Help! ESC problems.

[Kevin Hall]

Hey everyone,

Now I know we did the right thing going with a carbed, non-computer controlled car for CMC.

 

Maybe someone can help us out on Thomas' street car.

 

It's a 1990 TBI 5.0 RS Camaro with a 5 speed. We put a rebuilt motor in it a couple of weeks ago, its a roller cam motor with the later model heads, flat top pistons and a cam that was described as Vortec specs. It sounded about the same as the CMC legal cam, which works great in the CMC car. We kept the OEM Manifold, but added headers. The car runs great except the timing retards fully intermittantly, it is worse when it gets above 200 degrees, and sets a check engine light with a code 43. It seems to happen under load at about 2,300 to 2,700 RPM.

 

We have replaced the Knock sensor and module, which is what the code 43 is supposed to mean, but the problem is still there. It's better with higher octane gas, and when t he temp is about 180. It seems hard to believe that the compression is so much higher, or a relatively mild cam would cause this problem.

 

We also found that when we removed the knock sensor, no coolant came out, and it is solid, rather than threaded into the water jacket.

 

Anyone out there have any ideas on what is wrong, or if we need a diffetent PROM, know where we can get one?

 

Thanks-Kevin

#82 CMC Camaro.

[bsim]

Are you getting any knocking or pinging?

 

Obviously the knock sensor is there to detect those things, and retard timing to stop the pings. BUT, if there's a lumpy cam, or unbalanced motor, it would act like you're describing. I'd get another knock module and hook it up to your setup, leaving the 'old' one IN the car (essentially hold it away from the motor vibrations). See if that helps.

[cmc38camaro]

Kevin,

 

I'll second Brad's idea, connect up a second knock sensor away from the engine and make sure to ground it. If the problem persists it's a problem in the wiring or the computer. I had this issue on my corvette and it turned out to be the knock sensor wire shorting to ground near the starter. If remoting the sensor causes the problem to go away it could still be a wiring issue "fixed" by moving the wire.

 

The problem could be a sensor installation problem. I can't say I have heard of knock sensor going into a solid hole, I thought they all went into the water jacket. Are you sure its not just plugged with rust? In any case the knock sensors are ultra sensitive to how tight they are installed and how much thread sealant is used. Tighter and less sealant make them more sensitive. The torque spec on them is something ridiculously low like 13 or 17 in. lbs. You also want to make sure there is pleanty of sealent on the threads.

 

After that you are into more serious issues.

 

Good Luck,

Mike

[Kevin Hall]

Brad and Mike,

Thanks for the help, we may have a little of all three.

 

We're concerned the the motor we bought (turns out the motor we built using a 1980 core just wouldn't work out) may have had either rust build up or some sort of filler blocking the hole. Going solidly into the block, not the water jacket might be part of the problem.

 

As far as pings, my 86 corvette, and my wife's 84 V6 Camaro have/had much worse pinging yet no retard. I can't really feel or sense any knocking.

 

We have the knock sensor we took out of the car when we replaced it, I'll try rigging up the old one remotely like you suggest and see if that has any impact.

 

We were monitoring the signal from the control module to the ECM with a volt meter, and it seems there is no change in the signal when the situation happens, which makes me worried it's an ECM or PROM problem.

 

I really like Holley Carbs and basic HEI ignitions. The CMC car runs great, but probably wouldn't pass smog.

 

Kevin

p.s. any one interested in a great little 1980 5.0 Chevy motor? It even has a trick Holley manifold that would work with a stock 2 barrel carb. Balanced, flat top pistons, CMC legal cam, and fresh valve job, built with love and care by Hall Racing? We even had a new flywheel that fits it, (but not in the bell housing of a 1990 3rd gen.)

[nick-91Z]

1) the knock sensor definately needs to go into a water jacket. it is calibrated to "hear" knock through the water.

 

2) your problem is probably the location of the knock sensor. the ECM "tests" the knock sensor equipment by jumping the timing up very high for a brief moment and listening for knock. (you won't notice this, it happens too quickly). If the ECM doesn't "hear" knock, then it presumes that the knock sensor circuit is dead for whatever reason(s) and goes into "safe" mode - which is to pull the timing out and the car runs like crap.

 

get the knock sensor into a water jacket and odds are it will be fine.

 

there is other ways (which involve modifying the memcal/prom in the computer), but i'm presuming that's pretty far off-topic (and it involves referencing another message board) so I won't do that unless someone lets me know that it's ok to do so. not as difficult as it sounds, though.

 

do you have the knock sensor in the stock hole in the block? i wonder if someone filled the water jackets in the lower end of your block for strength.

 

edit: there are a few other off the wall things that would cause the knock sensor test to fail, such as a lumpy cam and/or low compression, because it's harder to make a motor with such low cylinder pressures to produce a good enough knock signal that the sensor can pick it up (what size are the chambers on the cyl heads you used?), and finally make sure the timing isn't really retarded, you have to disconnect one wire (to disconnect the computer from the distributor) so that you can set the initial base timing. if you don't do this, then you set the base timing (probably 6* or so?) with the computer advance included, which is not what you want. disconnecting the wire turns off the computer advance so that you can set the real base timing.

[Kevin Hall]

Nick,

Thanks for the insight. Your comments make sense and track with what we are experiencing. We have the sensor in the correct location, it does appear the block has been filled. Guess our bad for not building the motor ourselves. The lack of coolant at the sensor is the only anomoly any one can find other than a bad ECM.

 

The boy is going to see if the other side goes into the water jacket, and if so, try to reroute it to the other side. We really don't want to have to yank the motor again, its much more time consuming than the race car.

 

Good luck on finding some tires, your project sounds really cool. We've got some used 245/16s off the the race car and my Corvette, but we're down here in SoCal.

Kevin