Issue with Setting Idle Speed

[NCGermerican]

Hey Folks - posting in the general 944 forum but the car is an 87 turbo. I replaced the ICV and am trying to reset the base idle. It's a straightforward process using Clark's Garage. However, I can't seem to turn off the ICV via the jumper in the diagnostic port.

I'm assuming that when the jumper is working and the ICV turns off, there would be "some" change in idle. However, when I insert the jumper when the car is at operating temperature nothing happens. Plus, when I look on the F9 ODB+ software, it's showing that the ICV is still active (shows % open).

Any thoughts?

[Tom]

You might check with @FTECH9 on that to make sure the WOT and ISV algorithms are sorted. See, e.g., this thread and post #36 in particular.

https://carpokes.com/viewtopic.php?t=1163&start=30

That may have been limited to very early alpha units, but worth ruling out...

[FTECH9]

Hey Folks - posting in the general 944 forum but the car is an 87 turbo. I replaced the ICV and am trying to reset the base idle. It's a straightforward process using Clark's Garage. However, I can't seem to turn off the ICV via the jumper in the diagnostic port.

I'm assuming that when the jumper is working and the ICV turns off, there would be "some" change in idle. However, when I insert the jumper when the car is at operating temperature nothing happens. Plus, when I look on the F9 ODB+ software, it's showing that the ICV is still active (shows % open).

Any thoughts?

When Test Socket A Terminals B & C are sorted together, this activates the full-load signal. When the DME program sees the idle and full load signals simultaneously, it knows to command the ICV to a specific opening and hold it there. At no time is the ICV disabled, it just more or less stays as the same position. On a 944 Turbo with a stock 1988 program, this should be around 85.5% open IIRC. However, some performance chips have changed this value for some reason.

As an alternative to the factory base idle adjustment procedure, with the OBD+ Sport DME you can follow the same procedure, but not bridge terminals B & C. In this method, you would adjust the idle air bypass screw in the throttle body until the ICV open % is as close as possible to 85.5%. Both methods work well, but by adjusting the idle based on the ICV %, you factor out the error due to how RPM is measured off the ignition signal and shown on the imprecise tachometer in the dash. The factory calibration method requires the use of a special tool to accurately measure RPM off a special TDC sensor next to the two flywheel sensors. This special flywheel sensor is described as "timing mark sender for top dead center" in the PET catalog. If you do use the shorting terminals B & C method, be sure to adjust RPM using the FocusOBD software, it is much more accurate than the gauge in your dash. When I get some time, I will add a special "RPM" measurement based on the TDC sensor that will measure RPM as accurately as the special tool the factory service manual calls for.

One bonus reason to calibrate the base Idel using the ICV open % is to verify the ICV is working correctly. While adjusting the air bypass screw, the engine idle should stay very close to 840 RPM (+/- 40 RPM). Only when you adjust the Idel bypass beyond the mechanical limits of the ICV (0% or 100%) should the idle speed change down or up respectively.


- Joe

[NCGermerican]

Hey Folks - posting in the general 944 forum but the car is an 87 turbo. I replaced the ICV and am trying to reset the base idle. It's a straightforward process using Clark's Garage. However, I can't seem to turn off the ICV via the jumper in the diagnostic port.

I'm assuming that when the jumper is working and the ICV turns off, there would be "some" change in idle. However, when I insert the jumper when the car is at operating temperature nothing happens. Plus, when I look on the F9 ODB+ software, it's showing that the ICV is still active (shows % open).

Any thoughts?

When Test Socket A Terminals B & C are sorted together, this activates the full-load signal. When the DME program sees the idle and full load signals simultaneously, it knows to command the ICV to a specific opening and hold it there. At no time is the ICV disabled, it just more or less stays as the same position. On a 944 Turbo with a stock 1988 program, this should be around 85.5% open IIRC. However, some performance chips have changed this value for some reason.

As an alternative to the factory base idle adjustment procedure, with the OBD+ Sport DME you can follow the same procedure, but not bridge terminals B & C. In this method, you would adjust the idle air bypass screw in the throttle body until the ICV open % is as close as possible to 85.5%. Both methods work well, but by adjusting the idle based on the ICV %, you factor out the error due to how RPM is measured off the ignition signal and shown on the imprecise tachometer in the dash. The factory calibration method requires the use of a special tool to accurately measure RPM off a special TDC sensor next to the two flywheel sensors. This special flywheel sensor is described as "timing mark sender for top dead center" in the PET catalog. If you do use the shorting terminals B & C method, be sure to adjust RPM using the FocusOBD software, it is much more accurate than the gauge in your dash. When I get some time, I will add a special "RPM" measurement based on the TDC sensor that will measure RPM as accurately as the special tool the factory service manual calls for.

One bonus reason to calibrate the base Idel using the ICV open % is to verify the ICV is working correctly. While adjusting the air bypass screw, the engine idle should stay very close to 840 RPM (+/- 40 RPM). Only when you adjust the Idel bypass beyond the mechanical limits of the ICV (0% or 100%) should the idle speed change down or up respectively.


- Joe

This is amazing Joe - thank you. I definitely use the RPM reading on the ODB+ as it verifies my dash tachometer is off by over 100rpm's. At about 850 RPM's in OBD+ it shows about 1K RPM's on my dash.

[danmartinic]

On a 944 Turbo with a stock 1988 program, this should be around 85.5% open IIRC. However, some performance chips have changed this value for some reason.

This is very new and interesting info thanks

I have always set the idle so that there is no change when using the jumper as I assume the ICV is supposed to keep idle at the right level and therefore jumping it out should make no change if you set the screw right.

In other words, for lack of an accurate measure, I would use the ICV as the correct idle indicator.

Since the ICV doesn't actually "jumper out" but stays active, would my silly method still suffice? Or am I doing it wrong?

[icb]

Interesting post. I can verify during recent testing I did on my tach that it was off appreciably at both ends. I ended up having to do a "best fit" cal on it to have semi-accurate idle speed and red line readings. I suspect the accuracy errors are mainly due to component drift at this point.

[FTECH9]

On a 944 Turbo with a stock 1988 program, this should be around 85.5% open IIRC. However, some performance chips have changed this value for some reason.

This is very new and interesting info thanks

I have always set the idle so that there is no change when using the jumper as I assume the ICV is supposed to keep idle at the right level and therefore jumping it out should make no change if you set the screw right.

In other words, for lack of an accurate measure, I would use the ICV as the correct idle indicator.

Since the ICV doesn't actually "jumper out" but stays active, would my silly method still suffice? Or am I doing it wrong?

I would think your method would be "sufficient", the base Idel adjustment can be significantly off and more or less work. However, the ICV is doing more than simply regulating the RPM at idle. The DME program is actively adjusting the ICV during different off-idle load conditions. It is not clear to me at this time how far you can be off and not interfere with the different functionally Porsche/Bosch had in mind for this system. As I dive deeper into the how the ML3 Motronic works, I hope to have a better answer as to how accurate this needs to be and why. But Porsche did spend significant amount of money on every 944 to have an otherwise unused flywheel sensor installed for this adjustment. My engineering gut tells me Porsche is very proud of what they came up with given the level of accuracy they are expecting for this calibration.


-Joe

[danmartinic]

Thanks Joe.. well, they do allow a +/- 40 rpm lol

Sport DME OBD shows rpm. Where from? Accurate?

[FTECH9]

Thanks Joe.. well, they do allow a +/- 40 rpm lol

Sport DME OBD shows rpm. Where from? Accurate?

The OBD+ Sport DME measures RPM via the ignition signal which also drives the tachometer signal. The ignition signal is generated by an 8bit 8051 microcontroller which gets its timing from the flywheel sensors. While measuring RPM directly from the flywheel sensors would be ideal, the ignition signal is reasonable. Eventually, I will implement engine RPM measurements directly off the flywheel sensors for the best accuracy possible.

-Joe

[danmartinic]

Most likely beats this old piece of **** I have just like the one below; when hooked to the coil, shows something under 700 rpm while tach is showing at least 900

To think I used take this instrument as gospel when adjusting the carbs on my old VWs

Tach _ Dwell meter.JPG (42.94 KiB) Viewed 1244 times