help with: fault Code 34 - Hall signal

[siverson]

Hi,

I'm making progress (I think!) figuring out the problem with my 964 engine in my 914.

I've added an OBD-II connector to the harness, and have used a Durametric reader to read and reset the codes (and poke around at other values).

Short summary of the problem:

- Car is very difficult to start and idles horribly (if at all).
- Will rev fine, and at 3k-4k rpm actually sounds ok (but probably not perfect).
- If I reset the codes, and try to start the car, right away the fault code "34" appears (hall sensor).
- If I let it run for 5-10 seconds, then the code "21" (Hot film MAF) also appears.

What do you think? Should I replace both or just start with the hall sensor? Is this a common problem? There is very little information about the 964 hall sensor error code. This doesn't seem to be a part that would fail very often... ?






-Steve

[ischmitz]

First off, a 964 doesn't have a hot film MAF, it uses the barn-door style AFM. But this could just be a software glitch (naming) on the Durametric.

The hall sensor error code gets thrown when the hall sensor signal from the distributor isn't detected. It is used to identify the left or right bank when knocking occurs. Without this signal the DME automatically ignores the knock sensors and retards the igntion by 6 degrees under load. If you have never driven the car with a working hall sensor you wouldn't know until a similar car takes you on the straight. You leave HP on the table and once you fix it you'll notice. The comon reason for the failure is a bad cable right at the connector that plugs into the distributor body.

However, this in itself doesn't explain the hard starting and bad idle since that 6 degree value isn't applied at idle. The Durametric should be able to read actual values. At idle the timing is around 0 degrees.

I don't know what the MAF error is but this could mean the DME isn't seeing the engine load signal (AFM) at all and uses an emergency programm where is calculated fueling only from the RPM. That would make the car run horribly. Again, while the engine is running the actual values should be able to display the AFM output. Have you tried that?

Also, make sure the idle contact closes at idle. Check if the durametric can display the status of the idle and WOT switch while the engine is not running.

Ingo

[siverson]

Thanks Ingo, that is very useful.

The long block was built by Andial, and the DME programmed by Protomotive when they were still here in California (about 5 years ago). One thing they did do was replace the AFM with a 996 MAF, and re-calibrated the the DME to accept its values. They also replaced the engine speed sensor with a 996 part number. This was supposedly more efficient and it fit much easier in the 914 engine bay - and it's worked great for about 5 years. Photo below.

996.606.124.00 - air flow sensor
996.606.105.00 - engine speed sensor

Since this is the first time I've read the fault codes, I'm not sure if the MAF error is always there (due to the component swap) or the fault codes work correctly, too. But, anyways, from what you describe my major problem is not the hall sensor so I'll focus on MAF for now.

And good idea on checking the actual running values, I haven't tried that yet... (will do so tonight)

-Steve

[ischmitz]

I see. Check the Durametric menus. If it is anything like the Porsche testers they should be grouped in several catergories. On the Porsche testers they are called

Fault memory (stored faults, always available)
Drive link (actuate injectors, only avaiable with engine not running)
Input signals (all sensor signals, only available with engine running)
Actual values (all output signals, only available with engine running)

Yours should be similar. Let us know,
Ingo

[siverson]

Hi - just to follow up on this. I could never get a reading from the MAF, but the intake air temperature sensor still worked in the MAF, so I was guessing that the wiring/plugs/etc was still ok.

I took a gambled an ordered a replacement 996 MAF and that did it. Starts and idles fine. Yay.

But... Now that I have the Durametric hooked up for the first time I still see the "Hall sensor" error. But no others anymore. I did take apart the top of the distributor, pulled the hall sensor and poked around a little bit, but didn't see anything unusual. All connections/wires/etc look good. So I've put it back together and the same thing - hall sensor error.

Having said that, I also watched the running ignition timing via the Durametric and at idle it hovers around 3.5 degrees, and ramps up to 35-40 (or so; this is from memory) at higher rpms. I've never watched the timing before either, so I'm not sure if this is correct or not. I don't see anything obvious that says it's in "-6 degrees, limp mode".

Any suggestions? Should I just try replacing the hall sensor in the distributor?

-Steve

[ischmitz]

Have you measured the harness to confirm that it isn't causing the issue? Mine did the same thing and I had to cut open the molding at the 2-pin plug to the hall sensor to fix the harness. No more errors since.

FWIW, my 964 stock DMEs (various different revision levels) keep the timing at around 0 degrees during idle (Hammer reading, never confirmed with a timing light). I wonder why yours is different. Maybe custom programming. A normal 964 will not report a MAF. This raises several questions.

- is the reported timing from Durametric correct?
- has the timing been changed in your DME mapping
- Does your DME even care about the hall sensor
- Does your DME retard timing when knocking is detected

Ingo

[siverson]

> Have you measured the harness to confirm that it isn't causing the issue? ...cut open the molding at the 2-pin plug to the hall sensor...

Do you mean the harness wires may be broken somewhere? Either way, I'll check that. But don't you mean 3-pin plug for the hall sensor?

> - is the reported timing from Durametric correct?

I'm not sure, and I'll try to check.

> - has the timing been changed in your DME mapping

It may have been, and I actually think it has. I know they made some changes to prevent the car from stalling with a lightweight flywheel. It also idles at ~1,000rpm, not the typical ~800 or so.

> - Does your DME even care about the hall sensor
> - Does your DME retard timing when knocking is detected

I'm not sure on either of these... ?

-Steve

[ischmitz]

Steve,
yes, I am referring to the DME harness. You need to measure from the 3-pin connector to the 55-pin connector at the DME. There are two wires. Maybe one of then is interrupted somewhere along its way. In my case one of the wires came loose inside the 3-pin connector. This wasn't visible from the outside.

Use a multimeter and measure resistance bewtween the each wire on the 3-pin connector to its respective pin on the 55-pin connector.

Cheers,
Ingo

[siverson]

> yes, I am referring to the DME harness.

Good idea, and I just checked that, but no luck. There are three wires, and they all appear to make it fine all the way to the 55-pin plug:

- "power for hall sensor" (read +4.66V, maybe a 5V component?)
- "hall sensor signal" (could not check running measurements, but verified continuity to pin)
- and ground (ok).

Hmmm...

-Steve

[ischmitz]

Hm, as I said since you keep referring to your engine as a 964 here are the work instructions from the 964 WSM:



and then here is the same section from the 993 WSM:




Can you see the dilema: The 964 sensor needs +12V to operate while the 993 sensor requires 5V. Since you measure +4.66V this identifies your DME as an early 993 (non-varioram) 55-pin DME. From the outside this DME looks identical to the 964 counterpart. And I remember yours is painted so it is hard to tell. On the inside these DME's are very different.

Now you keep referring to your engine as a 964. If that is correct you also have a 964 distributor and with that a hall sensor that requires +12V. And would explain why yours throws the error.

Ingo

[siverson]

I've got an update and another question... The 5V/12V dilemma really had me confused so I double checked the hall sensor wiring harness. It turns out that the wires inside the rubber boot at the hall sender on the distributor had in fact disintegrated (you're first suggestion to check). I had previously checked continuity between the DME pins and the hall sender wiring harness, but had always done so independently. Long story short: inside the hall sender plug the "signal" wire and 12V wire were shorted on each other. So when I checked continuity on the pins it all checked out, but the signal and power wire were touching. I think I was reading 4.6V on the signal wire due to this "short", when it should have been 0V with the engine off.

Once I cut the damaged hall sender plug off, everything looked normal again and the DME was indeed feeding it 12V. Yay, so I tried my best to repair this plug with some new wire - photos below.

So... here are the new question (2 steps forward, 1 step back): I got everything buttoned back up, and was anxious to clear the DME codes and see if the hall sensor fault code is gone now, but... now the Durametric stopped talking to the DME?!?! Argh! Nothing changed here and this *was* working reliably. Further, I've double checked all the pins (OBD connector to DME) and power/ground and everything still looks great. I'm at a loss.

The Durametric green light comes on (as usual), and it blinks red when trying to talk to the DME (normal), but then always fails to communicate. I've pulled the power for the DME overnight and retried, disconnected the hall sensor again, etc, and still no luck.

- Any ideas why the Durametric and DME stopped talking to each other?

Recall, now that I have replaced the MAF sensor the car seems to run fine (I haven't driven it yet (seats are out), but it starts right up and idles/revs very smoothly), but I want to verify the hall sensor fault code is gone.


Other random, less important questions that have come up:

- How can I determine which distributor should be connected to which coil/ignition module. Or does it matter? I've always just done this by swapping them and one way *seems* to be a tad smoother, but I'm not sure. Is there a more scientific method to know for sure?

- EE question: As you can see in the photos, I replaced the last 6 inches of the hall sensor connector with 3 wires. Previously, and the rest of this harness has the ground wrapping the 2 other wires. I'm guessing this is for shielding or ? Is my lack of shielding on the 6 inch replacement going to matter or ?

Thanks again for the tips - this is actually a fun project when I get things fixed, and not so fun when they *should* be working?!?!

-Steve

The OBD connector that has stopped working:



Start of digging into the hall sensor short:



Rubber boot off. Hey, this doesn't look so good! (BTW, the rubber boot was in *perfect* condition and very pliable/etc...)









I know, the world's worst idea for a soldering workbench:




Hopefully this works better:

[siverson]

Bump. Any ideas?

[ischmitz]

Steve, I am still wondering about what motor and what DME you are using. Can you clarify?

Theearly 993 DME has two channels for the ignition that are controlled independently. For idle they fire with a slight time delay (sequential) and under load the fire simultaneously. Maybe the head geometry favors one spark plug over the other to fire first. Not sure. Just ues what makes it run best.

The 964 has only one channel and fires both coils simultaneously and it doesn't matter what coil is connected where.

I don't know why the Durametric stopped working. It needs GND, +12V, K, and L. Have you checked with them what the LEDs mean. Could it be that is "thinks" you are using it on another car (without license). As far as I remember depending on what version you bought it is only good for one(?) or two (?) cars. Not sure how it "knows" because 964 DME and the early 993 DME have no serialized information such as VIN number. That came with the later 993 (OBD-II).

Ingo

[siverson]

> Steve, I am still wondering about what motor and what DME you are using. Can you clarify?

It's a 1991 964 3.6 liter engine (rebuilt by Andial).

DME part # is 911.618.124.01.

Car is back together now and seems to be running great, but I still can't get the Durametric to read it. I'm going to try another OBD-I reader to see if that's it. And I'm also going to connect my wide band O2 meter to verify AF ratio is ok.

-Steve

[Kroggers]

Steve, thank you for the help on the other thread - and I will save this one as well as I have a damaged hall sensor connector on my distributor and as such might find the issues you have had valuable.

But I would like to go off-topic for two seconds please. Where did you get the cover for the engine shroud at the top of the fan with the distributor vent kit (the one you have painted orange) - I would like one for my engine...

[siverson]

> Where did you get the cover for the engine shroud at the top of the fan with the distributor vent kit (the one you have painted orange)

Just Home Depot sheet metal cut with tin snips and power coated orange. Drill holes, rivited in place. It's worked well for about 13 years now...

-Steve

[Kroggers]

Thank you again Steve