[rusEfi] Miata NA 1990, 1.6L stock #21

Your chance to introduce yourself and your vehicle
Post Reply
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: Miata NA 1990, 1.6L stock #21

Post by stefanst »

I'm looking at your log (2016-10-08-14-24).
There are quite a few things that seem odd. Some or all of this may be explained in the previous pages, but I was too lazy to read it all. So my apologies if I'm asking something that has already been answered.

- The log does not show an RPM reading. Since rpm is calculated from the trigger (cam in your case, I believe), there is no way for the ECU to inject fuel or spark a plug if there's no rpm signal. Was that log taken while cranking?
- If it was taken while cranking, then you have a problem reading the cam-sensor
- Your log doesn't include pulsewidth
- MAP of around 10kpa does not make any sense for an engine at idle or cranking. Should be at least 25kpa or so. Check your sensor calibration. Does MAP change when cranking (should be lower than sitting still)
- At the very beginning, your MAP is at 133kpa. This also makes not sense (you'd have to be 3000m below the earth's surface for that)
- You are using MAP and not AFM- correct?

Good luck!
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Miata NA 1990, 1.6L stock #21

Post by AndreyB »

I have a theory that RPM is zero because with new firmware we have a fatal error and trigger events are ignored to reduce potential damage. I am now working on https://sourceforge.net/p/rusefi/tickets/316/ so that each fatal error gets a unique code which would go into the log file same way as non-fatal error codes are already there. Should be ready tonight.

I am not sure why I do not see the fatal error message in the .csv error log but that does not really matter. Hoping to fix https://sourceforge.net/p/rusefi/tickets/316/ and test later tonight.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Miata NA 1990, 1.6L stock #21

Post by AndreyB »

https://sourceforge.net/p/rusefi/tickets/316/ is now implemented & tested. Please update firmware to latest and publish a .msl file.

I am hoping for something like errorCode 6133 "pin already used" which would mean invalid configuration, if that would be the case I would ask for the CurrentTune.msq file
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
tomiata
contributor
contributor
Posts: 234
Joined: Sat Aug 02, 2014 2:30 am
Location: Texas

Re: Miata NA 1990, 1.6L stock #21

Post by tomiata »

My starter motor seems to have died. I'll test after getting another starter. It may be the weekend before I can try it :-(
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Miata NA 1990, 1.6L stock #21

Post by AndreyB »

One step forward, two steps backwards :(

Did you try the latest firmware? Still interested in the error code even without cranking.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
tomiata
contributor
contributor
Posts: 234
Joined: Sat Aug 02, 2014 2:30 am
Location: Texas

Re: Miata NA 1990, 1.6L stock #21

Post by tomiata »

Hmmm, maybe the starter motor is not bad after all. I reseated the plug for the starter solenoid on the starter, and she will crank now.

I installed 10/11 firmware and will not start, similar to last try. But it does get some puffs and act like it is close to starting.
Attaching logs.

And I went back to 9/5 firmware as a sanity check, and she starts right up.
Attachments
MAIN_rfi_report_2016-10-12_22_13.csv
(2.1 MiB) Downloaded 645 times
2016-10-12_22_13.msl
(260.89 KiB) Downloaded 638 times
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: Miata NA 1990, 1.6L stock #21

Post by stefanst »

It looks like its actually catching. You're getting waaaaay above cranking rpm (250 or so).
Why it then dies is not clear. I'm still missing injector pulsewidth and ignition advance in your log.
If advance and PW are the same/comparable between newest FW and the last known-to-run version, you should be OK.
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Miata NA 1990, 1.6L stock #21

Post by AndreyB »

2016-10-12_22_13.msl shows cranking, then RPM gets above cranking threshold so it goes into running fuel logic, and at this point it looks like very little fuel is injected. MAP value is a bit strange - it does not go back up once you stop cranking, maybe that's actually how it works and manifold in fact still has vacuum?

I've also noticed you do not touch TPS according to that log, sometimes a bit of throttle helps.

Just added another datapoint into console log files, can you please download console version 20161012 and record a log with the older version, the one which starts OK, just to compare?
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Miata NA 1990, 1.6L stock #21

Post by AndreyB »

stefanst wrote:I'm still missing injector pulsewidth and ignition advance in your log.
When dwell drops from 8 to 5 is the transition from cranking to running mode. In running mode "runningFuel" would be used

/**
* Total fuel with CLT, IAT and TPS acceleration corrections per cycle,
* as squirt duration.
* Without injector lag.
* @see baseFuel
* @see actualLastInjection
*/
floatms_t runningFuel;

0.3 divided by two injection would be a very low value, pretty much only injector lag would be used. actualLastInjection datapoint which is

/**
* this one contains total resulting fuel squirt time, per event
* With all corrections and injector lag. See also baseFuel
*/
float actualLastInjection; // 64


I've just added into console logger, need an hour for it to build.
Attachments
screenshot.png
screenshot.png (39.76 KiB) Viewed 15299 times
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
tomiata
contributor
contributor
Posts: 234
Joined: Sat Aug 02, 2014 2:30 am
Location: Texas

Re: Miata NA 1990, 1.6L stock #21

Post by tomiata »

Good news, starter worked enough to do a good crank start with 9/5 firmware again. (Btw, I forgot to connect TPS last time.)
Bad news, starter quit working after that. Could not try cranking with 10/13 firmware.

But I attach logs from old and new fw in case they are useful.

I'll work on changing the starter, doesn't look easy on a Miata :-(
Attachments
Oct_13_fw_MAIN_rfi_report_2016-10-13_19_41.csv
(474.37 KiB) Downloaded 616 times
Oct_13_fw_2016-10-13_19_41.msl
(48.07 KiB) Downloaded 633 times
Sept_5_fw_MAIN_rfi_report_2016-10-13_19_37.csv
(327.35 KiB) Downloaded 607 times
Sept_5_fw_2016-10-13_19_37.msl
(102.74 KiB) Downloaded 618 times
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Miata NA 1990, 1.6L stock #21

Post by AndreyB »

Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Miata NA 1990, 1.6L stock #21

Post by AndreyB »

I believe some issue with MAP sensor decoding is the root cause of not running. Note how MAP sensor is sane while it starts and insane when it does not. It drops from hi value to a flat line low value even before you start cranking (maybe it drops once you start cranking but before firmware registers RPMs - but again, why is not the case in the log where it starts fine). And the flat line does not go back even once you are done cranking!

I am looking at the history of https://svn.code.sf.net/p/rusefi/code/trunk/firmware/controllers/sensors/map.cpp and the last meaningful change I see was at June 26th.

Once you have cranking, please invoke mapinfo and analoginfo command before/while/after cranking. Hopefully we would see MAP sensor voltage and MAP sensor decoding details. We are missing something trivial :(
Attachments
not_starts.png
not_starts.png (35.29 KiB) Viewed 15290 times
starts.png
starts.png (41.15 KiB) Viewed 15290 times
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
tomiata
contributor
contributor
Posts: 234
Joined: Sat Aug 02, 2014 2:30 am
Location: Texas

Re: Miata NA 1990, 1.6L stock #21

Post by tomiata »

Got the new starter on, 3 hours by myself and really needed some swivel sockets and longer socket extensions.

I can crank now, and I tried doing mapinfo and analoginfo before, during and after crank. I could not get both command entered while cranking though. The engine starts to catch on but immediately dies.

See if this exposes any answers. Thanks!

before:
Fri Oct 14 21:26:11 CDT 2016<EOT>: postMessage EngineState: confirmation_mapinfo:7
Fri Oct 14 21:26:11 CDT 2016<EOT>: postMessage EngineState: instant value=133.07kPa

during:
Fri Oct 14 21:26:18 CDT 2016<EOT>: postMessage EngineState: instant value=9.66kPa
Fri Oct 14 21:26:18 CDT 2016<EOT>: postMessage EngineState: map type=2/MT_MPX4250 MAP=9.66kPa
Attachments
2016-10-14_21_25.msl
(114.22 KiB) Downloaded 636 times
MAIN_rfi_report_2016-10-14_21_25.csv
(946.92 KiB) Downloaded 597 times
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: Miata NA 1990, 1.6L stock #21

Post by stefanst »

It's still catching and then dying. So I assume the amount of fuel its getting is wrong.
I still don't understand a lot of what's happening in your logs:
- Your MAP resting is 133- that's definitely wrong- Should be around 100
- Your MAP, while cranking, is around 10 - that's waaaaay too low
- Your MAP doesn't go back up between cranking attempts as it should

Since your MAP signal doesn't make much sense and it seems that your having a problem with the amount of fuel delivered, that's where I'd start troubleshooting. Was your MAP signal always like this? How was it calibrated? Which sensor are you using?
puff
contributor
contributor
Posts: 2961
Joined: Mon Nov 11, 2013 11:28 am
Location: Moskau

Re: Miata NA 1990, 1.6L stock #21

Post by puff »

could it be so that after the firmware upgrade something got mapped to wrong pins?
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Miata NA 1990, 1.6L stock #21

Post by AndreyB »

I am convinced that that's a software issue. 9kPa which goes down and never gets back is actually decoded zero volts, zero volts is a pretty special case :) Either +5v are lost of something else is funny. For instance https://sourceforge.net/p/rusefi/tickets/293/ was fixed mid-September. I am planning to get Sep 12, Sep 19 and Sep 26th binaries ready within the next hours so that we can start getting closer to what change in the firmware has broken everything. Good news is we only have a month range of versions with and without the issue.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
tomiata
contributor
contributor
Posts: 234
Joined: Sat Aug 02, 2014 2:30 am
Location: Texas

Re: Miata NA 1990, 1.6L stock #21

Post by tomiata »

stefanst wrote:It's still catching and then dying. So I assume the amount of fuel its getting is wrong.
I still don't understand a lot of what's happening in your logs:
- Your MAP resting is 133- that's definitely wrong- Should be around 100
- Your MAP, while cranking, is around 10 - that's waaaaay too low
- Your MAP doesn't go back up between cranking attempts as it should

Since your MAP signal doesn't make much sense and it seems that your having a problem with the amount of fuel delivered, that's where I'd start troubleshooting. Was your MAP signal always like this? How was it calibrated? Which sensor are you using?
MAP sensor is an MPX4250. I didn't do anything to calibrate it, just selected MPX4250 in TS. After I set it up and switched to Speed Density months ago, it just worked. I see there are calibration settings, but I didn't touch those since it was cranking and running (if it ain't broke ....) But then I noticed the spark misses, and later the fuel misses. @Thommm also saw spark misses. I don't know if he also confirmed the fuel miss issue. @russian pointed out the MAP value out of range a while back, but I haven't tried calibrating MAP yet. I don't think the miscalibration is causing the misses, but maybe I'm wrong.

Looking at MAP calibration I understand that 1 atmosphere is about 101kPa. Here are measurements:

Switch on not Running:
Sat Oct 15 12:07:19 CDT 2016<EOT>: postMessage EngineState: MAP ADC6 fast PA6 adc=0.99/input=2.47v/divider=2.50
Sat Oct 15 12:07:35 CDT 2016<EOT>: postMessage EngineState: map type=2/MT_MPX4250 MAP=132.80kPa

Measuring MAP sensor output with a DMM shows 1.74v. Rusefi shows my MAP input as 2.47v, which is wrong, and now I remember that I probably removed the pull up/down resistors on that input thinking they were not needed. I should have asked before removing those. I'm learning the hard way here. The 1.74v looks about right from the mpx4250 data sheet graph below:
mpx4250.jpg
mpx4250.jpg (46.31 KiB) Viewed 15225 times
Even with the mis-scaling, the input that Discovery sees will still be linear and follow the mpx4250 curve, right? Should be able to set to a "custom" MAP sensor and adjust min and max kPA and input voltage to match to get correct MAP data input?
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: Miata NA 1990, 1.6L stock #21

Post by stefanst »

MPX4250 doesn't need pull-up or -down resistors, so that's not your problem. What is the voltage at the corresponding pin on the brainboard, as measured with DMM? It should be half of the output of the MPX4250, so 0.87V or so.

One more thing: I believe the standard pin for MAP is PA0, not PA6- are you sure you selected the correct one?
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Miata NA 1990, 1.6L stock #21

Post by AndreyB »

stefanst wrote:What is the voltage at the corresponding pin on the brainboard, as measured with DMM? It should be half of the output of the MPX4250, so 0.87V or so.
I am expecting voltage to be correct before cranking, and wrong after cranking. After cranking the firmware believes the voltage to be close to zero.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Miata NA 1990, 1.6L stock #21

Post by AndreyB »

I suspect it's a firmware defect, can we please try different versions to pin-point the exact change?

Just implemented https://sourceforge.net/p/rusefi/tickets/318/ so please use latest rusEfi console from http://rusefi.com/build_server/rusefi_bundle.zip that would give us firmware version in the .msl log files

Here are three versions between Sep 5th and Oct 10th:

Sep 10th, #10571 https://svn.code.sf.net/p/rusefi/code/firmware_archive/rusefi_firmware_20160910.zip
Sep 20th, #10656 https://svn.code.sf.net/p/rusefi/code/firmware_archive/rusefi_firmware_20160920.zip
Sep 27th, #10707 https://svn.code.sf.net/p/rusefi/code/firmware_archive/rusefi_firmware_20160927.zip
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
tomiata
contributor
contributor
Posts: 234
Joined: Sat Aug 02, 2014 2:30 am
Location: Texas

Re: Miata NA 1990, 1.6L stock #21

Post by tomiata »

I suspect it's a firmware defect, can we please try different versions to pin-point the exact change?
Ok Sept 10 firmware works, starts fine. Sept. 20 firmware, no workie, almost starts and dies.

See logs attached, thanks working on this!
Attachments
Sept_10_MAIN_rfi_report_2016-10-15_20_01.csv
(575.89 KiB) Downloaded 534 times
Sept_10_2016-10-15_20_02.msl
(128.14 KiB) Downloaded 534 times
Sept_20_2016-10-15_20_04.msl
(172.43 KiB) Downloaded 533 times
Sept_20_MAIN_rfi_report_2016-10-15_20_03.csv
(1.25 MiB) Downloaded 546 times
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Miata NA 1990, 1.6L stock #21

Post by AndreyB »

Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: Miata NA 1990, 1.6L stock #21

Post by stefanst »

It's interesting, that even in the log where it fired you still have 132 or so kpa at rest. And, even more interestingly, about 500ms after you start cranking (as evidenced by the battery voltage dropping) MAP drops to 8kpa for a little bit before it recovers to more sensible values. That has to be an electronics, or a firmware problem. My guess is firmware, since we don't get any drops like that while the engine is running normally.
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Miata NA 1990, 1.6L stock #21

Post by AndreyB »

132 kPa is present in both "good" and "bad" versions, that's a calibration issue. But that's a smaller issue at the moment.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
stefanst
contributor
contributor
Posts: 703
Joined: Wed Feb 17, 2016 12:24 am
Location: USA 08530

Re: Miata NA 1990, 1.6L stock #21

Post by stefanst »

Correct- the 132kpa is calibration. But what causes the momentary drop to 8kpa?
tomiata
contributor
contributor
Posts: 234
Joined: Sat Aug 02, 2014 2:30 am
Location: Texas

Re: Miata NA 1990, 1.6L stock #21

Post by tomiata »

FW 10266 works, starts
FW 10267, no start. Looks like the ticket 293 was the culprit.

Thanks!
Attachments
fw_10267_2016-10-15_22_45.msl
(93.63 KiB) Downloaded 527 times
fw_10267_MAIN_rfi_report_2016-10-15_22_44.csv
(718.72 KiB) Downloaded 525 times
fw_10266_MAIN_rfi_report_2016-10-15_22_42.csv
(916.9 KiB) Downloaded 522 times
fw_10266_2016-10-15_22_42.msl
(201.85 KiB) Downloaded 511 times
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Miata NA 1990, 1.6L stock #21

Post by AndreyB »

revision 10627 has changed only one file - https://svn.code.sf.net/p/rusefi/code/trunk/firmware/chibios/boards/ST_STM32F4_DISCOVERY/board.h

Attached is the difference file. No PORTA pin settings were changed! I am a bit puzzled.

Can you please confirm that that there is no miraculous connectivity between pin PA6 (that's your MAP input right?) and any neighboring pins with a DMM? For instance PC4 was one of the ports which was affected by the 10627 change.

I need to leave house for a couple of hours, later today I will make some custom copies of the firmware with different sections of boards.h reverted to the original state. Hopefully we would isolate which exact pin is somehow affecting your PA6.

Another interesting question is why this happens on start of cranking. That must mean that we changing something on start of cranking. Not fuel pump probably since it turns off later. Maybe something like 'trigger error' pin or some other weird auxiliary device pin.
Attachments
board_diff.txt
(29.14 KiB) Downloaded 530 times
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
tomiata
contributor
contributor
Posts: 234
Joined: Sat Aug 02, 2014 2:30 am
Location: Texas

Re: Miata NA 1990, 1.6L stock #21

Post by tomiata »

Can you please confirm that that there is no miraculous connectivity between pin PA6 (that's your MAP input right?) and any neighboring pins with a DMM? For instance PC4 was one of the ports which was affected by the 10627 change.
Oh wow. It shouldn't be a meter probe. I only attached that after seeing the problem. I'll pull it out and examine my wiring on the bench. It will be late this afternoon before I can get to that. Let me do that next, thanks.
tomiata
contributor
contributor
Posts: 234
Joined: Sat Aug 02, 2014 2:30 am
Location: Texas

Re: Miata NA 1990, 1.6L stock #21

Post by tomiata »

Can you please confirm that that there is no miraculous connectivity between pin PA6 (that's your MAP input right?) and any neighboring pins with a DMM? For instance PC4 was one of the ports which was affected by the 10627 change.
Yes, PA6, inp10, is the pin I'm using for MAP. I looked at it on the bench and probed around. I don't see any stray connections or coupling to other pins. There is a pull up on inp10 I thought I may have removed, but it is there. And no pull down populated. And I probed signal from inp10 to the PA6 pin and it seems fine.
inp10.jpg
inp10.jpg (24.51 KiB) Viewed 15175 times
I put the board back in the car and with just switch on MAP output reads about 1.74v and PA6 pin reads about .93v. I tried running with the 10626 firmware and probed the discovery pins PC4 and PC5 while reving the engine to see if MAP could be coupled to those pins. I don't see any voltage changes on PC4 or PC5. Are there other pins you think might be affecting it, or just any pins on port B and C?
tomiata
contributor
contributor
Posts: 234
Joined: Sat Aug 02, 2014 2:30 am
Location: Texas

Re: Miata NA 1990, 1.6L stock #21

Post by tomiata »

I was able to build rusefi with old and new board.h and confirm that single file change causes the problem for me.
I built with the old board.h and merged in changes for one port at a time. I narrowed it down to something in the PORT E section (figures it would be the last port trying A thru E)

That's as far as I've gotten so far. You don't need to build trial versions with those changes. I'll look closer at PORT E after dinner.
Post Reply