T3 megasquirt II conversion

[Max Welton]

Progress. The sensor will live on the right side.

One thing to remember when mounting the sensor ... make sure it can be removed easily. It's going to have to be removed in order to replace the belt.

Tomorrow evening I will hook it up and make sure the MS picks up the tach signal. Then I need to box in that hold I opened up.









Max

[Max Welton]

Ok, before I haul off and hook this thing up ... can a VR sensor be damaged if connected backwards? Is there a backwards?

Max

[Devastator]

Yes, they can be hooked up backwards.

[Steve Arndt]

Ok, before I haul off and hook this thing up ... can a VR sensor be damaged if connected backwards? Is there a backwards?

Max

No you can't hurt it. It is a passive device.

[Max Welton]

Cool. It would be a drag if it was like the points-replacement modules.

Anyways, I have a good solid (cranking) tach signal from the VR. Now working on the shroud.

Max

[Piledriver]

Clarification: Hooking up a VR sensor backwards can cause headaches, as the waveform will be whack, and can cause an ECU to misplace or miss entirely the missing tooth. It will not hurt the sensor in any way.

You are probably OK, just throwing it out there.

[Max Welton]

So this is frustrating. With the initial positioning of the VR sensor the other day I had what appeared to be a solid ~300 rpm cranking tach signal. So I did some clean-up on the shroud (and repainted same) and mounted the TR on the fan (as shown by Devastator).

Next step was to set the tooth #1 Angle (under Trigger Wheel Settings). I did this by disabling fuel (unplugged the injectors) and had Mrs Max crank the starter while I watched the timing light. I had set the Number of Coils to "Single Coil" for this and connected pin36 to the stock coil. Although the signal seemed ratty, I put that down to the stock rotor/cap and installed the shroud and exhaust.

Then I started the car. Ran pretty nasty, with the tach reading reasonably but intermittantly dropping to zero. I still have the stock coil/distributor supplying ignition but the tach misbehaviour also screws up fueling.

So I figured I might be getting interference from the ignition, so I depowered the coil and just got a short log just cranking. Same thing. This zip has my CurrentTune plus the MLV and also an ignition log (first time I've used that).

http://www.maxwelton2k.net/squareback68 ... ntTune.zip

So I think I may have screwed up the VR sensor position while I was working on the shroud, having ruled out EMI from the ignition. I will now have to remove the exhaust and bellows again to see what I did. That's what I get for being impatient.

Unfortunately it is no easy thing now to adjust the sensor position.

Here are the pictures:
http://www.maxwelton2k.net/squareback68 ... index.html

Max

[Max Welton]

I have at least a partial explanation. The fan is somehow not on straight this time and can be seen to move radially when cranking. So the distance between the toothed wheel and the VR sensor changes through each revolution. And of course balance will be AFU.

So tomorrow it all comes back apart.

Max

[Piledriver]

Later T3 fans (maybe all of them?) have rubber in the drive hubs.
That rubber deteriorates... I have not tried fixing one, but perhaps others have?
If there is any play this will not end well.

[Max Welton]

Well, there doesn't seem to be any play. But the outer fan does have a little radial movement, but that is something that was already present. With some fiddling I was able to get the TW itself centered and moved the VR sensor in nice and close.

Didn't make much of a difference, though.

I updated the zip with a fresh log and CSV.
http://www.maxwelton2k.net/squareback68 ... ntTune.zip

Any ideas?

Max

[Max Welton]

OK, Matt Cramer thinks I'm seeing extra pulses due to emi in the engine room.

http://www.maxwelton2k.net/squareback68 ... othlog.JPG

So this weekend I plan to re-do the wiring from MS under the rear seat to the engine. I needed to add a wire for the second coil-driver anyway. One thing I will do this time is configure a DB37 where the bundle enters the engine room from the left rear fender so I can unplug it all at once whenever I need to drop the engine for any reason.

I'd like to do a tidy job of it this time. What are people using to wrap the wiring in the hot engine area?

Max

[jhoefer]

First, try adding a 10k ohm resistor in series with the sensor wire. Then try tweaking the VR conditioner.

http://www.diyautotune.com/tech_article ... _index.htm

And if you haven't already, (all) your sensor wires need EMI shielding. If your sensor wires don't have built-in shielding foil or braid, you'll have to add some (like mouser.com #517-1170-50/18). Though, it's often easier just to buy wire with included shielding (like mcmaster.com #8219K61) than to try to retrofit existing stuff. That shield also really should be wired to ground (engine block for example), but only connect one end of it to ground. If you connect both ends to ground, you can cause ground loops that can cause all sorts of weird problems. If you have a plug half-way between the sensor and MS, ground the engine side shielding to the block, ground the MS side of the shielding to the MS.

[Max Welton]

Excellent. I'll visit Radio-Shack in the morning.

Thanks!

Max

[Max Welton]

I re-routed the IGN wire for better separation with the HT leads and installed the 10k resister. The sync loss problem above 1k rpms went away. But I was getting drops (visible in the trigger logger) at idle.

So I went back and read http://www.msextra.com/doc/general/toot ... mples.html some more. There is an example of what the tooth logger looks like when the VR is wired backwards. Looked a lot like mine (when cranking).



So I reversed the VR wires and my tooth-log looked normal ... and the sync-loss at idle cleaned up.

It's electrical magic, I say. Magic!

So now I can set about pulling out the stock coil and installing my IGN4. After I run the second coil-driver wire.

This place is great!

Max

[Steve Arndt]

Happened to me too at first. I even used a scope. I didn't realize "falling edge" trigger was backwards from what I assumed by reading the wave form on the oscilloscope.