Low idle on rebuilt CX7 motor

[Ozynigma]

I am posting here again because no one, not even me, reads the CX7 forums

I have recently rebuilt my motor, mods are in my signature.

My idle is too low and close to stalling sometimes when in traffic. The engine also feels a little less responsive off boost but I could be imagining that. It just seems to take a bit of effort to dig out of hole of a slow idle to get moving.

The target idle speed in Versatuner is 850 rpm when warmed up for all settings, e.g. in gear, in neutral, AC on, AC off.

Here are the measured idle speeds.

In neutral AC off, 678 - 716rpm (697rpm +-19)
In neutral AC on and compressor running, 819 - 875rpm (847+-28), revs drop off when the compressor is not running.
In gear AC off, 602-642rpm (622rpm+-20)

So far I have checked and cleaned the MAF sensor and looked for any obvious vacuum/boost leaks. I have also vented my OCC to atmosphere and blocked the hose to the intake. Nothing made any difference.

I took one log that showed

AFR's are holding between 14.81 and 14.93.

STFT's varied from +0.8% to -0.8%

LTFT was 9.4%

So does anyone have any hints or tips on what I should be looking for that would keep idle rough and variable and well below the ECU's target?

Could MAF calibration be an issue and something to check? I am running on my tuners base map atm.

If its relevant I went to new 1 step colder plugs gapped to 0.7mm (28thou)

[CosmicArkie]

MAFCal would be the first thing I'd look at. If you're saying you are almost 10% lean at low rpm.

[Ozynigma]

Originally Posted by CosmicArkie MAFCal would be the first thing I'd look at. If you're saying you are almost 10% lean at low rpm.

Thanks, I am still learning what all this tuning stuff means. I swapped the turbo and had the battery out only 3 days ago so I am not sure if the LTFT would be stabilised yet.

I will do a MAF cal run asap, maybe late tonight when no one is watching.

[CosmicArkie]

Kinda depends upon how many cycles you've been through as far as stabilization. After you've done the tuning enough, you can learn to hold load, rpm, etc to get to that point quicker.

If you follow the MAFCal how to thread, you shouldn't get on anyone's nerves - you don't get over 35-40 mph.

[btstarcher]

What's your vacuum at idle?

[Ozynigma]

Originally Posted by btstarcher What's your vacuum at idle?

MAP varied in my log from 36.63 to 38.24kpa

By my calculations this is vacuum of 63kpa or 9.2psi.


I haven't done the MAF calibration log yet. Its a long story but its possible my base map is a bit lean and my current tuned maps are definitely dangerously lean (I checked yesterday 16AFR at idle, so I am not running them). This is because they were tuned around rich running problems with fuel leaks from an injector and the fuel pump that my car was having for the last year before the rebuild. So its not the tuners fault.

At this stage I just want to get the idle sorted and safely do another 500km of running in before getting it on a dyno and retuning properly.

A friend suggested last night that I look at whether the throttle body is sticking a bit at the end of its travel. So that is on the hit list too. If I give it a squeeze of throttle it smooths out around 900-1000rpm.

[btstarcher]

At idle you should be at around -10.5 psi. May have a vacuum leak.

[Ozynigma]

Originally Posted by btstarcher At idle you should be at around -10.5 psi. May have a vacuum leak.

Thanks.

I am going to have another look at my throttle body spacer and meth nozzle too. I had some water/meth/blue dye mix leaking out and I thought I had it fixed by tightening the nozzle up. I have had a lot of frustration with the throttle body spacer, I am close to tapping the cold pipe pre throttle body.

[Ozynigma]

I swapped throttle body and brake booster vacuum lines and resealed my throttle body spacer and WMI nozzle. No change to idle. So no obvious vacuum leaks around the TB.

I did a MAF log but I have Zero LTFT's above 30g/s flow so I will need to drive around some more for the ECU to learn fuel trims. I found the how to thread so this is going to be fun.

[KCi]

I have the opposite problem after rebuilt - high RPM idle regardless what i set in ATR.

It turns out to be bad vacuum. The shop had the OCC lines connected wrongly and the motor not seeing the right vacuum. Once this is sorted out, I bumped into low RPM and quick check on STFT showed it way off the limit. Good MAF calibration sorted this out.

[Ozynigma]

Originally Posted by KCi I have the opposite problem after rebuilt - high RPM idle regardless what i set in ATR. It turns out to be bad vacuum. The shop had the OCC lines connected wrongly and the motor not seeing the right vacuum. Once this is sorted out, I bumped into low RPM and quick check on STFT showed it way off the limit. Good MAF calibration sorted this out.

Thanks. That has me thinking. I have my Corksport OCC set up as instructed with VC and PCV running to a T and then through the OCC then back to the intake. I might try putting my VC back to the intake and my PCV back to the inlet manifold via the OCC and see if that helps.

Whats funny is that not only is the idle hunting around +-20rpm, but its not hunting around 850 +-20. I don't know why the ECU cannot correct up from the low revs with more throttle???

Also I have now checked my MAF calibration numbers and they are the same as the numbers published elsewhere here on MSF for the JBR True 3.5". So my baseline MAF calibration is correct for my mods.

[ALPINEST4RS]

Is the throttlebody clean? I had one hunting idle/stalling out when coming to a stop and I cleaned it. Took care of the issue.

[Ozynigma]

Originally Posted by ALPINEST4RS Is the throttlebody clean? I had one hunting idle/stalling out when coming to a stop and I cleaned it. Took care of the issue.

Yes. The one I took out was cleaned during the rebuild and I cleaned the one I swapped in yesterday. Brake cleaner and paper towels to remove pretty much all visible residue.

[Vansquish]

Not that this is particularly relevant to the present discussion, but your spark plugs should probably be regapped.

Mazda's own service manual changed the specs in 2010 for our motors. They recommend 0.024-0.027" at stock power levels.

[Ozynigma]

Originally Posted by Vansquish Not that this is particularly relevant to the present discussion, but your spark plugs should probably be regapped. Mazda's own service manual changed the specs in 2010 for our motors. They recommend 0.024-0.027" at stock power levels.

Thanks, I didn't know that.

Should I go to the upper end or lower end of that range if I am running well over stock power at 25psi boost?

I got my specs. from this thread, looks like it needs updating.

Spark Plug Database - Gen2

[torquemaniac]

@Ozynigma; I am also have some idling problem since my build

Since I dont have the balance shaft in place, the rough idling makes the car shake a bit, especialy when the heater is pushed to the max or the AC is on

The rough idling only happens in my case when the car is warm (cold idle is OK...)

I logged today my cold and warm idle, and noticed that the ignition timing when idling is wicked...the timing is lower than it should be (but the idling is not rough yet), then the ECU drops timing some more sporadicaly to try to get the idling rpm right on spot, but the reduction in timing actually increase the rpm...then the car start shaking real bad

I tried to find an idling timing table, but there is none...

Try to log the following when idling : load, actual lambda, desired lambda, rpm, ignition timing, maf voltage, relative throttle position, manifold pressure, engine coolant temp

You may also want to look at the VTCS system...it may be defective in some way

[Ozynigma]

Thanks VCTS is one of my next targets.

I rerouted VC to intake and no change. OCC is currently VTA.

[torquemaniac]

Originally Posted by Ozynigma Thanks VCTS is one of my next targets. I rerouted VC to intake and no change. OCC is currently VTA.

The Corksport OCC setup should have no effect on idling with the manifold capped

Did you get some DTC?

[Ozynigma]

No CEL and I just did a scan and no DTC.

VCTS seems to be functioning and when I moved it manually through its full range it made very little difference to the idle speed.

[Ozynigma]

So I might have a small exhaust leak at the donut join between my upper downpipe with cat and primary 02 sensor and my lower downpipe with no cat and secondary 02 sensor.

I could see some black staining on the edge of the flanges from exhaust soot when I put it back together.

Could a small exhaust leak cause the idle problems?

[Vansquish]

Originally Posted by Ozynigma Thanks, I didn't know that. Should I go to the upper end or lower end of that range if I am running well over stock power at 25psi boost? I got my specs. from this thread, looks like it needs updating. Spark Plug Database - Gen2

Yeah, that thread needs updating, there's no question about it.

I'd run in the middle to lower-end of that gap range, but if you're not experiencing blow-out, there may be no need to change anything for the moment. I would definitely try not to change too many variables at one time while you're troubleshooting your idle issues either.

[btstarcher]

So your OCC is VTA now? Because isn't that a vacuum leak?

Get rid of the OCC altogether, see if the problem persists.

[Ozynigma]

Originally Posted by btstarcher So your OCC is VTA now? Because isn't that a vacuum leak? Get rid of the OCC altogether, see if the problem persists.

Thats easier said than done. I will have to take the manifold off to reconnect the PCV to the manifold.

I now have the VC connected to the intake and the PCV VTA via the OCC. The PCV is a one way valve so I am not sucking any air into the system that way. In any case it would only be getting into the crankcase and not the intake system.

I am having trouble understanding why the ECU isn't applying more throttle and/or more or less fuel to bring the idle up to the targeted 850rpm. LTFT indicates its compensating for leanness but STFT are stable, so there is more room for correction if needed.

[torquemaniac]

Originally Posted by Ozynigma Thats easier said than done. I will have to take the manifold off to reconnect the PCV to the manifold. I now have the VC connected to the intake and the PCV VTA via the OCC. The PCV is a one way valve so I am not sucking any air into the system that way. In any case it would only be getting into the crankcase and not the intake system. I am having trouble understanding why the ECU isn't applying more throttle and/or more or less fuel to bring the idle up to the targeted 850rpm. LTFT indicates its compensating for leanness but STFT are stable, so there is more room for correction if needed.

Take an ilde log with ignition timing...

[Ozynigma]

Originally Posted by torquemaniac Take an ilde log with ignition timing...

I am getting the exhaust leak at the flange between the upper DP and lower DP fixed today. Its leaking between the two O2 sensors. Amongst other things it looks like I tightened the spring bolt connection up too tight on one of the five or six times I have had the turbo/DP in and out of the car.

I have taken the idle logs and they are on my laptop.

[Ozynigma]

No noticeable improvement with the exhaust leak fixed. I will do some more logs when I get back to Australia from OS.

[Ozynigma]

Originally Posted by Vansquish Not that this is particularly relevant to the present discussion, but your spark plugs should probably be regapped. Mazda's own service manual changed the specs in 2010 for our motors. They recommend 0.024-0.027" at stock power levels.

I just recapped the plugs to 24thou and it seems to have helped. They were all up around 32thou and dirty with carbon.

These plugs are one step colder. I will be running 24psi and WMI again shortly. I have been running spring pressure up to around 17psi boost while running the motor in.

ImageUploadedByTapatalk1450666577.817846.jpg

ImageUploadedByTapatalk1450666600.948855.jpg

ImageUploadedByTapatalk1450666619.068318.jpg

ImageUploadedByTapatalk1450666637.501860.jpg

[Ozynigma]

So I still have the problem. I have tried a few things and none have completely fixed the problem. I have;
Cleaned MAF sensor (and MAP sensor when motor was apart).
Regapped the plugs
Checked the MAF calibration of my tune
Zeroed the MAF calibration to 0.9V
Removed a 45 degree inlet extension before the MAF sensor
Added some extra throttle at low load and revs in the DBW Throttle Normal table

I have just done a new log. This log was done with the engine still warm but not at operating temperature. After idling in Park (neutral) with AC off for a few minutes I started moving between Park, Drive and Neutral.

I can see two things in the log.

1) The CPU flag for the auto transmission being in neutral did not change. This would account for my idle being worse in Drive but improved with AC on.

2) The tune targets 850rpm idle for all states of gear and AC when the engine is warmed up. However the log shows my desired RPM dropping and changing.

I did have the TCU unplugged when I had the engine out. I have also had the valve body of the transmission off for modification and back on again.

I think half my problem would be solved if the ECU knew the car was not in neutral.

[torquemaniac]

I looked at that log, and the actual RPM seems to match desired RPM pretty well (but desired RPM is not 850 all the way as commanded in your tune?)

And compared to your last log, ignition timing is a lot better (+10 versus -10)

[Ozynigma]

Originally Posted by torquemaniac (but desired RPM is not 850 all the way as commanded in your tune?)

Yes. Very strange. I have emailed Versatune for help.

[Ozynigma]

Originally Posted by Ozynigma Yes. Very strange. I have emailed Versatune for help.

@Steve@VersaTune

@Ugnius@VersaTune

[mazdakaps]

If you used the CS connection diagram you blocked the tube on the intake manifold? Maybe something isnt right there?

[Ozynigma]

Originally Posted by mazdakaps If you used the CS connection diagram you blocked the tube on the intake manifold? Maybe something isnt right there?

Yep I capped the inlet manifold port for the PCV. I have checked it by looking from under the car and it looks good.

[Ozynigma]

To close this issue out. Versatune have fixed it for me.

There was a problem with my software release version that did not give me editing access to two warm up idle tables.

As a result the ECU was ignoring the tunes minimum idle speed tables and trying to reduce idle speed to the stock values in the hidden warm up tables.

It does this by reducing timing, in my case down to -10 degrees at times. The reduced timing was necessary to counter the higher throttle table settings in the tune to match the desired 850rpm idle. Therefore the car was very sluggish getting off the mark and onto boost.