Hall sensor EFI madness

[lvporschepilot]

Hey everyone, need help from the collective as the AEM tech department is about as useful as a chocolate teapot regarding this issue.


Car and engine: 74 Carrera. Aem infinity twin plug 2.8 motor, 12 coil on plug, ITBs, the works. I am using Rasant's cam/crank signal unit which plugs into the distributor hole.

A couple months ago I dropped the engine to mess with the cams a bit, upon re-installation was when this issue arose.


Here's the issue: The crank or cam sensors give out i.e. lose signal, but only once the engine is hot. Everything is 100% a-ok when cold/warm, but once it's fully warmed up everything hall signal related becomes intermittent. I can get the engine restarted, but it's intermittent once the engine is hot.

here's what I've done so far:

-replaced both hall sensors. Honeywell LCZ used on crank, and ZF cam (both have heat ranges up to 150C). Same type sensors the Rasant unit came with.
-replaced the shielded wires for both cam and crank sensors, and grounded the shielding
-checked and set air gaps
-checked the pins and harness into the AEM Infinity brain itself to make certain they are tight
-checked power source for sensors and ECU overall


AEM has no idea what it could be. I sent the ECU and a datalog back to them, it all checked out fine sans the datalog obviously showing the crank/cam sensors coming in and out at fully warm. AEM said it has to be a wiring issue. Well, I've replaced the shielded wire twice now, soldered all connections, tried different brands of shielded wire just to make certain. It's not that. Rasant doesn't know what's going on either


anyone with ideas?

[stownsen914]

Do you have access to an oscilloscope to check the wave pattern of the signal?

[Mixed76]

Check grounds? Does the body of the sensor ground via the motor?

Sent from my Nokia 7.1 using Tapatalk

[lvporschepilot]

Originally Posted by stownsen914

Do you have access to an oscilloscope to check the wave pattern of the signal?

Oscilloscope shows normal signal, until it goes out, at which time the scope depicts it as a huge spike.

[lvporschepilot]

Originally Posted by Mixed76

Check grounds? Does the body of the sensor ground via the motor?

I've tried three different grounds for the signal grounds. First, I tried grounding separate from the ECU, i.e. the engine block, chassis, then I have tried grounding via the 'sensor ground' pin on the EFI brain, which is what AEM suggested. No change. The shielding is grounded at the ECU end, and that is a chassis ground, as I do not want the shield and signal to both be grounded to the ECU.

Heat is the trigger for this issue. I am wondering if I just received a faulty new hall sensor that does not perform as advertised at heat.

[Mixed76]

But why would you have four bad sensors all at the same time? Possible but unlikely, right?

Sent from my Nokia 7.1 using Tapatalk

[rwest]

Put an ice pack on it after it goes to spike and see if it comes back to square wave?

[lvporschepilot]

Originally Posted by Mixed76

But why would you have four bad sensors all at the same time? Possible but unlikely, right?

Sent from my Nokia 7.1 using Tapatalk

Agree, hugely unlikely. When I popped the engine back in, I actually did find the cam sensor was bad as I was getting no signal when logging while cranking. I replaced both while I was at it 'as long as I'm in there' type thing. Engine fired right up once sensors were replaced.

[Mixed76]

If they're on the distributor the sensors shouldn't be getting all that hot anyway. Aren't they in the relatively ambient air outside the shroud?

Sent from my Nokia 7.1 using Tapatalk

[lvporschepilot]

Yeah I don't think they could get to more than 200F, and even that's probably high. My air temp sensor in the air filter shows max temps of 140F, but it's not bolted to the engine. The distributor is, after all, attached to the engine. Not supposed to be an issue though with that rating.

[trond]

could it be noise messing with ECU. I had issues with VR signal drops and eventually found it was caused by noise from distributor rotor phasing being off at some load settings. Noise could cause this I believe. Ignoring the temperature observation but since you changed most other things already...

[tirwin]

Here’s a left field idea riffing off of trond’s comment. Is there anything in your setup that changes once the engine warms up? Like a sensor that grounds when warm? Just wondering if it’s not heat but some other signal that coincides with the engine being warm. Maybe that signal is interfering when it appears?

[lvporschepilot]

Originally Posted by trond

could it be noise messing with ECU. I had issues with VR signal drops and eventually found it was caused by noise from distributor rotor phasing being off at some load settings. Noise could cause this I believe. Ignoring the temperature observation but since you changed most other things already...

I don't believe so. 99% of the noise is eliminated when using shielded wire. I have the crank and cam using their own independent shielded wires, both grounded to the chassis, while the signal is ground to the ECU signal ground pin.

[r lane]

Very aggravating. Not familiar with your system, but is one sensor reading directly off the crank and the other directly off the cam, or does it all happen in some gadget that replaced the distributer? And you have seen the signal drop out with your scope? Have heard of starters needing to be shielded even when not in use.

[lvporschepilot]

Originally Posted by tirwin

Here’s a left field idea riffing off of trond’s comment. Is there anything in your setup that changes once the engine warms up? Like a sensor that grounds when warm? Just wondering if it’s not heat but some other signal that coincides with the engine being warm. Maybe that signal is interfering when it appears?

I'm not running any WUR type system nor do I have any fans that are triggered by heat. The amount of fuel injected at cold start and while warming up is greater but that's all, no additional grounds to speak with.


I'm investigating something now, it looks like the bottom of the fan housing might be touching the trigger wheel a tiny bit as I do see a trace amount of metal powder/flaking on the crank sensor. Belt adjustment may be needed. Wondering if perhaps there is some movement occurring once the engine is hot that is causing trace metal to flake and affect the sensor?

[lvporschepilot]

Originally Posted by r lane

Very aggravating. Not familiar with your system, but is one sensor reading directly off the crank and the other directly off the cam, or does it all happen in some gadget that replaced the distributer? And you have seen the signal drop out with your scope? Have heard of starters needing to be shielded even when not in use.

The Rasant sensor pack basically looks like a distributor without the cap. It hangs the Hall sensor off the side, which reads the crank. Inside is the cam sensor wheel, cam sensor just sticks in from the top and reads.

[winders]

So I tried to use a hall sensor for my 60-2 flywheel on my 3.6L race engine. Rasant built the harness for my MoTec M130 ECU and shielded the wires for the cam and crankcase sensors. The hall sensor worked fine right up until I picked the car up from being tuned. Then it started generating "missing tooth" errors. I thought I had a failed sensor so I bought a new one but the problem persisted. The harness checkout so we were at a loss. I checked ground and added more grounds. I tried several different air gaps. The MoTec oscilloscope screen showed a clean signal with nice square waves.

Then someone told me that I should be using a reluctor sensor in that application. So, we replaced the hall sensor with a reluctor sensor and all has been running great ever since.

[lvporschepilot]

Originally Posted by winders

So I tried to use a hall sensor for my 60-2 flywheel on my 3.6L race engine. Rasant built the harness for my MoTec M130 ECU and shielded the wires for the cam and crankcase sensors. The hall sensor worked fine right up until I picked the car up from being tuned. Then it started generating "missing tooth" errors. I thought I had a failed sensor so I bought a new one but the problem persisted. The harness checkout so we were at a loss. I checked ground and added more grounds. I tried several different air gaps. The MoTec oscilloscope screen showed a clean signal with nice square waves.

Then someone told me that I should be using a reluctor sensor in that application. So, we replaced the hall sensor with a reluctor sensor and all has been running great ever since.

If you don't mind, what reluctor sensors are you using? And are you using Rasant's assembly as I posted above?

[winders]

I am using a MoTec M 4-APX-001.

I am using a flywheel with a 60-2 tooth pattern for the crank sensor setup.

[jpnovak]

What toothed wheel are you using? A 60-2 pattern at the diameter of a 911 pulley puts the tooth/valley width narrower than the sensor. In this case, the release latch hysteresis can sometimes cause a missing tooth or extra tooth depending on the crank angle with the internal electronics. In this case a 36-1 pattern will have a wider tooth "angle" and better detection at higher rpms.

A magnetic or "reluctor" sensor will usually have a smaller detection window and not have the same tooth angle count limitations.