STL94LT1

Atok, you can't use the OBDII knock sensors with your 95. The LT4 knock module will work with your 95 knock sensors.

It is very easy to get the knock sensors too tight. Think 14 lb/ft is the right torque. I would loosen and retorque the sensors. I had mine too tight and was experiencing spark retard. Sounds like this might be a part of your problem, and maybe a bad LT4 knock module.

Atok

STL94, yeah but I Googled the LT4 sensor part number and find it is used on multiple cars and trucks, older than that LT4. Vehicles like C/K trucks back to the mid-80s, Buick Regal GN uses it as well. It seems to be a very common sensor. Not sure why GM picked it for the LT4. On the other hand, the sensor part # in my LT1 is not very common at all. It appears to be specific to the LT1. I wonder what the difference is?

Knock Sensors
1996 LT1/LT4 - 10456287 ACDelco #213-325
1995 LT1 - 10456126 ACDelco #213-96

You mentioned your base timing is 4* more advanced than mine at idle. Are you using a custom tune or stock?

Thanks

STL94LT1

I guess you can call it a custom tune. I do my own tuning.

I added "Closed TPS" timing to clean up the combustion and get my car to idle a bit smoother with the cam (224/230 605/622 112). Tom "tjwong" told me this trick. Tom also mentioned that the LT4 knock sensors wouldn't work in the pre 96 cars.

69427

I've got company visiting this week, so my forum time is a bit limited, but I promise to re-type (from memory) my lengthy lost ESC description (attention: caffeine required prior to reading) this weekend. The above sensor information looks accurate to me. There were several changes and additions to the ESC circuitry over the years, mainly related to diagnostics, and these changes required different sensor detail changes. I hope my description later will answer some questions that have been floating around. Thanks for your patience.

Carpenter

Thanks alot for your forthcoming information '69 427

rickneworleansla

Wheelman

For that

Strick

This is a good thread. Perfect timing for me. My 92 LT1 has Crane self aligning roller rockers that have been in for about 20K. I stayed with 1.5s as coil binding in the 92 was a problem with 1.6s until you change out the springs. The other day I jumped in the Vette and started out when "service engine light" came on. Code 43, ESC. My questions are: one, do these knock sensors wear out? Are they items that need periodic replacing? If the voltage check described in the Helms indicates no problem with the sensors, could I have knock? I worry about gas these days. I, too, am standing by for 69427's article. Maybe we should put it in the tech tips section so it won't be lost. Thanks in advance for your time and effort in writting this much needed article.

Carpenter

A thought....a specific question for '69 427: I have Crane 1.6 (noisy?) on my LT1. Performance wise (talk and theory is cheap)......would I see better ACTUAL (DD) performance out of these or going back to the factory 1.5 non rollers? (Based on the PCM'S retarding timing because of 'knock', of course?)

tjwong

Atok, are you getting just knock counts or actual knock retard? The reason I ask is because the two parameters are different. Counts can happen without actually incurring KR.

The LT1 sensors are compatible with the LT4 Knock module, but the LT4 sensors are not compatible with the 94/95 PCM. The reason is because the earlier PCM is looking for a different shunt resistor than the later PCM. The 96 LT4 is 100k while the 94/95 LT sensor is 4k ohms.

Atok

Thanks for the clarification Tom. I didn't think about the fault detection for the sensors. Makes perfect sense.

I'm 100% sure it's not true knock. The reason is I can sit in my garage in neutral (no load), just take it up off idle to 1000rpm. At 1000rpm I can hear a rocker ticking, it's here that the PCM starts counting knock counts very quickly. If I got to 1100rpm the rocker is quiet again and the knock counts stop. I see 0* of retard when doing this test because my MAP is still quite low. The KR table in the PCM doesn't start pulling timing until MAP is above 40 kPa.

When I go for a drive, I see small amounts of KR only, using the LT1 KM. Nothing huge, unless I put my trusty LT4 KM in. Then in 6th gear pulling up a very slight grade it'll pull tons of timing making the car feel very lazy. It'll pull timing under all conditions unless I'm coasting (low MAP again).

69427

Correct again. Additional circuitry was added to the ESC module in the '96 model year application for additional diagnostic capability. The 4kOhm resistor in the LT1 (and the last few years of the L98 engine) sensors were used primarily to give the ecm something to measure to determine if the sensor was hooked up. In '96 the new circuitry actually measured and buffered the sensor signal, which was then read by the ecm A/D circuit. The ESC module still determined what was actual knock, but the ecm now had an active means to tell if the sensors were working.

69427

You are correct that there are several ESC calibrations for the LT1. My goal was one calibration, consistent with the requirements of the ecm hardware for that model year. It would have been easier for me, time wise, and the manufacturing guys would have loved having to only build one part. Unfortunately, the real world complicated things. Due to differences in the various engine configurations in the LT1 family (slight differences in the B,D,F and Y engines) the noise signatures were slightly different in several applications. I was forced on occasion to customize the calibrations for individual platforms to get an acceptable sensitivity to knock, while not being affected by false knock. Retard caused by false knock has serious consequenses with regard to emissions, fuel economy, and converter life, and the engine calibration guys rightfully have no sense of humor when encountering spark curves different than what they plugged into the spark tables. Regarding the actual spec differences between these modules, yes, I am familiar with them. I'd explain them, but it would take a serious amount of time to type it all out. I apologize for my laziness.

Strick

Thank you Mike for your words of wisdom on this topic. My 92 LT1 flashed a "Check engine soon" light the other day and a code 43, ESC, was pulled up by my scanner. The light hasn't come back on since. The only thing I have that isn't OEM is the Crane roller rockers. I'll have access to the tech scan tool Monday and hope to find out what is going on. Any tips?

69427

To be honest, I've never personally witnessed an actual failure of the internal piezo crystal (the important part of the sensor), although with abuse in handling or installation it sure seems possible. What usually gives out is the shunt resistor in the sensor. Heat cycling, and unfortunately, vibration, will on occasion break the crimp that makes the electrical connection for the resistor. When the resistor breaks, the system impedance changes, and the ecm will detect this, and set a code.
I'm not familiar with the details of the Helms diagnostic test, but I can offer a few bits of info when trying to debug the system. Later L98 engines (with the ESC unit in the memcal) should measure about 2.5 volts dc on the sensor line when the ecm is powered up. The 92-95 LT1 engines should read about 1.7 volts under the same situation, while the 96 engines should, IIRC, read close to 5 volts. These different voltage readings (with the sensor connected) are due to changes in the sensors and ecm configurations to enable diagnostics of the sensors and harness connections. One bit of warning. I have been told, and have witnessed supporting evidence, that the connections (both harness and sensor) are not terribly robust when it comes to repeated connecting and disconnecting. My advice is to minimize the number of times the sensors are unnecessarily disconnected.

69427

If you are fortunate enough to have no significant increase in engine noise due to the change to these rockers, then I would keep them. A little more valve lift should be a positive thing. If you're getting significant or repeated false knock retard with these, then a change is definitely in order. It would be a shame to remove the new rockers after going to the time and expense of installing them. It appears that several members have had good results with the LT4 module. It's a touch less sensitive, so you might experience more light or inaudible knock that doesn't get responded to, but it should reduce the occurance of false knock retard. If the new module doesn't help, then unpleasant or time consuming changes would have to be made to the engine or ESC circuit.

69427

Disclaimer: I haven't looked at the software for the LT1s in a bunch of years, so take what I say regarding the software with a block of salt.
The rockers shouldn't cause any check engine issues that I can think of at the moment. The rockers usually just cause false knock retard, which isn't an electrical issue relating to setting a code. I suppose if the software read an excess amount of retard requests it might suspect a problem with the ESC hardware, but I don't know if this ability is in the software. Some of the tuners with access to the s/w may be able to answer this question regarding the level of s/w sophistication. Given that the check engine light was only a temporary thing, I would bet against the problem being the rockers, as the noise should be there all the time. A marginal connection at the sensor, ecm, or even the memcal might trigger a CEL. I wish you good luck tracking this down. Intermittents are always difficult to debug.

69427

I'll try to give a brief explanation regarding the sensors. In a perfect world where nothing breaks or wears out, all Corvettes from the early 80's to the LT4 (I'm not familiar with the details of the ESC system on the LSx engines) would have the same sensor (IIRC: 5.2 kHz., 100kOhm shunt resistor). The sensor frequency of 5.2 kHz (versus 6, 7, or even 8 kHz) was chosen due to the comparatively large bore of the engines. The fender mount ESC units of the early to mid 80's used this sensor specification. The 100 kOhm resistor was installed in the sensor to provide a bleed off path for electrical charge while the sensor was being transported, handled, or installed. It played no part in the function of knock detection, so after installation it didn't matter if the resistor ever broke or became non-functional.
Reality: Occasionally the sensor connection or harness wire would open up or become intermittent. The system would become deaf, and the ecm had no way to recognize this, and knock would continue to occur. Not the desired result.
To give the ecm (and the driver) some indication that the sensor was unhooked, the sensor and ESC electronics were changed with the introduction of the memcal mounted ESC unit.
The sensor resistor was changed to 4 kOhm, and circuitry in the ecm was added to measure the sensor line impedance. As long as the ecm could "see" the 4 kOhm resistance in the sensor, it knew that the sensor was hooked up and had continuity in the line. Although this worked well, there were two small occurance problems that could arise. #1: As long as the resistor was intact, the electrical requirements were met, although the ecm had no way of directly knowing the functional performance of the sensor piezo crystal. And, #2: If the resistor broke, the system impedance changed, requiring a sensor replacement. Recall that with the earlier system that the loss of the shunt resistor had no effect on the performance of the system. So, there was a small tradeoff with the diagnostic addition.
The '96 ecm removed most of the drawbacks from the earlier systems. Circuitry in the ecm now monitored the actual sensor signal waveform for activity while the engine was running. As long as the sensor generated a signal (of sufficient amplitude versus rpm) the diagnostic requirements were satisfied. This eliminated the need for the sensor impedance check, so the shunt resistor was changed back to the protective-only value of 100kOhm.
This is a very brief description, but I hope it shows that there were actual engineering reasons for the change of sensor specs over the years.

Tom Piper

I've always wondered why the '96 went back to the knock sensor used in much earlier C4s, now I know.




Tom Piper

Tom Piper

Sounds like a good rule of thumb would be to use an ohmmeter to determine a good connection from the ECM to the knock sensor with the connectors at both ends of the harness disconnected.

Tom Piper