[rusEfi] Mitsubishi 4G63 and 4G92 (#6)

[kb1gtt]

I see this MEMS accelerometer on the discovery, I also see from it's datasheet found here http://www.st.com/st-web-ui/static/active/en/resource/technical/document/datasheet/CD00135460.pdf on page 28 that the default condition for that chip is for Int1 and Int2 is 000 which appears to be configured to short to GND. PE0 and PE1 are connected to INT 0 and INT 1.

It does not look very pleasant to disconnect the traces, so fixing this on the discovery would likely require sending some packets over SPI to configure the LIS302DL, with luck we could get this to stay high impedance. However I don't have confidence we an guarantee this will always stay high, as it appears you can only change it from push/pull to open collector. So I think that it can always pull low. I think it's best to avoid using these pins. Either that or use the brain board instead of the Discovery board, as that brain board does not have this extra stuff on it.

[AndreyB]

Hi Andrey,
I have some questions for configured port for SPARK_1 (Firmware Release on 29 Jul, 2014)
I was test on Tuning Studio, Default configured as PC7=HIGH DRIVER 6. It dosen't work.

I've just checked the default configuration (GND<>PD6 to reset config) and ignition0 is at PC7, so Tuner Studio shows the right stuff.

The default configuration is a bit weird because ignition#0 is routed via the hi-side driver by default and ignition#1/ignition#2 is mapped to low-side drivers which does not make much sense. Anyway, Tuner Studio always shows the actual values and the document can me out-of-date. By the way I've just changed the document.

How do you test it to see that PC7 does not work and PC15 work? First test is to see the frequency at the pin with a multimeter if your multimeter has the Hz mode.

My multimeter shows me 40Hz @ PC7 and nothing @ PC15 with two jumper wires for trigger self-stimulation with the default 1200 RPM.

[Tambralinga]

Hi Andrey,
I have some questions for configured port for SPARK_1 (Firmware Release on 29 Jul, 2014)
I was test on Tuning Studio, Default configured as PC7=HIGH DRIVER 6. It dosen't work.

I've just checked the default configuration (GND<>PD6 to reset config) and ignition0 is at PC7, so Tuner Studio shows the right stuff.

The default configuration is a bit weird because ignition#0 is routed via the hi-side driver by default and ignition#1/ignition#2 is mapped to low-side drivers which does not make much sense. Anyway, Tuner Studio always shows the actual values and the document can me out-of-date. By the way I've just changed the document.

How do you test it to see that PC7 does not work and PC15 work? First test is to see the frequency at the pin with a multimeter if your multimeter has the Hz mode.

My multimeter shows me 40Hz @ PC7 and nothing @ PC15 with two jumper wires for trigger self-stimulation with the default 1200 RPM.

It should be the same but I'm confuse the in there Spartk_1
Ok, I'm got another one error in Dev console in the latest version 20140802 like below...



And could you please to check some LED-5 (STM32-Discovery) RED light status after connected to Tuning Studio.

Thanks to Jared for advice.
Mr.X

[AndreyB]

Looks like that's the same issue as in http://rusefi.com/forum/viewtopic.php?f=2&t=631 - I hoped that I've fixed that last tonight. Do you still have it with current version of the dev console?

[AndreyB]

And could you please to check some LED-5 (STM32-Discovery) RED light status after connected to Tuning Studio.

I've just fixed that - you would need new firmware and fresh rusefi.ini Tuner Studio project file, and you would need to re-configure anything you've configured. Sorry for the bug, I've changed it so it would show the red LED right away and not wait till you connect TS next time if I forget to adjust the thing I've forgotten to adjust this time.

[Tambralinga]

Hi kb1gtt,

Last time test by Skype.
I have a question about the connection SPARK-1.
If the SPARK-1 connect to High Drive, Is SPARK-2 connect to High Drive ?

The coil in the picture, Could you please to explain me How to connect / how to use ?


and



Thanks.
Mr.X

[kb1gtt]

Do you have a schematic or reference voltage for that J722t module? I found this http://www.stealth316.com/2-ignitionsystem.htm which is technically for a J723t. I see that J723t module would want a push of 3V not 12V. If you are driving with 12V you may be providing to much current and may have overheated / broken the ignitor for that coil pack. Can you find a schematic or some kind of reference that indicates what voltage or current you need to drive that with? If not I can make some suggestions on how to drive it with out over driving it. Also you can try 3V and see if it works.

Yes if spark 1 is connected to 12V (or 3V) high side, then all other sparks are also connected to that voltage high side.

Also according to this datasheet the typical dwell / charge time is 4mS http://www.datasheet4u.com/datasheet/J/7/2/J722T_ETC.pdf.html and it sparks on falling edge. If dwell is to long, it can also overheat the ignitor.

[Tambralinga]

I can change SPARK_2 to PC9 ?

[AndreyB]

I can change SPARK_2 to PC9 ?

Absolutely. Once you 'Burn' in tuner studio while engine is not running, you would need to reset rusEfi for the change to take effect.

Did you have a chance to start current version of dev console? Does it work for you now?

[AndreyB]

Before you start sparking, we really need to figure out the exact trigger shape Otherwise the ignition angle could be really messed up. Not only I need to know that we have one tooth and 4 teeth - I need to know exact angle widths and the relation between signal.

Option #1) if you can crank the engine with starter and record digital signal with dev console - publish the log file
Option #2) if you can run the engine with stock ECU and connect rusEfi sniffing in parallel that would be even better, but probably not possible with VR sensor
Option #3) find some documentation online?

For example, the ford aspire trigger is:

Code: Select all

	s->addEvent(53.747, T_SECONDARY, TV_HIGH);
	s->addEvent(121.90, T_SECONDARY, TV_LOW); // delta = 68.153
	s->addEvent(232.76, T_SECONDARY, TV_HIGH); // delta = 110.86
	s->addEvent(300.54, T_SECONDARY, TV_LOW); // delta = 67.78
	s->addEvent(360, T_PRIMARY, TV_HIGH);

	s->addEvent(409.8412, T_SECONDARY, TV_HIGH); // delta = 49.8412
	s->addEvent(478.6505, T_SECONDARY, TV_LOW); // delta = 68.8093
	s->addEvent(588.045, T_SECONDARY, TV_HIGH); // delta = 109.3945
	s->addEvent(657.03, T_SECONDARY, TV_LOW);
	s->addEvent(720, T_PRIMARY, TV_LOW);

I would need to code the same information for your trigger.

[Tambralinga]

The results Engine logs form Dev console.






T1T49971


code CAS : T1T49971

And logs form Dev console.
MAIN_rfi_report_2014-08-05 16_32.csv

https://docs.google.com/file/d/0B2_tOJp8KD51WU0wTUNqc19lWWc/edit

Thanks,
Mr.X

[AndreyB]

Perfect, we are getting closer!

According to the digital sniffer, you currently only have one signal:

That's probably the one on PC6. Where is your PA5 signal, the second trigger wheel?

[Tambralinga]

That's probably the one on PC6. Where is your PA5 signal, the second trigger wheel?

PC6 signal crank
PA5 signal CAM

[AndreyB]

PA5 signal CAM

This one does not reach the firmware.

[Tambralinga]

PA5 signal CAM

This one does not reach the firmware.

Hi Andrey,
Update logs from The latest Dev console below.

https://docs.google.com/file/d/0B2_tOJp8KD51NUJUQ1lmdFlnSDQ/edit

Thanks.
Mr.X

[AndreyB]

Tue Aug 05 22:17:11 ICT 2014: Console 20140803

Please update to latest dev console, because it is a little better with displaying errors like the one you have:

Code: Select all

Pin PA5 requested by wave input but already used by adc input,

You need to change the config so that PA5 is not used by ADC since it's used for trigger input, once you do that you are looking to have C1 and C2 channels on top of digital sniffer report. c1 & c2 would be your two trigger input channels.

[AndreyB]

PS: according to https://docs.google.com/spreadsheet/ccc?key=0Arl1FeMZcfisdEdGdUlHdWh6cVBoSzFIbkxqa1QtZ3c pin PA5 is ADC channel #5, so you cannot use this ADC channel.

[Tambralinga]

PS: according to https://docs.google.com/spreadsheet/ccc?key=0Arl1FeMZcfisdEdGdUlHdWh6cVBoSzFIbkxqa1QtZ3c pin PA5 is ADC channel #5, so you cannot use this ADC channel.

Hi Andrey,

I re-check The CAM2 now it appear, Could you please to see in the log like below.

https://docs.google.com/file/d/0B2_tOJp8KD51N0NvU2N2cjYtcjQ/edit

Thanks,
Mr.x

[AndreyB]

Very good, we are one step closer! Can you please flip CAM1 and CAM2 wires now and record one more log? I will be back in an hour and work on putting the angles into the firmware. The signal with one signal should be PC6 and the signal with multiple signals should be PA5.

What was the real RPM while you were taking this? Was the engine running or just cranking with a starter?

[Tambralinga]

Very good, we are one step closer! Can you please flip CAM1 and CAM2 wires now and record one more log? I will be back in an hour and work on putting the angles into the firmware. The signal with one signal should be PC6 and the signal with multiple signals should be PA5.

What was the real RPM while you were taking this? Was the engine running or just cranking with a starter?

OK, I'm update the logs again. But I'm not sure what the trouble was that started all of this.
https://docs.google.com/file/d/0B2_tOJp8KD51MVJxbUpHUWIzNUU/edit

Thanks,
Mr.X

[AndreyB]

Found a great picture with all the details about the trigger signal shape:

http://www.msextra.com/forums/viewtopic.php?f=4&t=35289#p259279

[AndreyB]

I've implemented the code for you trigger. It has 2 teeth on the inner wheel by the way. You should be good to start this car

1) update firmware

2) read about http://rusefi.com/wiki/index.php?title=Manual:Engine_Type

3) in dev console, execute set_engine_type 16 - this would change all values to the pre-sets from the mitsubishi.cpp so that you can get there in case of an incompatible software change.
(there is a bug - there will be the RED led after the command, just restart firmware - too tired to fix this bug now)

4) now create a new TS project with latest rusefi.ini

5) adjust your output pins and post the screenshots so that I can add that info into the pre-set file https://sourceforge.net/p/rusefi/code/HEAD/tree/trunk/firmware/config/engines/mitsubishi.cpp

This should get your engine running. Cranking fuel 6ms, running fuel 3ms in the template.

I've added some info into my first post in this thread just to make navigation simpler - http://rusefi.com/forum/viewtopic.php?f=3&t=622&p=9772#p9772

[Tambralinga]

I've implemented the code for you trigger. It has 2 teeth on the inner wheel by the way. You should be good to start this car

1) update firmware

2) read about http://rusefi.com/wiki/index.php?title=Manual:Engine_Type

3) in dev console, execute set_engine_type 16 - this would change all values to the pre-sets from the mitsubishi.cpp so that you can get there in case of an incompatible software change.
(there is a bug - there will be the RED led after the command, just restart firmware - too tired to fix this bug now)

4) now create a new TS project with latest rusefi.ini

5) adjust your output pins and post the screenshots so that I can add that info into the pre-set file https://sourceforge.net/p/rusefi/code/HEAD/tree/trunk/firmware/config/engines/mitsubishi.cpp

This should get your engine running. Cranking fuel 6ms, running fuel 3ms in the template.

I've added some info into my first post in this thread just to make navigation simpler - http://rusefi.com/forum/viewtopic.php?f=3&t=622&p=9772#p9772

Hi Andrey,
Let you check the logs please.

https://docs.google.com/file/d/0B2_tOJp8KD51MVJxbUpHUWIzNUU/edit



RPM 1300

Thanks,
Mr.X

[AndreyB]

Is that the right log file?

This one has Ford Aspire all over it. Did you try executing "set_engine_type 16" command in dev console? You need to put it into the command and hit Enter. Once you do that, there probably will be a red LED. Once you restart & reconnect, you should see Mitsu in the title bar to replace Ford Aspire.

Also this log file shows rusEfiVersion,20140803 - not the latest version. Did you update to the last version? Is this the right log?

As an option, close the console and delete all old logs to avoid confusing them.

[Tambralinga]

Is that the right log file?

This one has Ford Aspire all over it. Did you try executing "set_engine_type 16" command in dev console? You need to put it into the command and hit Enter. Once you do that, there probably will be a red LED. Once you restart & reconnect, you should see Mitsu in the title bar to replace Ford Aspire.

Also this log file shows rusEfiVersion,20140803 - not the latest version. Did you update to the last version? Is this the right log?

As an option, close the console and delete all old logs to avoid confusing them.

The latest logs. let you check it again.
https://docs.google.com/file/d/0B2_tOJp8KD51UkZZQlVDbmQ1bm8/edit

Thanks,
Mr.X

[Tambralinga]

Hi Andrey,

new firmware rusefi_firmware_not_sure> 06-Aug-2014.
Update logs for SPARK testing, continued from yesterday.



And the logs, below...

https://docs.google.com/file/d/0B2_tOJp8KD51cXpIRTc0NUFMVjg/edit

In TS can't tuning on real car online, the Engine should be stop.

Thanks,
Mr.X

[kb1gtt]

I understand you are now getting an accurate RPM is that correct?

About TS, can you tune at all? Is the issue that you have to stop the engine when you make an adjustment, or is it that you can't connect?

When you get something like RPM working, can you post some kind of a quick note? Both russian and myself don't understand what your status is, so we don't know what is working and what is not working. I believe your status is that you have fuel and spark controlled by rusEFI and it's running with an accurate RPM signal. However it's a slow task to tune. Is that correct?

[Tambralinga]

I understand you are now getting an accurate RPM is that correct?

About TS, can you tune at all? Is the issue that you have to stop the engine when you make an adjustment, or is it that you can't connect?

When you get something like RPM working, can you post some kind of a quick note? Both russian and myself don't understand what your status is, so we don't know what is working and what is not working. I believe your status is that you have fuel and spark controlled by rusEFI and it's running with an accurate RPM signal. However it's a slow task to tune. Is that correct?

Hi kb1gtt,
Your understand is correct
The problem is... When I change something in TS and burn, The result is LED should be red light and disconnected form FRK. So I solved by reset the board and then FRK is coming to run & start the engine everything should be fine.

Thanks,
Mr.X

[kb1gtt]

About FrankUSB not working TS communications, use PB1 jumper to specify during boot sequence what port will be used for TS communications.

If you remove PB1 jumper, then power the board, during the boot procedure, the firmware will decide to communicate with TS on the FrankUSB and will allow other serial communications on the Discovery USB.
If install a jumper on PB1 shorting PB1 to GND, then power the board, during the boot procedure, the firmware will decide to communicate with TS on the Discovery USB and will allow other serial communications on the FrankUSB.

[kb1gtt]

About burn resets, can you make sure you are using the console and firmware noted below.

console.png (22.73 KiB) Viewed 20347 times

-- The console is found from the SF repository, for me I launch it from here C:\rusefi_sf\misc\binaries\java_console, where C:\rusefi_sf\ is the top level of the checked out repository.
-- The firmware and TS config file can be found here http://rusefi.com/build_server/rusefi_firmware.zip

I have updated to add some notes for setting up TS here http://rusefi.com/wiki/index.php?title=Manual:Software:User/en#Tuner_Studio That section can certainly use allot more detail, but at least there are some notes about the steps for using that .ini file.

Once those are all validated, can you reproduce the issue? If so can you send me a picture of the screen that fails when you burn the data? I have this configured as noted above and have not been able to reproduce the issue. I think there is a firmware or TS config file issue, either that or I am not doing the same steps that are causing the problem.

If this corrects the problem, can you let us know it's resolved?