454 TBI for '98 350 Vortec

**dave w** · 03-13-2015

Originally Posted by rsicard

Appears the next course of action would be to disconnect the IAC valve to see if that stops the cyclic behavior with lean/rich injector pulse width and o2 on this 350 Vortec with 454 throttle body. Trying to achieve the objective of getting a good idle RPM and lean injector pulse. Maybe retarding the ignition timing will bring down the idle RPM. Yet the ignition timing seen on TunerPro and the actual with a timing light do not agree. Need to establish some kind of relationship between the two. Will disconnect the ESC and the IAC valve, set the distributor timing with the timing light, see what the PCM is commanding via TunerPro, then reconnect ESC with the engine running and see what happens with the timing light.

Need comments from the experts concerning this cyclic behavior. Maybe it is just an incompatibility of the binary with this particular engine?

I've been very busy lately, and the near future looks very busy also! Attached is a TunerPro RT screen shot of the near idle BLM averages ... which shows it's running very rich!!!! I noticed the AE is going active during idle, not good! I also noticed 4 degrees of knock retard in the data log, not good!

I wish could help more.

dave w

**rsicard** · 03-13-2015

Originally Posted by dave w

I've been very busy lately, and the near future looks very busy also! Attached is a TunerPro RT screen shot of the near idle BLM averages ... which shows it's running very rich!!!! I noticed the AE is going active during idle, not good! I also noticed 4 degrees of knock retard in the data log, not good!

I wish could help more.

dave w

Dave: Thanks much for reviewing the recording. That was the SECOND recording that I made just getting used to doing it. Future recording will engine run time on them also for reference purposes. Will have to run it open loop and modify the near idle VE table to lean it out. Don't understand where the 4 degrees knock retard is coming from as the knock sensor is not connected. Need advise on what constants etc. to set for the engine runs. I just blip the throttle and saw the AE going on, then quickly off, for a very brief time with each blip of the throttle off idle. Will lean it out and retard the timing to slow down the idle RPM. Still getting more experience, for a relative newby, each time it is run. Need to disconnect the IAC to see if that cures the closed loop excursions to rich and increased BPW and then lean and decreased BPW and lean indication on the O2 sensor at near idle condition. Will monitor the Innovate Wide Band O2 sensor also. Need to wire the Innovate into the A/D input and set up the parameters to display Wide Band O2.

Once the idle is stabilized at an idle RPM without the engine quitting, and have the ignition advance set properly for idle, then I will lean out the near idle. Have been having a hard time getting the engine to NOT QUIT near idle on the first few runs. Getting better with the number of runs.

**rsicard** · 03-15-2015

Originally Posted by dave w

I've been very busy lately, and the near future looks very busy also! Attached is a TunerPro RT screen shot of the near idle BLM averages ... which shows it's running very rich!!!! I noticed the AE is going active during idle, not good! I also noticed 4 degrees of knock retard in the data log, not good!

I wish could help more.

dave w

Dave W: The knock sensor wire is not connected and floating electrically. Will ground it and therefore no signal input to knock sensor input to PCM. Also will disconnect the IAC and oxy sensor and it should stay open loop. Also switch to binary for export 350 Vortec with TBI and then tune from there both near and off idle. Get it close to 14.7 AFR in near idle. Then retard timing to get idle RPM where it should be. Then tune closed loop off-idle for rough tune for non-load on run-in stand.

**dave w** · 03-15-2015

I think grounding the knock sensor wire is a BAD plan.

dave w

**rsicard** · 03-16-2015

Originally Posted by dave w

I think grounding the knock sensor wire is a BAD plan.

dave w

The floating knock sensor wire acts like an electronic antenna picking up various electromagnetic noises in the vicinity of the engine. The knock sensor wire is an input to the PCM. Likely GM/Delco designed that input to be low impedance so as to not be as susceptible to radiated noise. If it is high impedance, then it is susceptible to electromagnetically radiated noise in the engine EMI environment. GM/Delco may have designed an electronic notch filter within the PCM for the knock sensor input yet the floating wire still acts as an antenna.

Another subject, read on another forum that the timing as measured by timing light from the balancer should be matched with what the PCM is calling for. Really wondering if this is so. The timing of the 350 Vortec, does not agree with what the PCM is commanding. Should the two be sync'd?

**fastacton** · 03-16-2015

The knock sensor wire is meant to have some resistance on it and I believe that's why dave w thinks that's a bad plan (I agree). If you can't connect it to the proper knock sensor, then I would either de-pin it at the ECM, add a resistor before the ground or just turn off knock retard in the bin.

**rsicard** · 03-16-2015

Originally Posted by fastacton

The knock sensor wire is meant to have some resistance on it and I believe that's why dave w thinks that's a bad plan (I agree). If you can't connect it to the proper knock sensor, then I would either de-pin it at the ECM, add a resistor before the ground or just turn off knock retard in the bin.

fastacton: Thanks much for your comments. What is being suggested is poor circuit design in that a resistor needs to be added externally to the knock sensor input to limit current? It is quite likely that a resistor is in series within the PCM at the knock sensor input. GM can hardly afford poor circuit design damaging a PCM with a possible external shorting of this wire.

I will check to see if there is a circuit bias supply voltage at the knock sensor input wire. Then check to see if there is current flowing to the knock sensor itself to see what its value is. If there is, suspect something in the small mil-ampere value.

**fastacton** · 03-16-2015

Originally Posted by rsicard

fastacton: Thanks much for your comments. What is being suggested is poor circuit design in that a resistor needs to be added externally to the knock sensor input to limit current? It is quite likely that a resistor is in series within the PCM at the knock sensor input. GM can hardly afford poor circuit design damaging a PCM with a possible external shorting of this wire.

I will check to see if there is a circuit bias supply voltage at the knock sensor input wire. Then check to see if there is current flowing to the knock sensor itself to see what its value is. If there is, suspect something in the small mil-ampere value.

You may be correct, but I would error on the side of caution and I don't trust GM to have completely protected the ECM when it comes to things like this. I don't see the need to ground the wire when there are other options available(I probably wouldn't do the resistor since it's more work that the other options, but it is an option).

**rsicard** · 03-16-2015

Originally Posted by fastacton

You may be correct, but I would error on the side of caution and I don't trust GM to have completely protected the ECM when it comes to things like this. I don't see the need to ground the wire when there are other options available(I probably wouldn't do the resistor since it's more work that the other options, but it is an option).

Thanks again for your comments. Will test and not guess. There are many parameters which are likely just trims for the particular functions. Will look for the one that disables knock sensing. Or just hook the darn thing up to the knock sensor. Would just as soon disable some parameters and then tune fuel and ignition and add disabled parameters back in one at a time to see what the effects are.

**dave w** · 03-16-2015

Subroutines, Subroutines, Subroutines, GM has many of them in the PCM!!! If you disable the Knock Sensor, some of the subroutines in the programming will be adversely / negatively effected! If I were to defeat the Knock Sensor, I would install a 1 watt / 4K resistor to end of the knock sensor wire and the other end of the 1 watt / 4K resistor to ground. Been there / done that / works fine / lasts a long time / no adverse effects!!!

dave w

**rsicard** · 03-16-2015

Originally Posted by dave w

Subroutines, Subroutines, Subroutines, GM has many of them in the PCM!!! If you disable the Knock Sensor, some of the subroutines in the programming will be adversely / negatively effected! If I were to defeat the Knock Sensor, I would install a 1 watt / 4K resistor to end of the knock sensor wire and the other end of the 1 watt / 4K resistor to ground. Been there / done that / works fine / lasts a long time / no adverse effects!!!

dave w

Dave W: Thanks for the advise. Sounds like GM early PCM software may have been written as spagetti code. Good coding is not done that way. Will try the 1 watt resistor.

**rsicard** · 03-17-2015

Originally Posted by dave w

Subroutines, Subroutines, Subroutines, GM has many of them in the PCM!!! If you disable the Knock Sensor, some of the subroutines in the programming will be adversely / negatively effected! If I were to defeat the Knock Sensor, I would install a 1 watt / 4K resistor to end of the knock sensor wire and the other end of the 1 watt / 4K resistor to ground. Been there / done that / works fine / lasts a long time / no adverse effects!!!

dave w

A 4 K ohm resistor calculates out to 1.2 milli-amperes. Very little power disapated. Measured the voltage on the knock sensor line. Little less than 5 vdc. Also measured AC voltage on the line. At RPM got 600 milli-volts. Have gear driven camshaft. Need to go back to chain drive to see if the AC voltage goes down. Suspect the Knock Sensor may be broad-banded. Yet again, it may be tuned to a acoustic resonant frequency. So much for now.