dr bob

Rob Edwards kindly loaned me his JDS Spanner to catalog all the things I need to address at intake refresh time. I got two errors from the EZK. One for a failed Hall sensor, the other for failed rear knock sensor. This discusion focuses on the Hall sensor.

My first step was to look to make sure the Hall sensor was plugged in. It is. Next was a quick tour to 928 Intl's web catalog to get a replacement. Ouch! Then went back to the car to see what might be wrong. Sensor and connector look fine, but the sensor is hard to see since the main engine wiring harness is wedged between the connector bracket and the coolant pipes on the inner fender wall. That can't be right... I had put the cable back right where it was prior when I did the TB/WP project a couple/few summers ago. The first TB job was done at a dealership when I bought the car. Looks like the place they left it isn't the right place. So disconnect the battery ground, remove and clean the 14-pin and the alternator connection to the jump-start post. Undo the little 6mm bolt that secures the cable clamp to the lift bracket on the front there. I love that bolt... Then rerouted the cable correctly, under the Hall sensor and then back around over the top again by the lift bracket. Reinstall the clamp and that bolt. Did I mention how much I love that bolt? Then plug in the 14-pin and the alternator connection at the jump post. Connect the battery and go for a drive. See another black 928 stuck in traffic with me. Back home, new scan shows no Hall sensor error code.

Seems that the rather sensitive Hall sensor looks at the notch in the back of the cam gear to help the EZK decide what cylinder is up when a knock is detected. The EZK then retards the timing for that cylinder in steps. No signal, always retarded I'm guessing. Anyway, having the alternator and other sensor cables wrapped around the sensor and it's cable apparently confused it enough to indicate failure. Move the big cable harness, and the Hall sensor is no longer 'confused'.

So if you get that error code, take a look to make sure that the noise from the engine harness can't interfere with the operation of the Hall sensor. Beyond that issue, the cable really needs to be clear of the bracket and the coolant pipes just so it isn't pinched/rubbed/damaged as the engine wiggles around on those fresh hydraulic motor mounts.

One error down, one to go.

IcemanG17

Dr Bob
Bill and I replaced the hall sender in Sharky a couple months back.....PITA..not fun...... however once it was replaced HP was BACK.....so it was worth it to get a properly running 928!!

Whats odd is I also had a hall and rear knock sensor fault...but the knock sensors looked perfect when we replaced them...however later Bill found a crack in the rear one that gave the fault!!

the flyin' scotsman

If you do need a hall sensor shop around..........there's a vast difference in price from 'our' vendors.

Also, the favourite bolt Bob refers to also grounds the passenger side ignition coil as well as securing the wiring harness clamp. Its almost easier to remove the PS engine hoist to gain easier access to the favourite bolt.

Jadz928

Me too...

Yea, you did. Me too.

the flyin' scotsman

Bob..........is this how it looks now? (pic curtosey of Dwaynes great write up):

Jadz928

Mine's the opposite, which is why is was so difficult to get the bolt back on.

the flyin' scotsman

Can you explain opposite? the harness is on the other side?

dr bob

Referring to Dwayne's picture, my '89 does not have a coil ground cable attached to the bracket. Also, the cable routes to the rear under the sensor and its cable connector bracket, behind the sensor, then forward over the lift bracket to the jump post. The loop on the clamp is to the rear of the bolt hole on mine. The cable does not pass across the hole in the lift bracket.


My cam covers are not nearly as pretty as Dwayne's. Yet. That's on my this-year project list.


Brian-- The big improvement noticed is a 10+% improvement in digital-dash fuel economy. I never push hard enough on the skinny pedal to test any of the possible extra performance available.

I do have the new knock sensors already. It's so much work trying to change one with the intake in place that it will most likely go in during intake refresh this winter/spring.

Hilton

The hall sensor is triggered by a metal disc with a hole in it that sits behind the cam gear.

As noted above shop around - I think my replacement sensor cost me $100-ish, although the length of the lead on the bosch OEM part from the vendor was shorter than the original Porsche one, with no slack in it.

My hall sensor showed a fault, and after my intake job and fixing up the broken connector it all behaved ok.

A couple of months later I plugged the car into a JDS Spanner, and got a hall sensor fault again then kicked myself and swore while I replaced the sensor with the cam cover in place

Certainly the most frustrating fiddly "simple" job I've done (way worse than fixing the floppy driver's door handle).

John Speake

The EZK is very fond of storing spurious "fault codes" on Hall sensors and knock sensors.

The diagnostic system works by each ECU storing fault codes, the diagnostic tester interogates each ECU in turn for its stored codes, and then displayed them. So no intelligence in the tester other than decoding the fault code to the displayed text display.

Porsche recognised the trigger happy nature of Hall/knock fault codes in the EZK, and every 50 (IIRC) igntion switch offs, the ECU fault memories are erased.

I suggest the following method....... do battery disconnect for a couple of seconds, reconnect, then go for test drive. Then read codes.

Bill Ball

Even the $250 Hall sensor purchased from 928Intl came with a shorter lead than original and a different plug so the retainer clip would not fit.

Dwayne's orientation for the harness bracket is reversed from what I have seen as well. It sure makes getting at that hated bolt a lot easier if the harness is able to take that orientation.

the flyin' scotsman

I just stuck my head into the frozen garage (-18C). The wiring harness clamp is indeed reverse of Dwaynes pic...........the coil ground is also on that side.

Hopefully you can make out the pic, hand was a little shaky in the frigid air:

WallyP

dr bob,

If the mileage went up 10%, you probably don't still have a knock sensor fault. If I remember correctly, EITHER a Hall sensor fault OR a knock sensor fault has the same result - a six-degree retard on the timing. Fixing one doesn't change the timing...

the flyin' scotsman

I went back and read '87 spec book; interesting:

dr bob

Wally, I have no way to tell if the knock sensor fault is chronic or intermittent. It may do fine part of the time.

Others: The Spanner gives you the option to clear the ECU codes so you can start over without cycling the key 50 times or disconnecting the battery. I did happen to disconnect the battery anyway as part of the harness re-routing so no faults were maintained. Drove the car maybe 25 miles doing errands, and that wa slong enough to register another knock sensor fault.

The next step on that is to pull the connector there between the FPR and the damper, clean it and put it back together. Maybe it's a dirty connection. I'm seldom that lucky, but it could happen.

The engine still needs an intake refresh just due to time. A local team here is working to get a powder paint color that matches the original Zermatt silver, so my intakes and cam covers might be among the first to get that treatment if I get it all apart in time. I have a project that will keep me busy at the end of this month, a perfect time to let the painters and platers do their magic on the underhood bits.

I so look forward to that intake mess. But it will bring all the sensors back to working condition again, so it's a Good Thing.