Gents,

Can anyone can help me confirm some assumptions about the CAM sensor interfacing for a full sequential mode?

- Electrically-speaking, is the TEC3r looking for a pulse or just an edge?

- Does the required signal have to cross zero in order for TEC's input stage to see it (most reluctance sensors are read by zero-crossing detectors and when you feed those with a 0-5 or 0-12 square wave they miss pulses because zero is not crossed)

- If the TEC3r is looking for an edge, would that be a FALLING or a RISING edge?

- If it's looking for a pulse how can half-moon triggers ever work - they extend way past the required window.

The manual (Paragraph A.4.F) does say "as long as the sensor outputs a rising voltage to the TEC3R between 180o and 6o before TDC compression for the number one cylinder, it should work perfectly." – I would interpret this as RISING Edge.
Confusion is compounded by the User’s Manual figures. If I look at Figure A.4.7 (Ford 5.0L Half-Moon Trigger) and its shown engine rotation, the section labelled RISING edge would actually produce a FALLING edge at the output of the majority of common Hall-effect sensors. The output would stay LOW until the Raised Half transitioned to a Cut Half and then the sensor’s output would show a RISING edge with the “absence” of metal that would stay HIGH until the raised metal flew by again.

Of course, people who run a small tooth or a notch don’t really care about the sensor type or any of what I am asking. As long as they center that tooth or notch inside the window they’ll get one Rising and one Falling edge (maybe in reverse order) but a pulse will be seen by the TEC, nonetheless.

The reason for my post is that we are having a problem (intermittent CPS message in WinTEC4's gauge cluster) with a TEC3r-equiped car we are tuning. The 4-cylinder inline GM engine comes factory-equipped with a Hall CAM Position sensor which senses a raised section on the intake camshaft. Like majority of the Hall sensors, this one requires an external pull-up and when the raised metal section flies by, it pulls its output low – when the raised metal is absent the output goes back up high.

Attached PDF shows the camshaft profile orientation through two rotations of the crankshaft. Based on the PDF we get a RISING edge from the sensor at 90 degrees (crankshaft) rotation before compression TDC on Cyl 1.


We are doing some buffering of the CAM sensor output in order to add flexibility (polarity switching) but we are feeding the output of the buffers into TEC’s G10 input while trying the accompanying G4 pin to logic GND (which happens to also be the battery/vehicle GND).

Attached schematic shows the Sensor-TEC interface.

Still we get an Intermittent CPS fault. Any ideas what could be wrong?

We’ll have some scope captures ready for early next week which I’ll try to post.

Thanks,
Rick

Sorry about the massive image

Here is a scope capture taken between TEC G10 (HI) and G4 (LO) with the engine cranking (injectors disconnected).

WinTEC4 dashboard displayed a reading od 230 RPM
Looking at the scope capture Camshaft rotation frequency is 1.923Hz * 60 * 2 = 230.76RPM
The two readings match.

Looking at the waveform we see it's a nice and clean 0-5V square wave, so why does the TEC report an intermittent CAM Sensor fault? Any suggestions?

Rick

There are a few of use with this problem.Looks like you have dug into more then most.
Keep us informed on what you find. Im sure Jon is looking into this also.

I'm getting the same thing (only at startup not while running) now that I've updated to WinTec4. I wasn't getting it with the WinTec3 software.

I'm running a turbocharged GM 2.4 Quad4 in a '96 Sunfire.

Gary,

This particular project car we are working on, is a 2002 Sunfire GT with a Turbocharged 97 LD9 engine.
The engine cycle diagram above is for the LD9 (and probably the Quad4) engine.

I'll ask whether we can post some pictures in the Picture fourm.

I'll keep posting updates on the CAM sensor issue as we make progress.

Rick

There are a couple more things we can try

Without a confirmation from EMI, we can theorize that maybe the TEC wants a sync _|pulse|_, so we'll add a small one-shot circuit to create a pulse rather than an edge at 6-180deg (crank) or 3-90deg (cam) before Cyl 1 TDC Comp.

If this doesn't work we'll test the theory that maybe the TEC needs a zero-crossing (i.e. it wants the G10 signal to dip slightly below G4, the same as a mag sensor would produce) If the signal can't just swing from 0-5 or 0-Vign and needs just a small negative excursion this would be the answer.

I am tempted to think there may be something to the zero-crossing theory because the EMI Crank simulator outputs a Positive/Negative square wave.
On the other hand ALL TEC manuals say when using 5V Hall sensors, just feed their output into G10. This implies two things: TEC is ok with a 5V (HIGH) signal as opposed to V ignition (i.e. 12V) AND TEC doesn’t need a zero crossing signal – 5V Hall sensors don’t produce zero-crossing signals.

Also it’s never specified in the manuals what ground or return the CAM sensor interface should be referred to – In another thread Jon says to connect it to Sensor Ground, but generally one wouldn't want a switching signal (like the Crank or CAM) referred to the same ground that one's sensitive analog signals are referred to. I think G10 should be referred to G4 and G4 should be a high impedance input of a comparator circuit.

It would REALLY help if Jon or someone else from EMI can tell us what the input stage at G10 and G4 looks like (electrically). We don't need detailed schematics (that’s EMI proprietary info), just a general circuit description to allow us to adapt our CAM sensor scheme to the TEC. Also, regarding the WinTEC4 option MAG or HALL - what does that actually do? Does it activate a hardware switch that changes how G10 – G4 voltage is processed or is it purely a software feature?

Whatever the case, assuming we ever solve this issue - we'll post schematics, diagrams, parts lists and any other relevant info that can hopefully help the next guy.

Yes the LD9 is the 2.4 quad. Are you using the '95 2.3 quad intake cam or the stock 2.4? They both have the Cam Sync pad but the 2.3 has just a touch more lift (.006 I think) and about 2* more duration.

Are you or the owner on the J-body.org forums?

There are a couple of other Cam Sync threads wher Jon suggests using a 1k resistor between the 5v/12v power wire and the signal wire on the sensor plug to try to clear up some problems,I'm not sure what it's for exactly,maybe to slow the signal down or something. Maybe he'll jump in here tomorrow and clear up what's for???

There are a couple of pics of my car posted here http://electromotivetec.freeforums.org/lets-spice-it-up-t386.html

The 1K (or 4.75K in our case) resistor between the CAM sensor output and 5V or 12V is called a pull-up resistor and is absolutely required if you ever want to see the signal go HIGH. This particular CAM sensor's electronics have something called an open-collector output which means that they can only pull the signal LOW and let it float otherwise. The pull-up resistor takes the "floating" output HIGH - to either 12V or 5V depending on where it's tied to. Now the CAM sensor output can swing from HIGH to LOW and if you were to measure it you'd see a square wave.

Placing the resistor RIGHT at the sensor is probably suggested for wiring convenience but it wouldn't have much impact on cleaning the signal up.

There are Hall-effect CAM sensors which PUSH as well as PULL and don't require this resistor, but we aren't dealing with a sensor like that.

The intake cam is a stock cam (i.e. not from a 2.3L Quad)

Ok, first we're looking for an edge, not a pulse. You want the falling edge to land inside the correct window (as listed above). The cam input is simply a rail to rail opamp setup with an adjustable set point on the negtive input and about 10:1 amplification. In mag mode that set point is basically ground (same ground plane as sensor ground). In hall mode, that set point is about 2.5v. This is why we generally tell people to use a hall sensor for cam sync because with mag, if there is any noise on the line, there could very easily be false triggers. The resistor mentioned earlier for mag sensors essentially reduces input impedance whcih makes it such that any signal noise would have to be much stronger to get through. It has the side effect of reduce strength of the signal we want, but generally not enough to cause a problem.

I'm not sure why you're having trouble with this code. Maybe it's a false code. If the code is false, wait for the next release. I might have an imrpovement to the error checking that will solve that problem.

Doesn't the manual say that you are looking for a RISING edge?

Sorry, but now I'm confused and since I'm having Intermittent CAM Sensor error codes all the time, I'm trying to sort things out.

I'm sorry, I didn't have a manual immedately available and went on a what immediately made sensoe upon looking at the schematic. Disregard the last comment about the falling edge.