catalyst

Hello, I am looking for a bit of a sanity check. I have a ’96 with a MS2 DIYPNP(which I assembled). The engine is currently stock with a few exceptions, running Flow Force 640cc EV14 injectors and the MS2 has an external map sensor. Still have the stock AFM in place and using that MAT sensor. Also an innovate mtx-l. I’ve been slowly working on tuning this guy for the last couple months. Bit of a learning curve but it’s coming along. Some prior experience, this will be the second turbo system I’ve setup. Previous was a honda d-series motor in a civic, which turns out was basically easy-mode.

I have run into an issue that I can't seem to resolve with the provided tuning options available. If anyone would be willing look at these logs and offer opinions on my crappy a tuning I would be grateful. After much effort I feel like there might be an electrical issue. The problem occurs after a restart with engine at temp(hot). Initially I identified the issue as a lean idle, afr 17+ on hot restarts. This seems to be a well-documented problem, most were saying heat soak of the MAT sensor or NB’s with return less rail were having their own set of issues. Using the MAT correction table didn’t seem like best way to go. So I ended up using ASE and have added 2500 cycles to ASE and richened up the curve for CLT temps above 170. With this the car idles decent after a hot start albeit 11.5 to 12 afr for the first min or so. As soon at the ASE fuel correction tapers to around 14.0 the idle goes to crap, stumbling and almost dying in some cases. The MAT readback will have cooled to around ambient temps. I am seeing lots of idle variation and poor afr stability. Off idle things smooth out considerably. The ecu just can’t seem to handle idle and never seems to stop struggling unless the car sits overnight. I have logged the issued and see a difference in the data when it's hot, lots of small oscillations in several sensor readbacks makes me think I might have some electrical issue. I am seeing battery voltage also jumping quite a bit. Logs with title "weird noise" will show this toward the end

If I start up the car from full cold it runs well enough. VE table is not perfect but close. EGO is still disabled. The idle control valve is in closed loop and controls idle decent if the car was started cold. In the data logs there are two cases. First log is after I have driven the car around for 20 min and let it sit and idle and is running well at this point. Idle is tight and fuel is in check maybe a tad rich(13.8-14.1).

Turn off, wait 10 min. Upon restarting it’s like a completely different car, acting as described above. Two logs labeled “bad” exhibit the behavior. My feeling after a couple weeks trying to analyze the problem and zero success is leading me to something other than the tune is the root. Any ideas on what is causing this? I have attached 3 logs and my tune, The last log probably shows the most protuberance. I'm scratching my head at the moment



EDIT for links:

Good Idle after cold start and warm up

15 min after good idle

Shows some odd noise during idle

This shows ASE and more weird oscillations toward end of log

the tune

Matt Cramer

Unfortunately, the attachments don't seem to be working properly.

lsdlsd88

you'd better off uploading them somewhere like dropbox and post the link

catalyst

Oh no. Ok thanks for the tip, will get some links up in the next hour or so

catalyst

i edited the original post, added some google drive links

catalyst

So i noticed my Battery Voltage calibration was off by about .6V. I adjusted that this morning and logging a warm up and short drive. I immediately re-started the car, of course my problem still exists but i felt like these two logs may be a bit more clear that what i posted above. In case someone finds time to have a look or a 2nd look at this thread I am going to leave them here.

warm up log, let idle for a min or so at end.
https://www.dropbox.com/s/gwcg60llwh...drive.msl?dl=0

immediate restart, rough idle
https://www.dropbox.com/s/t4qrpru0g7...start.msl?dl=0

tune
https://www.dropbox.com/s/eyh7k3q85k...tTune.msq?dl=0

miguel.sagal92@gmail.com

I am having the same issue. I have a 1999 with mspnppro. It all started happening after I installed the flow force. First installed the ECU. Got it running on a stock Miata with perfect idle and autotuned map. Then I installed my flow force injectors, noticed the same issue happening except mine also slowly develops higher afrs after sitting for a while. It starts at 12 afr and after about 20 min of letting it idle it's at 17 afr. If I find a solution I'll let you know.

aceswerling

I've been wrestling with this issue on my '95 with MS3and ID 1000 injectors for a couple years. Like you, I mostly notice lean AFRs on hot restarts but I also see them when I'm driving if the engine is really hot.

I've made a number of changes, including different injectors, voltage calibration, and fuel temperature adjustments but nothing has solved the problem. The most consistent explanation I've heard is that the fuel injector solenoids change their electrical behavior when they got hot, which affects dead times.

On the other hand, I asked Injector Dynamics about this and they hadn't seen this behavior. Flyin Miata also reported they don't have this problem with the ECUs they support, either with ID 1000 or Flow Force injectors. Finally, it seems to me that OEMs wouldn't accept substantial behavior changes like this in Bosch injectors.

Most recently I installed some heat shields from DEI to try and keep some heat out of the injectors but they haven't made a difference.

All this makes me wonder how the problem can be caused by the injectors, or at least by the injectors by themselves. I'd really appreciate Matt's insight on this.

SchmoozerJoe

iTrader: (2)

I'd take what FM says with a heavy grain of salt.

This was a known issue with heat soak, both on the Hydra and the proto-type FM221 that I was trying to help bring to market.

While we were in Michigan, Greg Banish mentioned that it was due to the air temps not being calculated properly, as we'd see cars start to lean out and go up to two points lean on hot restarts.
There was a CLT to MAT table for the MS3 that Greg was dialing in, on a Miata they were using with an FMII kit on it, but I never got my hands on it.

FM wanted nothing to do with his developed base map for the MS platform and I failed to network properly and get a copy of it myself.

I did notice that FM picked up a thermal heat gun of some sort, after I dropped the project and left Colorado. Not sure how they "solved" it, but again, take what they say with a grain of salt.
We did eventually determine that the Hydra already had a built in calculated air-temp to help with this, so the effects on the Hydra were reduced, though my testing did eventually lead them to relocate the IAT to the IC outlet, instead of directly behind the radiator fans on the TB inlet hose.

I'm sure I have the charts and graphs around here somewhere that I recorded in the shop... but honestly, zero cares are given these days.

Long story short, I'm fairly certain there's at least one CLT to Air temp curve in the MS3 that can be adjusted to help with the hot restarts not leaning out, I'm sure that one of the guys here can weigh in for ya.

aceswerling

Thanks for the information. It's interesting to hear other ECUs have the same problem. Do the Hydra and FM221 handle starts differently than the Megasquirt?

I've moved my IAT to the outlet side of the intercooler to avoid that heat soak problem you noted. FWIW, I've noticed that the lean AFRs show up independently of CLT and MAT readings. As an example, I'll get the hot restart problem in cool air temps and after coolant has cooled off because the engine has been sitting a while. But I'll see lean conditions if the block is still hot, especially if that heat has had time to soak through everything in the engine compartment.

I looked at that CLT to MAT adjustment and people said it doesn't do much. I'm unsure it would help anyway since this behavior is independent of CLT and MAT.

It's funny, but my stubborn engineer mind can't let go of this issue. People have said to let EGO to handle the variability but I don't want to believe that's the only way. It's hard for me to imagine that's how OEMs handle it. Would anybody out there have experience with OEM ECUs to know for sure?