rusefi.com

[first part deleted for uselessness]

In order to do more troubleshooting on the Accel enrichment, could you change the "debug fields" under "Engine" -> "Base Engine Settings" to "TPS accel enrich" and drive around and log? If you find a way to make the car stumble under sudden throttle-movements, that would be helpful

russian wrote:I know what's going on

Just updated https://svn.code.sf.net/p/rusefi/code/trunk/firmware/controllers/algo/obd_error_codes.h file mentioned on the minimalistic
http://rusefi.com/wiki/index.php?title=Manual:Error_codes page

I bet your configuration has grandfathered both TPS low and high value detection thresholds as zero. My bad these settings were not exposed in the TS project - just commited, these would appear below "tps adc channel" setting. You probably want the defaults which are -10% for too low error and 110% for too high error.

Thanks alot! That cleared up the error.
I updated to Feb. 18 firmware Saturday AM for a trip and I drove 500 miles on it. This is working very well now. I'm on the RevD board and have closed loop fuel turned off.

@stefanst I tried to make the TPS debug setting as you asked. See the log attached from the last leg of the trip.

Thanks for the log. Unfortunately the TPS-accel fields were not captured. I'm not 100% sure about this, but you may have to flash the "debug" firmware version for this.

stefanst wrote:I'm not 100% sure about this, but you may have to flash the "debug" firmware version for this.

Nope. "debug" vs "release" versions of firmware is about lower lever constraint validations and compiler optimization level. Firmware debug fields work regardless.

Logging profiles is the answer. Duh!
Please include "debug f1" through "debug f7", as well as "debug i1", "debug i2" and "debug i3" in your logs.

Actually, I suggest to just log all available parameters while performing basic tuning. This way we get it all.

stefanst wrote:Logging profiles is the answer. Duh!

Actually I think it's my fault. I suspect we are looking at a SD card log. SD card logs do not have debug fields https://sourceforge.net/p/rusefi/tickets/355/

russian wrote:

stefanst wrote:Logging profiles is the answer. Duh!

Actually I think it's my fault. I suspect we are looking at a SD card log. SD card logs do not have debug fields https://sourceforge.net/p/rusefi/tickets/355/

Yes, these are sdcard logs. You mean the java console will log the fields now? I can do a short run of that tomorrow.

tomiata wrote:You mean the java console will log the fields now?

yes, console and TS logs have these fields.

russian wrote:

tomiata wrote:You mean the java console will log the fields now?

yes, console and TS logs have these fields.

Here is a short test console msl file to see if the data is showing up.

Yes

I completed another long trip over the weekend driving about 1000 miles total, and I saw no problems with the firmware this time!! This is with Feb. 18 firmware, revD discovery, and no closed loop fuel enabled.

I was trying to tweak the tune for better fuel economy and got it up to 29mpg. That is a little better than the stock ECU on this car, but a Miata 1.6L should be able to do better than that for highway driving. I had a '92 Miata a few years ago that would get 30-33mpg. I need to look into another problem with using oil I thought was leaking, but might be burning it. \

I saved off logs from this trip and the earlier trip in December to this drive folder:


@stefanst these are sdcard logs and would not have the TPS accel debug data you might be looking for.

Beware fuel economy attempts. It is common that fuel is used to cool your cyl. Running lean can lean can cause heat build up issues, which can lead to catastrophic failures.

kb1gtt wrote:Beware fuel economy attempts. It is common that fuel is used to cool your cyl. Running lean can lean can cause heat build up issues, which can lead to catastrophic failures.

Thanks for the word of caution. I was keeping an eye on AFR driving, and taking action if it got much over 16. Here is my AFR target table. I noticed when cruising at 75ish with a/c on and flat terrain it was over the edge of what I was setting as cruise range at about 3800 rpm and 80% load. I expanded the range one cell further. Does it look safe enough?

A lot of our friends over at miataturbo.net run lean to very lean cruise regions without any issues. As long as you're not overheating or getting knock, you're fine. If you're running on the edge of knock, I would have a buddy drive with you with a set of detcans to make sure you're not blowing up your engine: https://www.enginebasics.com/EFI%20Tuning/Det%20Cans.html

I'm using harbor freight det cans:
But for that you need a knock-sensor. On my buddy's old Ford pickup truck we just glued (JB-Weld FTW!) a $5 VW/Audi det-sensor from Ebay to the side of the block near the head. Works just fine.

Could you post your last tune please? Or are there no changes from the previously posted version?
How did the last drive feel? Any hesitation, especially under acceleration at low rpms?

stefanst wrote:Could you post your last tune please? Or are there no changes from the previously posted version?
How did the last drive feel? Any hesitation, especially under acceleration at low rpms?

Driving feels pretty good now. It took a while to get it back to a good state after recalibrating sensors and updating to newer firmware. I do have some knock when accelerating at low rpms I have noticed a couple of times. Latest tune attached. I just updated TPS/TPS enrichment table, but have not tried driving on it yet.

Also, I did a compression test today to check if it would explain losing oil. Numbers are:
1, 210
2. 210
3. 190
4. 180

And first compression cycle numbers were, 95, 70, 90, 50. The 50 worried me. How does that look for thinking about going turbo? I was hoping to work on that soon.

#4 always seems to suffer the most. It's the one most impacted by the "unique" Miata cooling system. Absolute numbers are not as imprtant as variations between cylinders. our highest is 210, lowest is 180. So that's about 15% difference. It's not awesome, but definitely not terrible enough to abandon turbo plans.

One word of advice though: 1.8l responds MUCH better to turbo than 1.6l. Considering you can usually get a decent 1.8l for under $500, I would first do that upgrade and then turbo it. Otherwise your exhaust manifold and maybe downpipe may not be reusable.

Hai guys! Number four is indeed usually a bit low but yours is okay. First compression cycle numbers aren't very useful because the engine doesn't always stop at the same point. It may already have been right before the compression stroke on the "50" reading, which doesn't give the starter time to spin the engine. It's best to look at the pressure difference between the cylinders, a variation of +-10% is normal-ish. Yours is at 16% which is kinda okay.
My turbo 1.6L (with good a power band and quick response) has a variation of 12% so don't give up on those turbo plans so easily. Turbo miata's are sooooooo much fun. I've killed M3's with a 1.6 turbo, I can definitely recommend the upgrade to turbo

By the way, an AFR of 1:15.4 is generally accepted as best for fuel economy, going over that reduces efficiency and makes it misfire more.

Hi Thommm, good to hear from you. I'm looking at options on turbos, but I have gotten busy with work and things. I'll put this on hold and keep this car as is for now.

The nice thing about the early Miata's is that emissions inspections are not required on cars over 25 years old in Texas, https://www.dps.texas.gov/RSD/VI/inspection/inspectionCriteria.aspx (And looks like not required at all if you live outside the major cities.)