I heard you can do this to improve track times. I have read complaints on hear about the knock sensor retarding the timing halfway down the track hurting times. Anyone care to shed some light on this? If its possible to bypass it for track us only, that would be great.

 

yes, take the wire off the knock sensor, and place a 3.3k ohm resistor inbetween the wire and a good ground.

 

3.9k resistor, not 3.3k

------------------
Ed Maher - Moderator @ The TPI & Carb Boards
92 Z28 Convertible - Quasar blue / Tan top
305 TPI A4 2.73 - 14.8 @ 93.1
Stock except ported plenum and dual cats
-=ICON Motorsports=-

- Definitely prototypes, high powered mutants of some kind. Too weird to live, too cool to die

 

I heard it was a 100 ohm resistor. Is it really a 3.9k ohm one?

 

I'm positive it's 3.9k ohms, i have a KS bypass on my car right now. There are some ECMs that use a 100k ohm KS, but AFAIK, all thirdgens use the 3.9k unit.

 

Ok, been meaning to get into this for a while now. When I measure my knock sensor it measures 99.8k. All of my electronics are from an '88 GTA and as far as I know it's the factory sensor. I went to the parts store, got a new one out of the box, measured it with an ohm meter and it also measured 99.8k.

So, anyone know what it's really supposed to measure in ohms?

Thanks,



------------------
_ ___ _
SMasterson

 

OK, i stand corrected, i guess some thirdgens did come w/ 100k ohm knock sensors. I thought they were a late model thing...

 

I am finding that "neutralizing" the knock sensor @ WOT definitely helps with getting consistent performance. More over, I have found that the Knock Sensor has a tendency to go deaf. IE. The Knock Sensor works when you can't hear it, and the KS stops working when you do.

Further, once the knock sensor stops working, then the ECM will give you the MAXIMUM spark advance it can, which further compounds the problem.

I found finding the ideal spark advance and capping the KS to 1* works best for WOT, while letting the KS do it's normal job at part throttle. The advantage to this method is that it allows you to attain max performance, but still have the protection of the KS for part throttle driving under load conditions (up a hill) or for a "bad tank of gas".

It involves going inside the eprom, but it's a lot cleaner and only disables the KS under WOT and lets it function for all other conditions.

You will most likely find that with the stock GM heads, that there is a very narrow "sweet spot" for spark advance where a couple degrees less makes your engine "doggy" and a couple degrees more causes ping/detonation. On my car, somewhere between 28-30* gives the best results.

Also, remember that the optimum spark advance changes with elevation. Low elevation needs the least and higher elevation more. If you set you car up for higher elevation (1,000+ feet), you may find that your car will ping at sea level. This is why I prefer to do it "inside the eprom" as you can set up your car to operate optimally for all elevations.

 

I noticed an interesting way that ADS Superchips has used to "override" the knock sensor. They intentionally set the advance in certain MAP/rpm points excessively high which causes the ecm to remove timing to the "max retard" value. This point is still a few degrees higher than the point the KS trips. Here is an example: I am working on my spark tables in low rpm region where you launch the car from a dead stop. At 1800 rpm and 100 kPa, ADS calls for 28 deg(main table) plus 4 deg(PE adder) + 6deg base timing. This is 38 deg advance at just 1800 rpm!!!! Way too much. If you look at there max knock retard value in PE mode, it is 20 deg. So, 38-20 = 18 deg advance. I personally cannot run near that much advance in my engine without pinging. WHat I figure is that this low rpm range is only seen for a very short time(dead stop to 20 mph) so the engine can tolerate a little detonation without it being damaging. Also it is low rpm only where ADS runs this high advance, above 2000 rpm the advance gets taken out.

------------------
Dave Zelinka

 

Those "high desired spark advances" with "wide knock retard potential" is one of the reasons a car can feel very inconsistent in its performance. What is worst, should that knock sensor actually encounter audible ping, the knock sensor will cease to function and then the ECM WILL advance your timing to 38*...making the detonation even worst and possibly damaging your motor.

I did a lot of experimentation, and assuming your car can tolerate 24* of effective total timing, and gives its best performance without pinging, you are better off to set your MAX timing to 24* and set the knock sensor to 0-1*. This way, if you should encounter audible ping, and the knock sensor stops working; the ECM will only advance you to 24*. Much easier to recover from (and have the knock sensor funciton again) and less chance of permanently damaging your motor than 38*.

But, this is only part of the knock sensor's job. I find you want your knock sensor operating for typical part throttle driving (which totally disabling defeats). Nothing worst that travelling to another part of the country, get a bad tank of gas (it happens) and then pinging like hell up a long grade; where you often don't have the luxury of "backing off the throttle" like you would in a WOT situation.

So that is why I prefer doing it inside the eprom, I can disable my KS @ WOT (which I am finding is the best approach for consistent performance), yet have it function normally for all other types of driving. You get to have your cake and eat it too.

 

Ok, I am beginning to understand this some.
Sorry to look like the rookie here. So now the knock sensor would be completely disconnected? Will it throw a code? I guess the resistor would fool the computer? And where is a good place the resistors at? Mineis a 89 Iroc. The car was a origional 350 but I have a temporary 305 in there now with a 305 KS. I need a quick fix. I want to look into doing it by the chip later.

 

The resistor will "fool" the ecm. Without one, it WILL trigger an SES code and you will fall into "Fail Safe" Mode. Your car will run (sometimes fairly well), but not optimally when in "Fail Safe" Mode. Fail Safe causes the ECM to ignore many sensor readings and use "pre-defined settings" within the eprom.

Mose people confuse this with "Limp Mode". Basically, Limp Mode gives you enough power to pull over to the side of the road...you will barely run.

 

Heck, I didn't mean to "correct" anyone! I was asking a question. (Big Grin)

I'm running 11:1 with aluminum heads and the Snap-on scanner NEVER shows me any knock retard. That's why I was measuring the sensor to begin with. I didn't think it was working and intended to replace it. But, knowing the auto parts guy as well as I do after sinking 6 grand into this project, he let me measure a new one before I purchased it.

I've only heard audible knock a couple of times and that was before I upgraded to 30# injectors. I was accelrating heavily and it didn't shift down to second.

So, anyone else got the 100k knock sensor?

------------------
_ ___ _
SMasterson

[This message has been edited by SMasterson (edited September 27, 2001).]

 

This "knock problem" seems to be something that plagues stock GM heads. I use to think only the iron heads were prone, but I have been told the Corvette heads are also prone.

Aftermarket aluminum heads are not nearly as prone. My friend with a 383 Miniram with AFR Competition 195s gets virturally NO knock whatsoever (audible or sensor). The only thing about his heads is he has larger combustion chambers.

With his heads, he MUST run higher spark to get the power. He hasn't "dialed in" the optmum spark advance, but his heads are easily handling 36+* with no knock. The only time I can run that high is with a functioning EGR @ part throttle. But no way @ WOT.

PS: He doesn't run an EGR due to his Miniram.

 

oops. thanks for pointing that out. atleast i was close



[This message has been edited by u r sofa king we tah did (edited September 27, 2001).]

 

hey glenn, i just had a thought, what if you were to put a resistor inbetween the knock sensor and the wire that is lower than 3.9k ohms, say 1k ohms. would it pick up less knock, and take a real good spark knock to set it off? just a though. also, i can safely run my 88 corvette at 12BTDC with no spark knock verified with a scanner, and thats with 89 octane. my iroc w/ iron heads i cant get higher than 8BTDC with 93 octane.

[This message has been edited by u r sofa king we tah did (edited September 27, 2001).]

 

UR, there is another issue with timing, in that with GM heads you don't need a lot of timing...they were designed that way.

Generally, larger combustion chambers want more timing and small combustion chambers less. The real key is to find the spark advance that yields the best performance. And if you can get the same performance from less advance, do it...your engine will prefer it.

There is discussion about a DIY "tuneable" KS to tuneout "other mechanical noises" that can give false readings. Guys have found relocating the KS can eliminate certain "background" noises like loose lifters, exhaust leaks, transmission noise, etc. But currently, I still prefer "tweaking" the eprom so you can still have it function for "day to day" driving for "hills/loads and possibly bad gas".

The thing to remember about the KS, is that it is a "first line of defense". It works before your ears hear it. Once you can hear it, the KS stops working and that is when the problem begins.

I am also not a fan of just "altering" the distributor base, as it can cause other problems like idling and low rpm acceleration as it can "bring on" the spark advance too soon, thus inducing more knock. These engines can be prone to knock at 1,600-2,200 rpm with a moderately heavy load.

[This message has been edited by Glenn91L98GTA (edited September 27, 2001).]

 

Glenn: I am not so sure it is aftermarket heads versus factory. I run Edelbrock aluminum heads on my 383 and I cannot handle alot of advance AT ALL. 30deg is absolute max for me at WOT high rpm. I run 22-28 deg WOT 2200-4500rpm. I intentionally set up my motor with a tight quench to aid mixture motion in the chamber which is supposed to increase burn rate as well as decrease detonation by brining all points in the combustion chamber to a more even temp(no hot spots) My quench height is 0.041". This is a very interesting topic as I too have seen some inconsistant performance from my setup. When I had the ADS chip in mine, sometimes it would launch REALLY good, and other times not so good. Since I remapped the whole spark table myself based on watching knock retard on the Diacom acceleration runs at several different MAP values, my engine performance is MUCH more consistant, but my launches are not as hard as they have been in the past. I am right now trying to figure out how to get around this. To give you an idea how different my spark numbers are now versus the ADS chip, at 2000 rpm 100kPA, ADS used 27deg in the table, I use 11 deg to get no "knock retard" from the ecm!!!

------------------
Dave Zelinka

 

Yelo, it seems the size of the combustion chamber (along with 'swirl and tumble') are contributing factors. Edelbrocks have a smaller combustion chamber than stock L98 iron heads. My buddy's 383 has 72 cc combustion chambers and he has no knock whatsoever @ 36+ degrees.

The bottom line, is that some heads just don't like a lot of spark advance and they were designed that way. The best guide is to find out what advance gives you the best performance; and use the lowest spark advance possible that yields equivalent performance.

FYI, the majority of my 200+ eprom burns have been to "test" my optimum spark advance, under a variety of conditions. I have even found elevation (due to air density) to have a major effect. I live at 1,100' and had my spark advance tuned to the point where 1/2 more degree would induce audible ping. I then took a trip to the Coast (sea level) and I was pinging whenever I booted it. So I had to readjust the eprom further to compensate for the denser air by backing off the spark advance a degree. Further, with higher elevation, I am finding I can run even more advance...I haven't hit the limit yet and I am still "tweaking" that area.

With a simple "bypass" of the KS, you loose that "tuninig" ability, as the only way to compensate is within the eprom.

 

Also, Yelo, try getting the spark advance up quicker @ low rpm. You are probably finding the same thing I am; lower the advance will eliminate the knock, but your engine feels "doggy", advance it and you engine will "liven up" but may induce audible knock.

It is a balancing act that takes a LOT of testing to find that "sweet spot"; and if varies through out the rpm range. But it is definitely the one area that really affects the engine's performance. Fun testing too.

Lastly, have you noticed that a lot of knock seems to occur during an upshift? This is generally where I hear audible ping first. I am working on altering the actual instructions in the eprom to purposely retard the spark during an upshift.

[This message has been edited by Glenn91L98GTA (edited September 27, 2001).]