Mini cooper mpi 1300 (mems rover ecu) 1998 trigger setup

[Adamw]

21 hours ago, retro said:

EVO basemaps have the  trigger 2 edge set to ALL. (was only reason I was suspicious)

The edge setting is not related to the sensor, it is more related to giving the ecu the correct pattern.  The evo trigger system uses 2 different width teeth on the cam to identify both phase and cylinder pair, for the ECU to be able to measure the width of a tooth it needs to see both the rising and falling edge of the tooth.  

In your case however the cam sensor is only used to identify phase, trig 2 sync mode is "Cam pulse 1X", so the ecu just wants to see 1 cam edge per cycle. 

 

22 hours ago, retro said:

Our reference timing was 0 TDC, we would expect twice the pulses yes so one at TDC 0 and one 360 out.
So using a normal light, but changing the offset by 1 moved the crank angle 2 degrees? (so twice the pulses yes but doubling the offset???)

Using a normal light I wouldn't expect a change in 1 degree offset to move crank by 2 degrees. (then again I wasn't working the light to confirm this..)

You are correct, a non-advance timing light (or an advance timing light that is set to zero), should show a 1deg change on the pulley if you change the offset by 1deg.  So he is looking at something wrong I think - or maybe the marks arent scaled correctly or something?  I would do more like a 10deg offset change to confirm as it is often pretty difficult to see a 1deg change.   

You have possibly mentioned it already but this post is getting big - what does it have for coils/ignitors? 

[retro]

haha yeh this post is getting big, I was tempted to make my own post for getting this car working, but ah well if anyone else fits a Link to a Rover Mem MPi mini they have the joy of wading through all of this
However we have a custom 36-2 crank trigger and now a different cam trigger also so its kinda meh.....

Ah yes I forgot the EVO has a different cam trigger my bad.

Yeh regarding the timing light I think next time I'll work the light, I'll change by 10 and see if it changes as what he was describing doesn't sound correct to me.... Its the factory timing teeth on the lower cover, they are teeth marked in 4 degree increments but each valley would be 2 degrees so id say he just reading it wrong.

Coil is the factory MPi wasted spark unit.
You have possibly mentioned it already but this post is getting big - what does it have for coils/ignitors? 
NEC100710 - Mini Coil Pack | Mini Ignition | minisport.com Mini Sport

[Adamw]

Thats a dumb coil, does it have an ignitor/ign module connected between ECU and coil?  

[retro]

Hmmm I don't see an igniter anywhere in the wiring diagrams, so unless it was built into the rover mems ecu I assumed it was a smart coil with 4 pins...
Purple 'A' wire is switched 12+
Then the 2 outputs from the ecu to the coil. So technically only using 3 wires though..

Good point... crap how is rover triggering the dumb coil then... maybe some high coil drive direct from the ecu?

I'll measure the primary and secondary coil windings and see if there is continuity of any sort.

[retro]

Dang, just wired the car. Oops looks like we adding an igniter, now to find something suitable, I see link do a 3 channel. We just need any 2 channel that will work with this dumb coil. Will prob just chuck a generic Hitachi 2 or 3 channel on there.

Hopefully haven't caused any issues with the ecu (all 4 spark plugs were sparking) but only run it at 200rpm cranking speed so far.

Rover MEMS - MPi/SPi (chez-alice.fr)

 

[Adamw]

Usually no harm done, the Link ign outputs are actively current limited to about 2.2A.  Just the spark would be very weak, and it would be sparking on the wrong side of the dwell pulse which probably explains the timing drift.   

The 3 channel ignitor that Link stock is a 0227100203 or equivalent.  The most common 2 channel part number is 0227100200.  The 3 channel ones are typically cheaper and easier to find than the 2 channel for whatever reason, either will do the job.  Stick to a good ign brand like Bosch/Beru/Huco/Hitatchi, and avoid the generic parts stores ones. 

Or another often easier and cheaper option is to change to a VW or Subaru wasted spark coil with built-in drivers.

[retro]

Appreciate the info, the owner has ordered a Bosch unit from pans, so should be quality not a knock off. Looks to be the same as the link/haltech also.

Yes if it was my car, id run a smart coil.
I started with the MX-5 1600 basemap as that appeared to closely match the fuel and timing curves this A-series will want. I just worry the 480cc injectors won't be big enough to maintain under 20% duty cycle with this aftermarket cam (we shall see)

The car is a fully restored classic mini van, full panel and paint in a nice white colour, a resto mod theme, black bumpers black mags, so I could prob post some pics info on it on ya social media or here once done.

I'll be putting it on the Dynapack and maybe make a whopping 70-80hp at the hubs

 

[retro]

Got the car started and running on Sunday night.
The trigger offset was 360 out, but because wasted spark and quite lumpy cam it was a little tricky to spot at first as it was still happy running on 2 or 3 (maybe slightly on 4 haha) Checked exhaust port temps and plugs to find that.

Got that sorted and have the car revving and idling ok, my question is around injection timing.
I only have the wideband on the center exhaust ports 2-3. There is no bung on 1-4 for me to check currently.

Is the logic: if the injection time is too early I assume ports 2-3 get richer (as robbing the charge from 1-4.
If the injection time is too late the same thing happens?

Tricky to tune with only 1 wideband, but changing the timing from 410 to say 470 increases the rpm and slightly leans out mixture.

 

[Adamw]

You really need at least an inner and outer probe to tune these.  

 

7 hours ago, retro said:

Is the logic: if the injection time is too early I assume ports 2-3 get richer (as robbing the charge from 1-4.
If the injection time is too late the same thing happens?

No, the problem is cyl 2 and 3 are affected oppositely.  For example, assuming using start of injection timing - if we move the injection event earlier, cyl 2 loses some fuel to cyl 1, and cyl 4 loses the same amount to cyl 3.  So cyl 2 gets leaner by about the same amount that cyl 3 gets richer.   They effectively cancel each other out so the lambda wont see a change.  

[retro]

Yes start of injection and 410 as a starting point you recommended earlier.

Ok so if we have an o2 sensor on the middle port 2-3 (there is only 1 exhaust pipe out the middle so only option)
then another sensor on 1-4 (these are individual exhaust ports)

As you are saying: If cyl2 (inner) looses fuel and gets leaner (by giving it to cyl 1= outer richer)
Then cyl 4 (outer) looses fuel and gets leaner (by giving it to cyl3 =inner richer)

Yes the inners cancel out, BUT wouldn't the outers also cancel out? The net result would be the same even with 2 sensors?

So having 2 widebands, one on the inner (shared pipe) and then one on the outer (likely shared if we put the bung down the merge of 1-4) Then we will be none the wiser...

I'm not sure how to get this dialed in then..

Or is the idea JUST have a bung on 1 or 4, then try get either of the individual ports to match 2-3.

P.S when changing the injection timing from 370 all way to 470 I didn't really see a net lambda change, but RPM went up and smoother at different settings..

I thought inners would be richer when getting injection timing wrong, because it would sit on the valve and wait for next cycle as described here.
The siamese port injection problem (starchak.ca)

"

   from A to B the outside cylinder is sucking,

      from C to D the inside cylinder is sucking and

      between A and D they share the port.

the inside valve takes anything injected into the port from B to A which is 450° of the  720° port cycle or 63%,

      the outside gets what’s injected between D and B which is 180° or 25% and,

      they share between A and D which is 90° or 12%.

"

 

[Adamw]

22 hours ago, retro said:

because it would sit on the valve and wait for next cycle as described here.

He is using "batch fire" squirting the injector at the same crank angle every 360deg. 

We are doing sequential - the injector squirts first for cyl 2 when its intake valve is open (between C & A), then 180 deg later it squirts for cyl 1 when its intake valve is open (between D & B )

 

22 hours ago, retro said:

Or is the idea JUST have a bung on 1 or 4, then try get either of the individual ports to match 2-3.

This is all you need to do, get one of the outers to match the inner, or get the two outers to match each other (with no probe in the middle) would work also.  Ideally you would have all 3 probed but just 2 will give you enough idea of where it is happiest.

[retro]

Ahhhh yes that description is a "batch!!"

thanks for clearing that up. (I’m glad u are onto it haha)

yeh I was thinking technically u would have all 3 pipes with o2 sensors based off your description above. 

but yes we will use 1 of the outers, no point putting the bung on the merge of the 2 outers else it wouldn’t help me dial it in as it would just work out the same as you describe above.