Idle Dip

[93_Clubman]

In case you've not seen this on the EGR:
http://www.mazda-speed.com/forum2/index ... 742.0.html

[doctor_bob]

That was the page we worked from when pulling it off last night actually

I think we've narrowed it down to the oxygen sensor actually based on some advice from the awesome fellas over at mx5 mania - they suggested disconnecting the o2 sensor to let the factory mixture settings take over and test it, sure enough it was like flipping a switch for rooting the throttle when connecting and disconnecting it. I'm replacing the EGR valve assembly anyway as I picked one up pretty cheap from them and I didn't like the fact the one in there at the moment leaks & sticks...

I hope I haven't done anything to the engine letting it run with the faulty sensor, if it was running lean...I've heard that can kill engines

[93_Clubman]

Good to hear it seems the prob has been identified - don't think it will have caused any long term probs as the EGR is blanked off/ removed by some, particularly in the US.

Btw, as you mentioned earlier that you're trying to get your head around all this stuff, you might be interested in the following:

In open loop for cold start & WOT (wide open throttle) or approx 4k+ RPM.
In closed loop once operating temp reached up to approx 4k RPM.

Quoted from Miata.net:
"Closed loop = O2 sensor rules, tune should be near perfect
Open loop = All other sensors rule (O2 is ignored) tune tends to be rich.

If the TPS is not closed or wide open the ECU looks to the CAS & AFM to see if you're cruising.
If so, it goes into closed loop with the O2 sensor & ignores the others.
If you accelerate, it goes open loop & uses all the sensors to set the tune (rich).
When you close the throttle above 2K RPM fuel is cut altogether."

So, with your TPS or O2 sensor disconnected you're running in open loop only.

[doctor_bob]

The EGR wasn't what I was worried about causing damage, but rather running on a faulty O2 sensor; if it detected incorrect oxygen levels it could potentially overly lean or rich the mixture, is this correct?

It would seem the Miata.net information only further confirms the o2 sensor being (at least partially) to blame. The bogging down throttle would instantly dissipate when changing from "closed loop" to "open loop" territory in your quote, so I guess this is the point at which it cuts the crappy sensor readings out of the picture. Makes perfect sense!

edit: Just got me thinking, if it's running great with the o2 sensor disconnected, forcing it into open loop all the time, am I correct in thinking this would also cut the TPS out of the picture? ie, is this another thing I should be checking for faults?

[93_Clubman]

if it detected incorrect oxygen levels it could potentially overly lean or rich the mixture, is this correct?

An increase in fuel consumption. i.e. running rich, can be a sign of a faulty O2 sensor.

am I correct in thinking this would also cut the TPS out of the picture? ie, is this another thing I should be checking for faults?

Worth testing/ checking if the problem persists.

[doctor_bob]

OK! I think this is actually finally sorted... although I'm hesitant to jump to conclusions too quickly until I've let it completely cool down overnight and dríven it for a bit tomorrow.

Got a working O2 sensor in there, and the throttle response was back and it didn't sputter or dip on idle like it had before the whole saga began (ie pre injector cleaner & egr valve clean), which was great...but straight away noticed some issues with the revs being too high:

From cold it went to about 1800rpm and slowly edged down to about 1300 until the oil temperature was registered on the guage, at which point I started driving. After driving for a bit, I assume when the car reached normal operating temperature, it would drop to around 1200-1300 when first coming to a stop but then ease back to around 850 after a couple of seconds of not touching the throttle. I started thinking about the ambient idle adjustment thing I was told not to bother with as the culprit. Then I started thinking about how it was the most promising lead from google alone from the point of a view of a computer guy punching in keywords, before talking to you guys, and the fact that the guy I bought this from was in the same industry as me... seemed a logical step to think he'd adjusted the revs higher, and it had worked to a degree in terms of compensating for the drop but not really addressing the issue, and perhaps we'd solved it with the EGR valve (it was incredibly filthy and there were large particles around the valve seal. I wouldn't be surprised if it didn't fully seal pre-clean) or even the injector cleaner.

So after driving it home I thought it couldn't hurt to try following that guide after coming to the conclusion it was the same GND/TEN combination to adjust the idle in an NB as in the guide from various researching... Shorted them and straight away the revs jumped to around 1250 or so. Adjusted as per the guide except downwards obviously, and left it just a smidge higher than the ECU idle as I didn't want to overdo it... took it for a drive and it feels completely normal again, no high idle revs, no idle dip, full range of throttle response without bogging down. I'll know for sure soon, but it looks promising to my (admittedly limited) logic.

[93_Clubman]

Yeah, it's always prudent to fault-find issues from factory or default settings or setup, particularly if intermittent. Once it's running right in that configuration, then by all means adjust, modify, or adopt a different approach to testing or checking. From my limited experience in small scale ICT, this also seems to be a common approach.

[doctor_bob]

I'm feeling pretty confident now. Forgot to mention, I kept the old EGR valve in there when adjusting everything last night. From cold this morning it was a marked improvement, probably about as good as when I bought the car, but still some slight rev seeking (give or take less than 50rpm) I noticed now that I know what I'm looking for... so I swapped in the EGR valve from mx5mania after giving it a quick soak in carby cleaner (second hand part...) in doing so noted the action & seal of the valve & piston was a HUGE improvement on the one that came with the car. No leakage out the piston either. Feeling pretty confident this time around I swapped it in and the revs were nice and stable right from start. Acid test will be coming home from work this arvo.

Thanks to everyone in the thread for their tips, especially mr 93_clubman, whatever your real name is, thank you hugely for your help....digging up some of that info around open/closed loop; Would not have found that, and wouldn't have solved the problem as quickly without knowing. If you're in Sydney let me know, I think I owe you a few beers

[93_Clubman]

No probs - happy to help as it can be frustrating & time consuming trying to prolem solve intermittent issues when new to a particular car.

[doctor_bob]

More for the purpose of future reference... the idle dip came back yesterday after a small window of driving well on the cleaned EGR valve. Not so sure now whether it actually helped or if the car was just having a "good day" (I mean it was an intermittent fault in the first place prior to the O2 sensor debacle). The same process of disconnecting the O2 sensor restored good idle function. At this stage I'm going to rule out the a faulty replacement O2 sensor under the assumption the previous symptoms associated with it were quite different. So I guess that leaves the TPS module... I suppose I'd look again at the O2 sensor again if the TPS module doesn't help. I think I have all the necessary references to perform a replacement, so yeah, more of an FYI progress update in case this thread can be helpful to others in the future...

Edit: Actually I do have one question, can anyone tell me if there would be any realistic chance of causing more damage to the car by simply driving around with the O2 sensor disconnected until we resolve the issue completely (ie permanently run on open loop)? If there's any risk I guess I'll just garage it for now... but if I CAN keep driving it without consequence... well... you know, it's an MX5... so yeah

[93_Clubman]

Edit: Actually I do have one question, can anyone tell me if there would be any realistic chance of causing more damage to the car by simply driving around with the O2 sensor disconnected until we resolve the issue completely (ie permanently run on open loop)?

It will just run a bit rich like the many cars out there whos O2 sensor isn't functioning, as they tend to go after a period of time/ kms travelled.

[doctor_bob]

Sorry for old thread bump (again). I think this is finally fixed, thought I should update this in case someone else uses the thread for future reference.

Found a brand new TPS module online from the US, got it including shipping for about $40 (!)

Finally put it in yesterday after waiting a few weeks for slow international postage service... looks like this did the trick. The throttle response seems slightly "tighter" too. Fingers crossed...

...of course now that I've said something here it's probably going to present itself again in a new and even weirder way

[93_Clubman]

Tks for letting us know - fingers crossed.

[doctor_bob]

So...it's back, getting worse again...

Each thing I did made an improvement in one way or other though;

- Replacing the EGR valve made it so it wasn't violently shuddering when it did droop, and it was all gunked up; looked like it wouldn't make a seal around the valve, so I'm not surprised at that result
- Replacing the TPS module made the throttle response better, and it seeks around less. Now when it's drooping it will be fairly steady in whichever RPM it decides to settle on.

I'm going to pull the whole throttle body off this weekend hopefully and have a close look at the whole thing. On the plus side these issues have forced me to get a better understanding of what's going on under the hood, so it's not all bad. By process of elimination it must be the IAC valve, right?

If it's not fixed by giving the IAC valve a clean I'll look at replacing it... if that doesn't work I'm going to take it to some experts like mx5mania or somewhere because after that I'm all out of ideas (and expertise...) The IAC does make sense though, right? Here's hoping it just needs a clean...

[93_Clubman]

By process of elimination it must be the IAC valve, right? ... The IAC does make sense though, right?

Haven't read back through everything, but reluctant to say this must be a particular thing. Especially as doing so relies on correct reasoning so far on a vehicle (NB8A) which is well known for elusive & sometimes multiple issues like this. However, it's certainly worth checking the IACV.

If you haven't already it's also worth checking the wiring to the coilpack as it deteriorates where it deflects just prior to plugging into the coilpack. Apparently the combination of deflecting & not being well supported in this particular spot, probably compounded by alternating high & low temperatures, weakens the wires. It's not immediately apparent to the eye, but if you move this part of the wiring loom around while the motor is running you'll get a response in terms of idle if there's a problem. Good luck!