VR sensor suddenly not responsive... I'm going insane!

[badjuju]

This is weirding me out, and I got sick of updating my same post, since no one responds to it, so pardon this message if it needs to be somewhere else.

After getting my spark firing, I noticed that my RPM was a little erratic during cranking (at this point I still have not started the car).  I'm running... 1.0.73, I think.  I looked into getting a shielded wire for my trigger, but couldn't find anything except for 24ga.  I bought it, because a friend of mine told me it works fine for his Megasquirt setup.  After hooking the little bastard up and grounding it, I got no rpm at ALL during cranking, then once in a while it would chime in with a 177rpm or so, then back to zero.

I figured I had crossed a wire, or allowed the inner wires to be grounded to the shield or something, so I started testing.  I get like 260kohm between the shield and both wires.  I try another strand of the same wire, and I get the same thing  Should I be getting an actual resistance measurement between the shield and the wire?  I check for voltage from my pin27, which is where I'm hooked up to, and I get like .2 volts.  I slowly rotate my cam while my gf checks my voltage, and as the sensor runs over different teeth, she reads varying voltages from .4 to .5, to .2.  I cut the wire and check voltage on pin 27, it's about 2 volts.  I decide that the resistance between the shield and wire is allowing some sort of voodoo witchcraft inductive short, or maybe there's just a short in the wire.

I took the wire out of the shield, inserted a new conductor into each of two big sleeves of shielding, grounded one end of each shield to the head, and connected the trigger up in the same manner as I had it originally, before I noticed the erratic rpm signal.  Nothing doing.  I'm getting the exact same results no matter what I do.  This is great news for my wiring ability, but WHERE do I go from here?  I'm totally clueless.

Humbly yours,
Alec

[rob@vems.co.uk]

Does your P259 still work?
If you've been messing with the wires on the primary VR then you may have compramised the continuity between the grounds.

[rob@vems.co.uk]

Actually I've just been through your posts - you've had plenty of responses on the threads I can see.

[badjuju]

Frustrating thing is I had like a week of updating my own thread w no responses.  Please forgive my jumping the gun, but this project has been collecting parts for over three years, and I finally got the space to actually work on it.  I'm super excited that it's almost ready to start.  You know how it is.

Is P259 just a generic output?  Do I need to tell VT which channel my VR sensor is on or something?  Your (excellent, btw) user guide doesn't have pin 27 listed as a P259 output.

Also, come to think of it, when I changed the wiring for the VR, I just ran a shielded wire to VEMS, soldered it to pin 27 and the sensor ground, then further up the line, soldered the sensor ground line to the new wire I ran.  I'll try changing that back when I get home.

[mattias]

It should be stated that the P259 outputs should not be used without a guaranteed diode protected load, a relay without protection diode may kill it. What follows is that the dead P259 will drain all current from the voltage regulator on board, leaving nothing to drive the trigger signal A/D circuitry - effectively disabling any VR trigger.
 The signs of a dead P259 is that all it's outputs usually short to ground which will drive all things (relays) permanently. Normally people connect the fuel pump relay to it - when the chip is dead the pump will always run with the ignition on, regardless of the engine running or not (normally grounds only with rpm signal, and a few seconds at power on to prime the fuel system).

Use free I259 (ignition), S259 (stepper) or INJFET (injector) outputs to drive relays if you want to keep the P259 alive.

---

Measuring between ground and signal on the VR trigger, you should read a few hundred Ohms - this is the actual wire windings in the sensor.

Measure voltage between ground and signal, and swing something magnetic (not stainless, just plain iron works) in front of the sensor, you should read first a positive voltage and then a negative as the object leaves the sensor. This is a sign of correct wiring.

[badjuju]

Sooooo, P259 is flipping the fuel relay on and off, and it doesn't just turn on and stay on, does that tell me that P259 isn't messed up?  I'll switch it up to S259 to help protect it, but will an unfried P259 that is turning a relay on and off cause this problem?  Or will it only happen after the P259 dies?

[rob@vems.co.uk]

If theres anything happening with the P259 then its not dead.
So you need to look at the voltage coming out of the sensor - pull the plug off the VEMS and put your DVM probes across ground and VR in.
Also - disable any secondary trigger.

Rob

[dnb]

Another good thing to try is to play a trigger signal from a laptop soundcard into the ECU using the trigger simulator in Vemstune.  This will rule out any ECU problems.

[badjuju]

so, just so I'm clear...

tonight when I get home, I'm going to unplug P259, change my grounds around, test voltage, hope for more than .2 volts, see an rpm signal, change my input for my fuel pump relay, start the car for the first time in three years, and crank open a beer in celebration?

[Sprocket]

I figured I had crossed a wire, or allowed the inner wires to be grounded to the shield or something, so I started testing.  I get like 260kohm between the shield and both wires.  I try another strand of the same wire, and I get the same thing  Should I be getting an actual resistance measurement between the shield and the wire? 

There should be an infinate resistance between the shield and the open ended conductors. A 260kohn resistance between the shield and conductors will have comprimised the signal.

You have connected one end only of the shield to the common VEMS ground?

[badjuju]

I was told on MSEFI.com that the wire and the shield should have some amount of resistance between them?

I did not connect the shield to the common ground, I connected them to the head, right where the CPS bolts down, at cyl 1.  The sensor grounds are bolted to the head at cyl 6.

[Sprocket]

If you have two individual wires both with pvc insulation on them, you should get an infinate resistance. It is a resistance, but it is infinate, meaning that the actual value is so large that the value has no end. 260kohm is a fairly low resistance and would indicate that there is some sort of short between the conductors and the shield. I wouldnt be happy with any resistance lower than 20megohm in such a wire, and even then its still suspect.

I just measured a 10m length of industry grade 1mmsq Belden screened wire and that has infinate resistance between the conductors and the shield/ screen/ drain wire.

What is the wire you use and where did you get it?

[Sprocket]

I did not connect the shield to the common ground, I connected them to the head, right where the CPS bolts down, at cyl 1.  The sensor grounds are bolted to the head at cyl 6.

Are you saying that you connect the power grounds of the ECU to a different point to the sensor grounds?

I can confirm that is BAD

Please tell us that the four power grounds and the sensor ground are connected together at a common ground point, for example, the same single ring crimp terminal, and that all the sensor grounds from the IAT CLT TPS CPS or any others, all attach to the sensor ground wire, between the ECU and the common ground point, and not at the common ground point itself.

If the grounding is correct, does the head have a good ground path to the rest of the block?

[badjuju]

just finished wiring it to the letter, as you explained it just now, using http://195.159.109.134/vemsuk/forum/index.php/topic,52.0.html as an example.  Grounded to the chassis right under the box.

With my shield grounded, I get no signal at all.  with the shield ungrounded, I get sporadic rpm.

wtF is going on?

Edit: I disabled the fuel pump for now, since I don't know which I259, P259, etc, channel, goes to which pin on the ec36.  Suggestions?

[Sprocket]

Are you using the same shielded wire with the 260kohm resistamce between the shield and the conductor? If so, take all of that cable, and throw it away. I don't care much for what anyone said on MSEFI, the wire is useless. there should be INFINATE resistance between ALL the conductors and the shiled. End to end each conductor should read zero ohms as its such a small length.there is still the chance that the sensor wiring polarity is inverted, so don't rule that out yet. You need to resolve the shielded cable issue first.

You say 24 gauge, im assuming thats AWG and not SWG? which means the wire has a cross section of 0.2mmsq. A little small in my opinion anyway

Other things to consider is the air gap on the trigger itself, most VR sensors are between 0.2 and 1mm air gap.

Also, as Dave said earlier, prove the ECU is not the problem by using Vemstune to play a trigger signal that you feed into the VR sensor input on the ecu (without the real sensor connected of course) from the headphone socket on your PC. It really needs the ECU setting up on a bench to do this, but will prove whether the problem is the loom and sensor, or the ECU.