mopnbucket

Hey folks, I've got an idle issue after doing some work and I think I know what's up but need to figure out WHY it is the way that it is.

My setup: '93 Miata with OEM engine, harness, and 5-speed transmission. MSPNPPRO / MSPNP3.

When I start the car (cold), the engine idles around 1500 rpm with the clutch pedal held in. If I drop the clutch out in neutral, however, the idle picks up to about 1900 rpm. When I depress the clutch pedal again, RPM falls back down to 1500.

My understanding is that I probably have a busted neutral safety switch circuit somehow. I'm trying to figure out how to diagnose which piece of the puzzle might be busted.

• I just changed my clutch fluid, master cylinder, slave cylinder, and added a braided stainless line, so if that new hydraulic setup is not working, the clutch isn't actually getting disengaged.
• The NSS switch itself might be toast.
• The megasquirt might have an issue. It was working before, but I wonder about this because I think the shop forgot to disconnect my battery when they were doing some welding (roll cage) and looks like they fried my AFR gauge, so wondering if MS could have been impacted too (though it seems fine).

Any tips for how to best diagnose these issues? (especially the third one: how do I find out how the NSS shows up in Tuner Studio / how it's labeled?)

Thanks!

EDIT: attached tune and datalog to original post. Thanks for the suggestion!

SpartanSV

When asking a question about how your car is running ALWAYS post your tune and a log of the issue.


You don't have a NSS issue. No idea why you're thinking that.

mopnbucket

Cool cool cool. I'll get a log when I'm home next. Tune is attached.

For the record though, I'm running the same DIYAutotune-provided basemap that seemed like it was working fine before and the only things that have changed between "working fine" and "broken" are the things I called out in my original post. Just trying to help decrease the scope if I can. Usually when something breaks and you just touched the thing that broke, the thing you just did is a good place to start investigating, no?

Well, I am a total #$% newb, so no surprise if I'm off-base.

That said, I'm thinking that because a bunch of folks have had a similar problem with their miatas and cited a broken NSS as leading the ECU to think the car is in gear when it's in neutral and adding fueling to compensate for the load. But who knows? "I'm an idiot and screwed up my clutch line swap" is just as viable an answer.

Ted75zcar

iTrader: (2)

You need to confirm that you are in closed loop (CL) idle during both of these states. Refer to the megasquirt manual to figure out how to do this via the status fields.

mopnbucket

Attached is the promised datalog.

Gotcha, based on the current tune I think it's open loop (because basemap). I can easily change this, but can I ask why this matters? Wouldn't that be EVEN MORE reason why the RPMs shouldn't jump based on the clutch (i.e., if there was slightly more load with the clutch engaged and the engine spinning transmission internals in neutral, wouldn't idle speed drop with the same amount of fuel being put in the engine without closed loop feedback)?

Thanks for helping me learn this stuff!

SpartanSV

Your log doesn't show us the change you're talking about. It just shows the engine starting and it immediately jumps to a high idle.

mopnbucket

Keep scrolling my friend:

SpartanSV

I see a 1500-1900 oscillation. Clutch in is 1500, clutch out is 1900?

You said the car idles at 1000 rpm clutch in. That isn't happening in your log. Which is it?

If your car actually idled at 1000 rpm before messing with the clutch hydraulics and it now won't idle anywhere near 1000 rpm then I would bet lots of money on you having a vacuum leak

mopnbucket

Correct.

You're correct. My initial post "~1000" was for memory and off-base. The log shows the actual behavior. I've fixed my original post.

mopnbucket

Hey folks, seems like this thread is more open-ended than it should be, so let me scope it down to a much more concrete, newb question:

I have an MSPNP Pro (MS3) for a 93 Miata. If I want to figure out how a thing is wired into the megasquirt and then find it in tuner studio, how should I go about it? Put it another way how do I map a pin in the miata harness to the name of an input/output on the MS, and then how do I map that to the name of a thing in TunerStudio? The specific thing I'm trying to figure out in this specific case is the clutch switch / NSS which are 1V in the OEM harness, but I've run into this mapping problem enough that I'd love to learn how to figure this out more generally.

AFAIK, all I have to work with is this page describing specifically how my MSPNP Pro is set up:

MegaSquirtPNP by DIYAutoTune.com

And the MS3 manual:

MS3 base V3.0 Hardware Manual 1.5

How do you go about piecing things together? Are there other resources I should be looking to?

Thanks.

SpartanSV

Did the car idle at 1000 rpm before the current issue?

Your idle speed change is coming from the 7+ degree shift you have between clutch in and clutch out. This is being caused by the gigantic step in the ignition table between 1500 and 2000 rpm. You have a slight change in MAP between clutch in and clutch out and you have such a huge jump in timing that it's changing your idle speed significantly.

Whatever you've seen about a NSS causing this issue isn't going to apply here. That may be a function of how the stock ecu works, but that isn't how your system works.

Ted75zcar

iTrader: (2)

~cough~

SpartanSV

He's using open loop idle.

Ted75zcar

iTrader: (2)

I know...

SpartanSV

Then enlighten me please. Are you suggesting he turn it on?

Are you referring to the idle engine state?

You have me thoroughly confused.

Ted75zcar

iTrader: (2)

of course. Idle is a function of several things, load, timing, fueling, manifold pressure, ... why would anybody expect it to be stable if these parameters weren't actively managed and properly tuned? Even the stock 1989 ECM implements closed loop idle.

1000 rpm idle should have been the first clue.

I will say that in my days I have seen a speed density reference that was pinched in the pedal assembly causing the engine to think it saw nearly 100kpa load and go crazy rich.

SpartanSV

My point is that he's always been running open loop idle and he didn't think there was an issue until now. If he previously reached a 1000 rpm idle and now he can't then there's likely a mechanical reason for that. Most likely a vacuum leak. Do you disagree?

If he always had a high idle then I agree the next step is closed loop idle tuning.

The pinched reference is a common NB thing. Doesn't apply to OP.

Ted75zcar

iTrader: (2)

As you pointed out (I have not looked at the logs or the tune), there is a 7 degree timing shift (presumably due to the change in ign load). That alone can change idle speed dramatically.

Edit: the whole switch topic can easily be verified simply by disconnecting the switch plug and confirming the effect either stays or goes away

Ted75zcar

iTrader: (2)

As far as a vac leak, the posted screen shot doesn't have the marker values visible so it is difficult to say, but it appears as if load decreases during the higher rpm areas not increasing.

mopnbucket

Hey Ted, thanks for all the input. I totally agree that I should move to closed loop.

That said, the context is that this is not a tuned car (yet). I'm running the base map as I'm just getting this car running after a substantial set of changes and trying to figure out why this idling behavior changed when the tune did not (i.e., implying that something else has changed unintentionally). I will of course tune properly once I get things working at baseline, including moving to closed loop idle. I'm just not there yet...trying to figure out what I fucked up before I go changing more stuff.

Duh good call. Ok I'll do that just to cross one more thing off the list, even if it's not likely the issue.



Right, so it looks like both PW/VE and MAP decrease after the clutch is released, but there's a TINY blip upwards in MAP (and thus fueling) presumably right as the clutch gets released. So is the idea that the tiny increase in MAP (maybe a vacuum leak) causes an increase in fueling which drives an increase in RPM and once that RPM goes up then we get a big jump in timing that sustains the RPM even as MAP and PW/VE drop? Here's a screenshot with more values highlighted for those that can't look at the log (ignore the bogus AFR in the first two; wideband hadn't heated up yet):

Clutch pedal depressed:


Blip in MAP and fueling as clutch pedal is released:


After clutch pedal is released: