MSII EDIS problems
Posted: Mon May 17, 2010 9:15 am
Hi all,
I am trying to sort out a MSII/EDIS install on a Porsche 911SC. I am running ver 2.89 firmware on a V3 board and am using the Ford EDIS ignition module and coil pack connected to MS via the relay board. MS came from DIYAutoTune and is allegedly EDIS-ready.
Using the junkyard Ford EDIS module that I received when I purchased the box-of-stuff from someone who had given up on the project, the ignition appears to work overall, with 2 issues. The timing jumps around a little at idle (mostly on target but occasionally 20 deg or so off for what looks like 1 or 2 revs), and there is an apparent ignition cut out at around 4000 rpm. The problem used to be intermittent but now happens all the time.
The attached data log shows the problem. It seems that EDIS may be losing sync with the crank wheel b/c the dash tach (which is fed from the EDIS IDM line) jumps around when the problem happens - much more than the actual RPM change. On a longer data log that I took yesterday the RPM graph actually dropped to 0 at one point.
I figured the EDIS module might be bad (though I know these are pretty rugged) so I bought a replacement at Autozone. It does not exhibit the 4K rev problem BUT the timing is stuck at 10 degrees all the time. It is running in limp-home mode even though it is getting a SAW signal from MS (I confirmed this with an oscilloscope). I tried another module from AZ also, no change. I swapped in a new coil pack. No change.
I am using the junkyard wiring harness, which looks to be in good shape, shortened appropriately to fit. The VR sensor and PIP/SAW lines are shielded (in separate bundles) and the shields are connected to each other and to the appropriate pin on the EDIS module. I am running the IDM line to my dash tach, which works fine. I disconnected it just in case it was causing trouble, no change.
I am using resistor plugs (NGK BPR5EGP at stock 0.035 gap) and wires (originally for a Ford Crown Vic) and have the coil pack capacitor wired in per the diagram. I am using the cap that came with the junkyard module. It is wired in closest to the ignition module plug instead of the coil pack, which I have since read is not optimal, but the wire connecting the 2 is less than 12" long, so not sure if that is a big deal. I tried piggybacking a larger cap (470uF) in case I had major noise on the power line but that did not help.
My junkyard EDIS module is a Ford 91BB12K072BA (93 Aerostar). This is not on the list of approved modules in the Megamanual but it cross references to the same AZ part that most of the others do (F142). The replacement modules I tried were an F142 and an F159. Looks like all of the part numbers in the Megamanual cross ref to one of these.
The VR sensor is mounted 0.030" from the toothed wheel. Wheel is made of thin steel, not as robust as most that I see in the pics here. I checked the VR output on a scope and it looks good. I may try messing with the spacing to see if that helps somehow.
Sorry for the info overload, and thanks for any tips on where to look next. MSQ and datalog attached.
Scott
I am trying to sort out a MSII/EDIS install on a Porsche 911SC. I am running ver 2.89 firmware on a V3 board and am using the Ford EDIS ignition module and coil pack connected to MS via the relay board. MS came from DIYAutoTune and is allegedly EDIS-ready.
Using the junkyard Ford EDIS module that I received when I purchased the box-of-stuff from someone who had given up on the project, the ignition appears to work overall, with 2 issues. The timing jumps around a little at idle (mostly on target but occasionally 20 deg or so off for what looks like 1 or 2 revs), and there is an apparent ignition cut out at around 4000 rpm. The problem used to be intermittent but now happens all the time.
The attached data log shows the problem. It seems that EDIS may be losing sync with the crank wheel b/c the dash tach (which is fed from the EDIS IDM line) jumps around when the problem happens - much more than the actual RPM change. On a longer data log that I took yesterday the RPM graph actually dropped to 0 at one point.
I figured the EDIS module might be bad (though I know these are pretty rugged) so I bought a replacement at Autozone. It does not exhibit the 4K rev problem BUT the timing is stuck at 10 degrees all the time. It is running in limp-home mode even though it is getting a SAW signal from MS (I confirmed this with an oscilloscope). I tried another module from AZ also, no change. I swapped in a new coil pack. No change.
I am using the junkyard wiring harness, which looks to be in good shape, shortened appropriately to fit. The VR sensor and PIP/SAW lines are shielded (in separate bundles) and the shields are connected to each other and to the appropriate pin on the EDIS module. I am running the IDM line to my dash tach, which works fine. I disconnected it just in case it was causing trouble, no change.
I am using resistor plugs (NGK BPR5EGP at stock 0.035 gap) and wires (originally for a Ford Crown Vic) and have the coil pack capacitor wired in per the diagram. I am using the cap that came with the junkyard module. It is wired in closest to the ignition module plug instead of the coil pack, which I have since read is not optimal, but the wire connecting the 2 is less than 12" long, so not sure if that is a big deal. I tried piggybacking a larger cap (470uF) in case I had major noise on the power line but that did not help.
My junkyard EDIS module is a Ford 91BB12K072BA (93 Aerostar). This is not on the list of approved modules in the Megamanual but it cross references to the same AZ part that most of the others do (F142). The replacement modules I tried were an F142 and an F159. Looks like all of the part numbers in the Megamanual cross ref to one of these.
The VR sensor is mounted 0.030" from the toothed wheel. Wheel is made of thin steel, not as robust as most that I see in the pics here. I checked the VR output on a scope and it looks good. I may try messing with the spacing to see if that helps somehow.
Sorry for the info overload, and thanks for any tips on where to look next. MSQ and datalog attached.
Scott