RPM range, so whatever I can do to minipulate it in my favor is good with me.

I guess my question is this:
Would the car be quicker with little/no knock to advance the timing and a richer air/fuel ratio; or would it be quicker with around 7 counts and a leaner ratio?


Thanks a lot guys this really helps me out,
Kelton

[DlandryTSI]

Yeah I think you are right. 5 counts and under and the ecu ignores it. 6-7 and I think it pulls 2 or 3 degrees of timing.

Its been a while since I've used a 1g ecu ;)

Try to tune for 5 and under. As long as your timing is steady at around 21-23 degrees you'll have better power than with 6 or more.

--Dave

[KeltonDSMer]

Yeah, that's what I thought too. Thanks a lot for all the help, I really appreciate it.

Oh, does anyone have experience/knowledge about using the VPC? I am trying to find a way to buy this 91AWD tsi from a real cool guy named matt in thorton, and it has a VPC. I never learned about them because I figured I wouldn't be buying a $1000 fuel computer. I like how it usees a non-restrictive air sensor and that you can read all the other sensor values it uses, (like intake temp and such). He is a smart guy for sure and posts pretty frequently at Buschur Forums, so I know it hasn't been sketchly modified.
He lives at similar altitude too, so I wouldn't have to tune anything after bringing it to Boulder if I'm able to get the car. It has only 89000 miles on the odometer, and a rebuilt head, (freak accident where the A/C belt broke and got cought in the timing belt, bending some valves). Is there a chance of piston damage from valves bending?, because I know the 1G pistons are releatively soft.
He seems like a really good guy that wouldn't try to pull anything weird, so I don't think he would leave anything out if something else was wrong with the car.... Anyone know Matt?

Thanks again guys,
Kelton