injectors turning off after a while

[jabran200]

I thought I had finally sorted my issues out, I can crank from cold runs smooth etc and was ready to road tune but after a while (not sure when as it is a bit random but happens when warmed up) it will turn off and I am pretty sure the inejctors are turning off.

After the car has stalled I crank but I only have spark and no fuel no mattery what I do, The only way I can manage to start it is by switching out to a powerfc and it will start normally. I can then start with the megasquirt but after a few minutes will stop again and I would have to switch out.

What could cause the issues? injector drivers? anything i should check on the megasquirt? Do the logs show anything odd?

log and .msq are attached. right at the end of the log shows the issue, afrs go off the scale as well which also indicates a fuel issue atleast.

[jabran200]

Just been sitting here going through the logs and I am not 100% but it looks like the cause could be the duty cycle dropping off as the rpms drop at the same rate/angle?

[Matt Cramer]

The pulse width is staying the same, so the drop in duty cycle is expected. Can you check if you're getting fuel when you encounter the no-start condition?

[jabran200]

Thought I was onto something there lol

So warmed up was running for a good 10+ minutes and then it died, I have a datalog but the laptop battery dies (I don't know if it will help ontop of the exisitng one)

I plugged a noid light on (previous spliced the injector plugs for aftermarket ones so have 4 extra plugs for injectors) just to see what that shows and during cranking after it has shutdown I have spark, fuel pumping through the lines but the noid light is not blinking. So Injectors aren't pulsing getting power I guess.

So immediatley after checking if fuel is coming I plug in my other ecu which fires it up straight away, discconect it after a few seconds and plug in my megasquirt and runs although for only a couple of minutes.

I though maybe the code is messed up so reflash the code on there etc and let it run and start it after it has been standing for 5-10 minutes and it started but turned off after some time.

I checked the transistors on the heatsink after it shutdown u5,q12 and q3 are slightly warm to the touch barely noticable? I assume that would be normal, they rest felt cooler

[Matt Cramer]

It could be the injector drivers are failing when they warm up - that would be my guess.

[jabran200]

So try and replace U7?

[SymTech Laboratories]

Is your ECU built on a v2.2 or v3.0 board? Your settings indicate you have high impedance fuel injectors. Can you confirm you're using either high impedance fuel injectors or low impedance injectors with inline resistors?

[jabran200]

high impedance with no resistors ms1 v3.0

I got to thinking and realised that I added a capaitor (a few weeks ago but never really kept it running for long) on the boot jumper to clean out soan hour and then called it a night, normally it would turn off after 15-20 minutes.

I will update tomorrow and see but it appears to be sorted.

[jabran200]

seems to be fine now after removing the boot capacitor and the 12v pullup. I think it was more the pullup than the boot jumper causing the issues but it ran for a lot longer today and I went for a little drive and really got the temps up so I would say it is sorted for now. t

Thanks for all your help guys

[jabran200]

Gutted, I thought this was fixed but I guess not. Only drove it a couple of times since "fixing" it and they were short journeys.

From your guys previous input and my "logical" thinking I see this, datalogs shows that megasquirt is indeed giving out the right pulse for the injectors so this then must mean something is wrong within the megasquirt like the Q1 and Q5 and I should start by replacing these.

Now that I think about it Q1 and Q5 are used to drive a full batch injectors as they are all wired in batch. The ecu has two pins for the injectors but they both join together in the loom. So if one died the other one would actually keep it working (i think) and the likelyhood of both of them dying is quite low I would think.

I think U4 is the actual FET driver (Q1 and Q5) so would it be the best bet to change that first?

[SymTech Laboratories]

That's a logical next step. If Q1 an Q5 have an intermittent issue, the problem is more likely to rear its head once the MOSFET's are warm.

[jabran200]

I am struggling to find a IXDI404PI-ND, are there any well known alternatives that may be easier to find?

[SymTech Laboratories]

I am struggling to find a IXDI404PI-ND, are there any well known alternatives that may be easier to find?

IXYS spun off their high voltage and power semiconductor branch to form Clare. They discontinued a lot of their parts in the process, but fortunately Clare makes pin compatible replacements. Take a look at Digi-Key part number CLA360-ND/Clare part number IXDI604PI.

[jabran200]

Thanks, I am not 100% if it was the chip now as I drove it yesterday for about 4-5 hours and moved the ecu to a cooler locations as before it was lying under the passenger floor mat (temporarily) and I guess either the db37 got moved going over speed bumps or the unit got warm due to the exaust runnign underneath the passenger side and I did notice it was fairly warm under the mat when I went to inspect it when it wouldn't start.

I was mapping yesterday and giving a good thrashing so it may be the heat? The unit gets noticably warm, is that normal? I assume they woudl as the heastsink touches the casing?

[SymTech Laboratories]

Thanks, I am not 100% if it was the chip now as I drove it yesterday for about 4-5 hours and moved the ecu to a cooler locations as before it was lying under the passenger floor mat (temporarily) and I guess either the db37 got moved going over speed bumps or the unit got warm due to the exaust runnign underneath the passenger side and I did notice it was fairly warm under the mat when I went to inspect it when it wouldn't start.

I was mapping yesterday and giving a good thrashing so it may be the heat? The unit gets noticably warm, is that normal? I assume they woudl as the heastsink touches the casing?

It's getting noticeably warm on the outside of the ECU near where the heatsink is? You might want to remove the top portion of the enclosure and feel for any hot components; none of the components should be too hot to the touch.