Hey everyone,
Well I'd prefer not to blow up my engine because of GM's little knock test, so I was wondering how I would go about disabling this. I'm running the '8746 with a Prominator on er'.

Tony

 

If you disable the code 43 in the prom the forced knock test will no longer be performed.

 

I've never heard of a problem traced to the K/S test, and I really doubt the K/S test would hurt an engine. Not to mention sacrificing the reliability it adds. It just adds some timing for a moment and listens make sure it gets a response from the K/S.

 

See the 'Patch Method Demo' sticky at the top (thread may be in another sticky). It has a patch to disable the forced knock test and another for a rev-limiter. Both handy to have.

RBob.

 

Isn't that response detonation though? I'd only run with it off when I had the juice turned on.

 

What?.

The K/S TEST, adds a few degrees timing, and then listens for a response from the knock sensor.

Any detected detonation from the K/S forces some reduction in timing (obviously when all the enables are meet)..

 

The only response the knock sensor can give is "there's detonation present" though. I'd prefer not to take it to the point where that occurs.

 

All the self tests done in the life time of the engine, wouldn't equate to the damage done from one tank of garbage gas, without the K/S.
Not to mention what an asset it is in tuning an engine.

With the risks of detonation, from running Nitrous I really can't understand why you'd want to remove anything, that mins the risk of detonation.

 

The forced knock test on a stock engine may be OK. A few pings here and there won't hurt anything. At times the test can fail to produce knock which then causes other problems.

On a modified engine the forced knock test will scare the boogers out of the driver. The ECM waits until it is in PE mode and adds a ton of timing (ca. 10 deg). If that doesn't make it knock the next time it adds more timing (ca 20deg). I had this test done on my mild 327 build once, many years ago.

And once was one too many. Sounded like all eight cylinders were being beat on with a sledge hammer. I couldn't lift fast enough.

Moving on to the newer ECMs ('90-'92 SD) GM decided to measure the resistance of the knock sensor. This is a better test of the integrity of the system.

Back to the current posters question: running nitrous and a forced knock test, and I'll bet the tops of the pistons end up in the bottom of the pan. There is little control over when the test is run. It becomes a 50-50 chance of way too much timing while the juice is lit, or being just OK.

RBob.

 

Does the K/S test bypass patch leave the K/S active?.

In a nitrous appl., it'd just seem prudent to make sure the K/S is alive and well. Maybe, set the test timing WAY down, rather then looking to hammer the engine. And/or lowering the thresholds to a real min., to get the test out of the way, before the nitrous is enabled?. ie 25%, when in fact the TPS enable is say 90%.

 

Yes. Also, the knock run-away test is active. Other then the ECM forcing knock the ESC system is left intact.

RBob.

 

Been running that patch since it came out. Love it! Also, with the 8746, I noticed that the aldl is spitting out msb. This isn't a good indicator of knock, the lsb is the real deal. You'll get knock retard just from the lsb before it adds to the msb and you'll be sitting there scratching your head. So do NOT go by the ALDL knock counter, that's only after the ESC picks up 255 "small" counts. Last time down the 1/4 I had zero msb counts but 20 degrees of knock retard in all of 3rd gear. Needless to say lockers has saved my engine and my sanity.

 

IMHO, the k/s is an enemy of racers........ANYTHING that can reduce consistency is. What I program should be applied w/o correction every time. This summer I'm going to run a dry nitrous system. The plan is to add 25% fuel and reduce timing 2*. Then lean it out to just to the edge, bring it back a touch. After that, then play with timing. I'm 100% sure that removing the k/s is the right thing for me. It's been about 4 years now w/o one.

 

The only time I'd remove the knock sensor feedback is with a loud chassis dedicated race car. Things like the transmission shifting hard, solid bushings, and solid mounts can really cause the knock sensor to go crazy. But in a "quiet" ride it can be a life savor. Think about how consistent your times would be if you were just 2 degrees too much for 4 years and no knock sensor. My transmission 2-3rd shift sends false knock and that's with a low sensativity ESC , so I might have to disable it under heavy throttle until I rebuild the trans.

 

I think that removing the factors that lead to the detonation in the first place is better then just removing the ecms control over it.

If your absoltutly hellbent on getting rid of the knock sensor, then dont remove it. Instead, remove the ecms control over the timing and still keep the sensor so you can have its output in your datalogs. Better then nothing.

JP, damn, trans makes that much noise when it attempts to shift? Sounds like you better make that rebuild sooner then later

 

cough, cough....false knocks. Sometimes, they are impossible to remove...change the engine's harmonic vibrations and you'll see what I mean. Besides, people got along just fine before the K/S was made. It's harder to blow up an engine than most people think. My 406 has been at 39* total timing WOT from 3600rpm(til redline), it's ran 14.0:1 a/f ratio WOT several times in 97* ambient temps(125* track/AIT temps) and a host of other crazy stuff people would never try. No knock sensor, no O2 sensor.....

My "angle" is that one should learn to sense detonation. It's just a part of being a good tuner. Never "rely" on a sensor..

 

Well, I dont think that one should totally rely on a sensor, but its input can be useful as its another piece of information that you have to work with.

 

I totally agree, when talking about a "mild" engine. The stupid k/s just goes crazy on perfomance engines. I have tuned my friend's cars and they all had false knocks!

 

Theres not much debate if your getting lots of false knock from the setup your using. Disable it already

 

11s,
IMO, Your idea of adding 25% fuel will not work right with any particular dry "shot" of nitrous. Nitrous roughly comes out at a particular flow rate. To get the fuel right, you have to add a constant flowrate of fuel. Fuel flow rate normally injected by NA injections increases with RPM (say for example your PW is 8 ms at all WOT, at 3000 RPM the flow is half of 6000 RPM). To do it right, your fuel added in % PW would have to be inversely proportional to RPM (50% multiplier at 3000RPM, 25% multiplier at 6000 RPM). This of course neglects air displaced by nitrous. If it is MAF, then no problem with the displaced air, because the MAF only measures the air. MAP will take some guess work and tuning.

You can calculate how much you need based on torque and A/F and BSFC and injector flowrate. Basically if your torque at one RPM goes up 50%, you can say you need at LEAST 50% more fuel PW at that RPM. Divide the guesstimate nitrous torque curve by the NA engine torque curve to get at least a good guess for the multiplier, and then go conservatively more.

 

The MAF is a dutycycle based system, so an easier way may be to add additional duty cycle in before the calculated duty cycle is multiplied in with the DRPs to get the pulsewidth.



EDIT:OOPS! Forgot your not using MAF anymore. Nevermind.

 

You are saying that adding 25% fuel to the PE vs coolant temp table (OR 25% directly to the entire wot section of the VE table) will not actually add 25% fuel across the rpm range? I'm planning to log pulse width and make sure it increases 25%. EDIT: also log wot a/f ratio to confirm it's pig rich at all rpms(before testing the nitrous). Any suggestions are welcome.

It's actually killing me because I would much rather make a nitrous routine that'll add fuel+ subtract timing when X pin is grounded. Seems like ASM has a steep learning curve.

 

No, 25% fuel will add 25% to the PW. Which will be 25% more flow of fuel at each RPM. So a fuel flowrate of 10 g/s will be 12.5 g/s, and a flow rate of 20.0 g/s will be 25.0 g/s. Nitrous will come out at (for example 100 g/s) a constant flow rate. You will need to add a constant fuel flow rate to compensate it. A constant flow rate of (for example 10 g/s) will require 100% addition at a base flow 10 g/s, and 50% addition at a base flow of 20 g/s. Get it?

 

The fuel flow per each cylinder intake cycle will be 25% more but the overall fuel flow rate will increase with rpms since the injectors will be open more often. The problem comes that if the nitrous is a constant flow, then the nitrous content of the intake charge of each cylinder declines with increasing engine speed as the engine will be taking in more air.

I still like my DC idea. Multiply the extra needed DC by the number of drps between injector firings x the drp to get the additional needed pulsewidth. Tack it on to the BPW and, in theory, your done.

 

I get it.

 

Dimented is right also, and the DC method will allow you to tune specifically for the nitrous, i.e. you can change the VE of the engine and not have to worry about it screwing up the nitrous portion of the tune. Of course if you change fuel injectors or fuel pressure, it's another story (maybe use the injector flowrate in the equation and calibrate a fuel flow value in g/s rather than DC).

Make a patch that'll do that and debug it, maybe even get a baseline calibration, and I'm sure it'll be used by many (but having enough injector is going to be a limit). Maybe wire and plumb in an additional injector and put that in the calculation. I'm pretty sure the injector drivers can handle 1 to 3 more injectors. Could also just use a separate driver board. These would have to be activated/deactivated by the nitrous solenoid.

 

My injectors are at 60-70% DC right now. Wonder if they'll max out with a 125 shot. I can just crank up my fuel pressure to drop DC to about 50% for headroom....

 

could you explain that in a little more detail?

 

Just as an example, say your nitrous system needs 50 extra punds per hour of fuel and your total fuel capacity for all eight injectors is 200 pounds per hour. If you divide the extra needed fuel by the total fuel flow of the injectors, you get your duty cycle.

50/200 = 25% DC. The injectors will have to be open for an additional 25% more time to meet the extra demand.

The drps between injector firings determine how long the injector has to fire. Normally in double fire mode, there are four dist. pulses per firing. The Distributer Reference Period in the ecm is the time between pulses. The total time the injectors have to fire is 4 x DRP. On the fourth pulse, the ecm will again fire the injectors. The injectors have to be open for an additional 25% of this time to deliver the additional needed fuel.

Say the DRP is 5 ms at 3000 rpm. There are 4 in between each firing so thats 20 ms. Multiply that by .25 DC and you get the extra pulsewidth, which 5 ms. Add the result of this calculation to the BPW and you get your extra fuel.

 

The general eqn. is 4 x DRP x DC = PW in double fire mode.

 

Holy crap guys I just wanted to kill the test, not disable the knock sensor completely. Will disabling code 43 do this? Or just prevent it from throwing a code if it fails the test?

Tony

 

True, and those people understood the importance of plug reading, and did it regularly.

You must have missed my posting with the link to abnormal combustion. And while you ***MIGHT*** get by with some detonation, if it cascades into preignition, all it takes is a couple combustion events to kill the dome of a piston/ rings/ etc..

Don't get too carried away with making sweeping statements based on a limited amount of information. When you get to making too many assumptions, it will eventually bite you.

 

duly noted..

 

Yes, disabling code 43 will prevent the test from running. The side affect being that PE SA and HiWay SA are also disabled.

That's the beauty of the patch. It sets the status bit the same as if the test was already successfully run.

RBob.

 

So the ecm tests for knock at WOT?