Electromotive XDi-2 ignition system problem

I had the same issue with my electromotive box. I have the tecGT which incorporates the XDI box. Car would not start at times I was getting voltage to computer. It ended up being a bad connection at the fuse box. It was intermittent. Sometimes would catch sometimes it wouldn’t. Please check that. I have not had a problem since I cut wire and put new spade connectors on there. Starts every single time now. Drove me insane

 

Congrats!

Now it would be good it some one else with an electromotive would chime in because the signal you have looks inverted..the straight line in normally on the right and any snoopy line on the left and what you have normally means the the +/- pins sensor pins have gotten swapted. It will still work inverted but timing is less stable. Maybe try to get the missing teeth on the scope image and make sure it goes up not down after the gap.

 

That might be a "missing teeth" feature that the system uses for a "zero" reference (and, since your trace is only about 6 teeth long, sometimes it is in your sample, but most times not).

 

The trace should be going left to right. I've never seen it otherwise.....but I guess anything is possible.

That said, the missing tooth has exactly the leading edge it should with a long swoop in it negative from the last tooth going positive. The small swoop in the trailing edge means the trigger wheel teethvare just a little too wide for the sensor you're using and the steep clean neg to positive means they are a little too close together for the sensor. But it's not horrible and you've confirmed it's wired right so you're read to go drive

 

Awesome news.

Mark is correct, older data is on the left of the display usually. I’ve used some tektronix scopes from around that era and they were normal (older data left, newer data right).

 

No, the back is sadly the swoop side

No, i think they are right, the missing tooth for sure goe +then minus just as it should so im pretty sure its wired right.

its working so maybe don't mess with it

Running the wave form will be basically the same, it will be higher voltage so you may need to change the voltage scale and it will be faster so you will need to change the time scale, then it will look the same as cranking.[/QUOTE]

 

Here's one I made that looks a bit like yours but worse.
Image Unavailable, Please Login

It is for sure wired correctly but the right side falling edge is much worse than what you have.....this one goes flat so there is really no way to know where exactly the ecu is going to trigger. Your falling edge looks like this rising edge, there is a swoop but the line crosses the 0V axis at only 1 place and has a decent slope at the crossing point. This was me not thinking the design through before building it for the millionth time. I just layed and cut a 50% high/50%low tooth pattern and used the QV's OEM bosch vr ignition sensor the was designed to work with a 6mm OD pin as the trigger, not an 8 or 10mm wide square tooth...and when the trigger core pin is narrower than the tooth you get a flat in the signal because for a moment the mass under the trigger core is not changing and no voltage is generated. I had to swap it to a hall sensor.

Yours is no where near this bad! and should be fine. To get a perfect signal you need to eight use a sensor with a wider core pin or trigger wheel with narrower or V shaped teeth, then you would have a nice steep falling edge and you might have a spark that is +/- 0.1 deg instead of 0.2 or0.3 deg...it would be like that, a small improvement in accuracy that probably doesn't matter. Older ECUs and newer inexpensive ECUs can't do any better than about +/- 0.25 and would never know the difference but newer stuff designed to meet emissions rules is 0.1 deg worst case total error (so +/-0.05) or there about so getting the hardware right does matter. I bought a cheap hall sensor from a cheap ECU supplier and ended up throwing it out, if you look closely at the trace the time between falling edges varies (ignore the rising edge, it needed a resistor to clean that up) it was awful. but 3rd time was a charm for me......
Image Unavailable, Please Login

 

no, opening the gap will only reduce the voltage peaks not change the shape in any meaningful way.

I don't specifically knwo but I've never seen a VR setup that didn't set when it saw voltage above some small number, the trigger at 0V.

It also apparently has a minimum rpm somewhere between 120rpm and wherever its cranking now. I worked on an open souce ECU project that used the same chip family GM, haltech, motec all use and it did all the time stuff in a coprocessor called the eTPU, enhanced Time Processing Unit. Setting that up in the code was a little tricky because it was a balance between wanting maximum resolution which meant running at a high clock speed, but then you risk timing out at the missing teeth during cranking. the eTPU was 24bit, it replace the TPU that I think was 16bit and probably what is in your electromotive, or a version of it. a 24 bit will count to 16,777,215 but a 16 can only count to 65,535 so while we could support about 50rpm to about 25krpm (same chip and code many of the F1 teams used in the crazy rpm days) and have 0.1 degree max total error, a 16 bit can not so they will need to limit both the lower and upper rpm if they want decent accuracy and I suspect that is what you ran into cranking at 120rpm.

 

I think it will be fine with the ecu itself the biggest factor. The wave form isn't perfect but it has a decent slope at 0V, I would guess the sensor won't cause more than like 0.05deg error vs 0.01deg if you fix it... yes it could be 5x less error but it's still a very small number and probably only 10% the error the ecu will create. 0.5 deg is usually right at the line of what you can be pretty sure you're seeing on the dyno and what a 16bit ecu can do....you can just pull it out of the noise. So a 32bit ecu that is 0.1deg error will always beat a 16bit (which I'm pretty sure you have) for hp, but it's 1% maybe and you'll never see the difference between 0.51 and 0.55 deg.

Now that its working, I would just drive it and be happy

 

There is no replacement for dyno tuning to get stuff right. Do you have higher compression in the build?

 

mk e:

do you have any recommendations for an XDI replacement. I am experiencing an occasional loss of spark which has left me stranded. Now it seems to be fine so until the problem arises it's a bit hard to reliably fix. i am leery of taking it out since no spark is pretty fatal. This is on my '78 308 GTB. I've had the system for over 15 years with no problems until now (though I don't drive it very much). Any dvice would be greatly appreciated. Thanks.

 

We've got a replacement in the works. Looking at ease of switching out repin etc.

Since electromotive went away we've been trying to sort this.

I do know there's no similar dip switch control option. So any new ECU will be a programmable via PC/Android with more depth of control.

Just a heads up

 

makes me think its worth having a look at the wiring for something line that

As for replacing it....literally ANY ECU you can buy today will do at least as good a job. At the cheap end a rusefi uaefi for under $200 or any link or haltech or whatever probably starting around $1200-$1500. for the controller, you can probably keep the coils and sensor you have. You should add a MAP sensor if you don't have one. Or it sounds like Scott is putting a kit together that I'm sure will work well.

 

That's what we're trying to do. The xdi was great because it was just ignition, simple and clean. Everything now is EFI. No stand alone ignition.
Nick picked up another EFI company and I want to test and check them out. Ideally putting something together for those with XDI that need to switch up.

 

Thanks for the input mk e. I'm going through the entire XDI system now and will look into your potential replacements. I suspected wiring and when I started looking around something might have gotten re-positioned/reconnected so it's working fine now. I need to replicate the problem so I can be confident it is remedied or I might need to get towed again.If I can't replicate the problem I may have to go with the nuclear option (replacing it).

 

SDS EM-5: CPI is a standalone ignition system.

 

I'll check them out. Thanks for your help here, JCR.