Kugawestie's MX5 SDV Build Thread - Now An Upgrade Thread

[Westford]

Hmmmmm......in the video it looks like the fuel pump isn't running after initial priming. If you could disconnect it from the loom and wire the pump with temporary permanent live so it runs continuously this would prove it one way or the other (as a fault finding exercise only)

Westford

[CRAIGR]

Seems there is either no fuel or no spark when releasing the key from cranking position Gary.

Would suggest initially a temporary live to the fuel pump to make sure it continues to run when key is released from cranking position

[Dave Eastwood (Gadgetman) - Club Chairman]

Last time I saw an engine do that it had no crank/cam signal to the ecu, so poor rpm signal; the ecu would prime the fuel rail, the engine would spin and catch on the starting map, but the ecu never saw the rpm rise to a "running engine" figure, so never turned the fuel pump back on. The engine would burn the fuel in the pipes and die.

[KugaWestie]

Ok, so it sounds like a good test is to wire the fuel pump direct to the battery as a test measure. I will give this a go and see what happens.

Thanks for the suggestions  

If this works, does that mean the crank sensor is not working?

[Westford]

If this works, does that mean the crank sensor is not working?

I would say if it runs with the pump wired, then the crank sensor must be working, otherwise the ignition system wouldn't fire at all.

Westford

[Dave Eastwood (Gadgetman) - Club Chairman]

Sorry, just clarifying original post - that should have been "without a good rpm signal", not without a signal altogether.

(That'll teach me to multitask while posting   )

[KugaWestie]

Sorry, just clarifying original post - that should have been "without a good rpm signal", not without a signal altogether.

(That'll teach me to multitask while posting   )

No worries, we will solve this between us all!

So poor rpm signal would be caused by ?

[Dave Eastwood (Gadgetman) - Club Chairman]

In my case it was a poor connection in the crank sensor plug on the loom side. It made contact, but only just. Noise can be another culprit, but I wouldn't have thought that would apply in this instance.

I would check the basics as far as you can with an oem ecu, and eliminate as much as possible. Check that all sensor seem to be giving valid signals, is your rev counter fed from the ecu - does it move when the engine catches?

By pass the ecu controlled fuel pump relay and temporarily power up the fuel pump continuously. (being very careful, for obvious reasons).

Hard to see on the video, but when it catches is it doing so on all four cylinders?

[KugaWestie]

In my case it was a poor connection in the crank sensor plug on the loom side. It made contact, but only just. Noise can be another culprit, but I wouldn't have thought that would apply in this instance.

I would check the basics as far as you can with an oem ecu, and eliminate as much as possible. Check that all sensor seem to be giving valid signals, is your rev counter fed from the ecu - does it move when the engine catches?

By pass the ecu controlled fuel pump relay and temporarily power up the fuel pump continuously. (being very careful, for obvious reasons).

Hard to see on the video, but when it catches is it doing so on all four cylinders?

I am pretty sure all four are catching.

Rev counter not connected up yet as I am a section of loom missing    hopefully that will soon be sorted thanks to smokey

When I checked the injector connections using a bulb, I checked 2 and 4, and they were both the same.

My understanding is that 1 and 4 spark together for back up, and 2 and 3 do the same, this is why I checked 2 and 4

[hilux]

In my case it was a poor connection in the crank sensor plug on the loom side. It made contact, but only just. Noise can be another culprit, but I wouldn't have thought that would apply in this instance.

I would check the basics as far as you can with an oem ecu, and eliminate as much as possible. Check that all sensor seem to be giving valid signals, is your rev counter fed from the ecu - does it move when the engine catches?

By pass the ecu controlled fuel pump relay and temporarily power up the fuel pump continuously. (being very careful, for obvious reasons).

Hard to see on the video, but when it catches is it doing so on all four cylinders?

I am pretty sure all four are catching.

Rev counter not connected up yet as I am a section of loom missing    hopefully that will soon be sorted thanks to smokey

When I checked the injector connections using a bulb, I checked 2 and 4, and they were both the same.

My understanding is that 1 and 4 spark together for back up, and 2 and 3 do the same, this is why I checked 2 and 4

A live to the injectors doesnt mean they are being earthed so as to fire.

The crank sensor picks up the revs when its turning over and the ECU energises the injectors by earthing them in turn.

Check the relevant earth connection from the ECU

The injector plug should be 5 wires I would have thought - one live feed and one to each injector (they are earthed via the ECU)

I think you`ve doen really well so far but if you dont mind I note you are using a bulb for continuity/live testing so if you do not have a multi-meter I highly recommend you get one - they arent overly expensive.

You can test each ECU pin out to its end component for resistance and continuity in minutes.

I would still say its the relay or the crank sensor but make sure the ECU earth is good.

[TIG Brother]

The ECU has about 4 or 5 earths. 2 larger black wires which will earth out some where around the dash board area, which you will have to drill / make these go into a white connector block. There are then 2 or 3 ( cannot remeber exactly ) that run off into the the front loom area. One will be around the left hand side where the main battery earth strap goes onto the chassis, it will go into a multiple white earth block which should also earth to the bolt point. The other runs off down the drivers side and then goes on to earth onto a small stud underneath the throttle body at the front of the engine. Check to make sure these are all there.

Also again highly reccoment the multi meter, it will make life so much easier.

[Dave Eastwood (Gadgetman) - Club Chairman]

+2 on the multimeter, Maplins do them from very little cash.

Hadn't realised you had no instruments and a bit of loom missing. I don't know the Mazda set up, so it may be no help. But ordinarily, I wouldn't be trying to troubleshoot without either the missing bits in place, (even if just laid out loosely on the chassis), (or if I knew what all the missing bits were, just simulating the important ones with spare cable)

As Tig said, a whole raft of grounds for an ecu is common, one (or more) will be a power ground, then you have one (or more) "analogue" or sensor grounds.

The injectors like Hilux said will typically have power to them and an ecu switched ground. How many wires you have depends on how the MX5 fuel injection works. Full sequential injection would need a power feed to each injector plus an individualy switched ground to each injector. With that set up you also normally have a cam sensor too, so thath the ecu knows which cylinder is on a firing stroke.

Alternatively it may be a batch fired set up in which either all the injectors fire at the same time - commonly used on aftermarket ecu's on simple engine set ups like Westfield ones. This only has two wires, a common power and a common switched ground.

There is a sort of halfway house between the two types, where the injectors are fired in two groups of two. Obviously, this is a three wire set up; common power, plus two switched grounds.

[KugaWestie]

Right, back in the garage after a trip to the shops with the family    

I have directly wired the fuel pump from the battery and disconnected it from the loom. Cranked the engine with the fuel pump constantly running

No difference - so I think the fuel pump can be eliminated.

Next up - spark plug test

[KugaWestie]

Spark plug number on out, engine turned and getting a good spark onto the cam cover

[KugaWestie]

I think smokey mow has worked it out

The wiring loom that connects to the coil pack has 4 wires, and the coil pack itself only has three pins.

The wire that has no pin connected to it is black with a white stripe. Looking on the wiring diagrams for the 94 engine which this is, this black and white wire goes to spark plugs, rev counter and ecu

So looks like I need to find a 4 pin coil pack