Been having some starting issues for a while on a cammed ls6 swapped into a track-only miata. I think it started around when I had the car rewired with a cutoff last year. The power wires were all changed and the stock miata relay/fuse box and ignition switch was ditched for custom stuff. Also got a cam right before that, but didn't have a problem between then and the wiring job.

Symptoms:
- 50% of the time starts fine.
- other 50% of the time it spins, catches, and then dies immediately. further cranking doesn't catch. power cycle the ignition and it usually starts.
- have a cutoff which cuts off everything included +12V constant to PCM. I always switch the cutoff off when parked
- Runs great once started. I've tracked it with this problem.
- When it doesn't start right away it seems like it catches after power cycle and as soon as I release starter ...
- Happens when cold or hot.

Tested/swapped these things:
- IAC (was making lots of clicking noises before)
- Power wiring. Everything was hanging off the starter bolt so I moved the ignition/PCM switched power to the cutoff. Now the dash power doesn't cut off due to voltage drop at starter.
- Injectors don't leak. Put new o-rings in cause I forgot to oil some of them.
- 60 psi fuel pressure at the rail
- no maf - speed density, tune hasn't changed at all.
- tested CKP/CMP with my oscilloscope when cranking and fuel pump off. See below. Looks ok. Could be bad sometimes?
- battery fine
- starter fine
- plugs fine
- plug wires fine

Ideas I have left:
- +12V constant to PCM still coming off starter bolt. Moving that to cutoff or maybe even direct to battery this weekend.
- +12V constant gets cutoff. Maybe it's losing some important idle/startup settings. Long term trims turned off in tune so it's not that.
- spark/coils?
- ECU went bad? I have a spare, but this seems really unlikely.
- Swap CKP and CMP sensor for the hell of it?

HPT log of 3 failed starts and then IGN power cycle and (rough because of extra fuel?) start. Right now I'm thinking electrical because of the symptoms and what changed before it started happening. Maybe the +12V constant is dropping too low and the PCM and spark is failing sometimes. Any ideas?

 

Where are you located ? Near me ? Did you build the harness or is it aftermarket? I would eliminate the car wiring portion. essentially connect the constant +12 v to battery and the switched ignition to a toggle switch then to battery. Sounds like intermittent switched +12 volts. Sounds like an ignition switch contact is burned.

 

I'm in CA. Engine harness is painless which I've had 0 problems with. It's always been on the car. The new wiring is a painless 50005 toggle switch+relays and new battery->starter/alternator/fan/fuel pump/PCM power wiring. Basically power distribution. There's no keyed ignition left. The first thing I'm going to do is pull that second set of wires off the starter. There's a big voltage drop at the starter and I wouldn't be surprised if that was the problem. It was causing the dash to flicker on startup when that was hooked up to that same spot. Overall the wiring job that was done is really clean except for this.

It's possible it's the toggle switches. I could try bypassing it or just hook the oscilloscope up and make sure it isn't losing contact.

 

Why are you connected to the starter terminal? as a power source ? Not a painless fan as we had several of their harnesses with pins in wrong holes. For a show car look with fuses, relays and switched ignition, I use current performance harnesses. Without switched ignition, PCM does not retain obd2 codes. Oops just noticed the 50005 is just a fuse panel.

 

I think it was done that way for wire length. Battery was in the trunk. I'll fix it, but it's bundled with some other wires so I have to get the car on the lift this weekend. Seems highly unlikely it's the painless stuff. The stuff added when the problem started is just relays and fuses with a switch panel. If I had known the power was going to be sourced from the starter I would have taken the car somewhere else or made the installer do it another way. It's easy enough to fix with some time, though.

 

I'd run a 10 gage wire from battery directly to alternator and a power distributin post. Then run the car off that. Use starter exclusively for starter functions. Just imo.

 

Ya, that's what I'm doing now. I'm using the cutoff post as a distribution block. It's got some beefy posts. Battery and cutoff were moved to the passenger footwell and are about ~1 foot away from each other. Starter is now ~2 feet away from cutoff. This last wire is more of a pain because it's wrapped up in some transmission wires and CKP wire.

I was trying to figure out what the PCM uses constant 12V for other than keeping memory. Maybe it uses that for powering all the output circuits and the switched 12V is just to trigger an internal relay.

 

Constant 12 v is only used to retain obd2 codes far as I have seen in the past 10 years of building harnesses, programming and doing plenty of swaps.

 

Haven't rewired it, but I read about some starting problems caused by the fuel pump. When I power off the pump it drops from 58 to 54 psi at the rail. After 2 minutes it's down to 40 psi. An hour later it's at 22 psi. Not sure if this is bad or not. It's an aftermarket dw301 pump I've had for a while. If the wiring changes don't work out I might steal the walbro out of my subaru and swap that in.

 

That is not a bad bleed off rate.

 

I moved the wire to the cutoff which was a PITA since it was wrapped up in the main harness. Started several times without issue. Going to try starting it tomorrow morning when it's colder and see what happens. If I still have a problem I'm going to look at the tune.

 

No luck. Didn't start up this morning on the first crank. Now I'm thinking it's tune related. It's always had some cold start issues when temps drop below 50. Need to give it gas to keep it alive until it warms up. Going to play around with it today and see if I can figure it out.

 

Thanks to aknovaman for helping me figure this out. It looks like cutting off the power to constant +12V to the PCM and having it lose the MAF DTCs with my SD tune is the problem. The PCM needs to relearn that every time I switch the cutoff back on which is done when I try starting it. Then when I cycle the ignition it knows the MAF has failed and runs in SD mode and starts up properly. I'd like to have the PCM relearn the DTCs before the first start, but it looks like there isn't a way to do that. So my options are:

- Hook up constant +12V direct to the battery. I wanted to avoid this since I'd prefer everything gets cutoff with the switch and the car gets parked for a while.
- Hook up the MAF. The MAF connector on my harness was cut off so I need to get a new one. Might just get the ls7 connector since I already have the ls7 card style MAF in there for the IAT.

 

Actually, I'm a little slow tonight. I already have the LS7 breakout connector that I can cut up and hook both the MAF and IAT wires to.

EDIT: I was hoping that having a non-floating MAF signal might trigger the DTC, but one of the conditions is RPM > 300. http://www.ls2.com/boggs/dtcs/DTC%20P0102.htm. Looks like the easiest thing to do is run a MAF.

I'd hook up the ECU directly to the battery, but it's a track car with a 3 lbs 10Ah LiFePO4. I measured 36 mA draw from the ECU a while back (maybe that was more than just the ECU?). While I was hoping Peukert's constant might mean a slow discharge would prolong the life of the battery, it only took about 1.5 weeks for the battery to drain pretty far down which Li* batteries don't particularly like. I'd rather just put everything on the cutoff. Another option is a Pb battery for at home and then switch to the LiFePO4 at track events.

 

Did some more searching and came up with two more options:

1) Set the lowest table entry in the MAF table to 12-12.5 g/sec so the car gets fueled before it has a chance to fail the MAF:

http://www.hptuners.com/forum/showth...force+maf+fail

2) Mess with the MAF fail conditions. Look like HPT doesn't have all of these, but EFIlive does. Only bad part is I'm running a custom HPT OS so EFILive probably won't be able to open and save that. Other option is to look at a stock tune, make this one change in EFILive, diff the two binaries, and hexedit my custom binary to include the changes assuming everything looks like it's in the same spot.

EDIT: Arg, looks like HPT encrypts/compresses the file because a small change shows a complete change in the diff...

LABELS
Parameters Value
{C2901} MAF High Frequency Fail 1 13500.000000 Hz
{C2903} MAF High Frequency Fail Limit 18.000000 Failures
{C2904} MAF Low Frequency Fail 1 1300.000000 Hz
{C2906} MAF Low Frequency Fail Limit 6.000000 Failures
{C2907} MAF Test Min Engine Speed 400.000000 RPM
{C2908} MAF Test Min Run Time 2.000000 Seconds
{C2909} MAF Test Min Batt Voltage 8.000000 Volts
{C2910} MAF Test Max TPS 95.000000 %

 

hook the 12 volt constant to the battery , and buy a battery tender , that way the ecu will remember the tune and your battery will stay charged , if they work with li batteries

 

There's no good tender for a Li* battery. They like to be nearly fully charged and left alone. I gave up on trying to hack the lower part of the MAF curve and hooked constant +12V to the battery post on the cutoff. There's now a Pb battery in the car with a tender on it. When I go to the track I can warm it up on the Pb and then swap in the Li battery and go through the MAF fail process one time in the morning. Annoying, but less annoying than not knowing what the problem is and have it fail at seemingly random times.

The extra parameters in EFI Live aren't really going to solve the problem only potentially make it fail into SD mode faster. Seems like there's no way to set the DTC without trying to start the car.

Most of the reason I don't want the MAF is throttle response.

 

how bout a small lead acid like a bike battery with a isolator that you can hook to the constant , and just keep that on the tender

 

That's a good idea. An isolator (diode) might work and someone mentioned a 9V battery. I can keep the ECU alive with that when the cutoff is switched off. Let me think how I can wire that up so nothing explodes and they don't kick me off the track. A Pb battery is going to be too heavy. The Li battery I run is 3.2 lbs - don't want to add anymore weight than necessary.