Firmware developent status

[AndreyB]

New feature: configuration is reset to default state if pin PD6 is ground. This is useful if you've managed to save an invalid configuration, this jumper serves as "chip erase + reflash"

This begins a week of usability and error handling

[AndreyB]

Working on TunerStudio support for Simulator - http://rusefi.com/wiki/index.php?title=Manual:Software:Simulator

[Nobody]

Working on TunerStudio support for Simulator - http://rusefi.com/wiki/index.php?title=Manual:Software:Simulator

If you can add arbitrary waveforms... I have 2 hardware dual channel units, its great for throwing trash at software and finding limits that you would likely not run into in real life.

[AndreyB]

If you can add arbitrary waveforms... I have 2 hardware dual channel units, its great for throwing trash at software and finding limits that you would likely not run into in real life.

Please elaborate on this idea. Arbitrary waveforms processing as trigger signal or generating as fake trigger signal? The windows simulator does not have any real I/O and it's only testing your windows box performance.

[AndreyB]

A little bit of usability: gauges are now detached on double-click, you can also control fake/mock voltage value if you are connecting to the firmware simulator

I know there are some java developers here, wonder if anyone would be interested in getting more involved with the console

[AndreyB]

Dev console can now locate & launch simulator with just a button click - see http://rusefi.com/wiki/index.php?title=Manual:Software:Simulator

[AndreyB]

PS: does anyone want to make a screencast of downloading http://rusefi.com/build_server/rusefi_simulator.zip - starting simulator - launching TCP forwarding - connecting Tuner Studio?

[AndreyB]

@ reported an issue: Tuner Studio 'Burn' freezes the whole firmware - see https://www.youtube.com/watch?v=bLf5ymacpig see https://sourceforge.net/p/rusefi/tickets/77/

Looks like it is related to writing into internal chip flash memory, it just freezes the whole firmware - including the blinking status light - for a second or two. Is there anything in the low-level flash nature which might be causing this by design? All I am doing is writing ~8Kb.

See https://svn.code.sf.net/p/rusefi/code/trunk/firmware/controllers/flash_main.cpp
See https://svn.code.sf.net/p/rusefi/code/trunk/firmware/hw_layer/flash.c

[AndreyB]

Just a little bit of documentation: http://rusefi.com/wiki/index.php?title=Manual:Software:Trigger

[AndreyB]

Usability improvement: dev console shows which physical pin is connected to spark and injection outputs if you leave the mouse over one of the digital charts.

[AndreyB]

Usability improvement: better trigger signal error detection, current version would blink the orange LED if the trigger sensors are not providing the expected signals.

[AndreyB]

Usability: new I/O testing commands, see http://rusefi.com/wiki/index.php?title=Manual:Software:dev_console_commands#I.2FO_testing

[AndreyB]

Working on OEM ecu stimulation usability, this has resulted in a bunch of new commands:
enable self_stimulation
disable self_stimulation
set_trigger_input_pin INDEX PIN
set_logic_input_pin 0 INDEX PIN

So for example you can do

enable self_stimulation
set_trigger_input_pin 0 none
set_trigger_input_pin 1 none
set_logic_input_pin 0 pa5
writeconfig

This is getting us closer to having four logic sniffing pins for OEM stimulation mode.

[AndreyB]

To be honest I am on the edge of my automotive skills - I am struggling to get the test mule running on stock ECU. While I am waiting for more tools I am working on more gauges for TS and LCD screen support:



I desperately need car people with rusEfi hardware

[AndreyB]

dev console now has some keyboard shortcuts & text attributes in message area - that should make it a bit more user-friendly. Thank you @ for both ideas!

[AndreyB]

A critical issue was just reported: https://sourceforge.net/p/rusefi/tickets/96/ Please do not run any real engines before this gets fixed!

[AndreyB]

#96 is fixed.
Here is how one would be able to see the watchdog in action:

Code: Select all

set_engine_type 21
rpm 200
set_global_trigger_offset_angle 60
stop_stimulator_at_index 2

stop_stimulator_at_index X is the new command which stops trigger simulation after X index. A command to resume the simulation is resume_stimulator

Every 50 milliseconds a method is invoked which checks if there were no trigger events within the last 250 ms. If that's the case, all injectors and coils are turned off.

[AndreyB]

Minor usability changes based on @'s feedback.

[AndreyB]

I am going to be in Russia for the next two weeks, so I should be in bed about 3pm CST and back online around midnight in the US.

[AndreyB]

fuel pump relay is now implemented as a hard-coded flexible logic general purpose output:

Code: Select all

#define FUEL_PUMP_LOGIC "time_since_boot 4 < rpm 0 > OR"

This gets us closer to being able to specify expressions like that one for user-defined outputs, like variable timing or whatever.

[3400tZ]

Awesome man!

What's the reason for the expression to not be in order like "time_since_boot < 4 OR rpm > 0". I'm sure there is a good reason, I'm just curious

[AndreyB]

What's the reason for the expression to not be in order like "time_since_boot < 4 OR rpm > 0".

RPN is super-easy thus efficient when it's time to evaluate the expression. there are always no brackets and you do not have to look forward, just a one-pass algorithm taking and putting stuff on stack. I will create an online converter from RPN to "normal" converter to save everyone from sanity

[AndreyB]

I hope I've just implemented #52 - now you can push throttle pedal all the way down during cranking to stop fuel from coming into the cylinders - this case you hopefully might clean the excessive fuel out of the cylinders.

This feature is ON by default, you can guess what enable cylinder_cleanup and disable cylinder_cleanup commands do.

[kb1gtt]

Wish list item, if 0RPM, and TPS cycled, inject a squirt of fuel. This simulates the pumping feature that used to exist on carburetor vehicles.

[AndreyB]

if 0RPM, and TPS cycled, inject a squirt of fuel. This simulates the pumping feature that used to exist on carburetor vehicles.

This goes on top of the TODO list - this should potentially help @ and @

[kb1gtt]

How hard to make this like this, 0RPM, TPS has gone 70% or more in less than .25 seconds, then prime pulse. Such that you can slowly push the TPS and get into anti flood with out delivering an additional prime pulse.

Also once you have your engine properly tuned, you don't need these features, so you probably want to be able to turn this feature on and off via software. I have grown accustom to cycling the key on and off to get fuel injected systems to deliver an additional priming pulse. Simulating the carburetor is more for the carb people who are hesitant to go fuel injected.

Oh also this can be handy people who like to turn the key off and pump the gas, then turn the key back on to get a back fire when you turn the key back on. I've cleared out my exhaust more than once doing this.

[AndreyB]

Finally first engine in Russia:

#8 Citroen Berlingo http://rusefi.com/forum/viewtopic.php?f=15&t=540

[video][/video]

[AndreyB]

#101 - timing advance: flip sign is ready. Now positive advance value means 'before TDC' as it should be

I am mostly focused on getting Neon ready to fire up on rusEfi these days.

[AndreyB]

offtopic: that's me driving this white Rover

@12:00 you can see a turbo brick leaving the track to be crasher by a very picturesque stone

[video][/video]

[AndreyB]

Made some progress in cranking fuel logic department - cranking fuel is now controlled by two curves: coolant temperature correction and cranking_duration correction - more details @ http://rusefi.com/wiki/index.php?title=Manual:Software:Fuel_Control