MS3 pre-1.3 beta11 released

[jsmcortina]

This is a development code that follows pre-1.3 alpha10.

Main changes since alpha10
Bug fixes:
Fix "stuck" fuel-cut in launch.
Don't show config error for knock if already a config error.
Apply tc_retard to ign timing.
T/C retard set to zero when off.
Add CAN MSG_SPND fix from Jean
VVT angle calc, handle when min/max not set.
VVT angle calc for wraparound and allow 5 deg more.
CAN RTC was sometimes calling can_inc_ring() in error
Apply {loadhigh} to various load gauges.
Send BCD encoded year to set RTC on Extender.
VSS2 uses correct bits to select sensor position.

Enhancements:
Change Subaru 6/7 re-sync method during cranking.
Finish C code part of Generic PID (PWM only to start). Still working on ini.
Generic closed-loop control feature.
Add min RPM for AC idleup activation. Allow it to activate even on crank.
GPS reading from IO-extender.
Add rev2 SDcard datalog mode for interleaved GPS data with 64 byte.
Add min/max limits to SpkAdj. (untested)
Add dwell table.
CAN polling load balancing
Allow sig-PIP to run COP (untested)
Add engine shutdown feature. Send 0x8501 to table 7 offset 683
Crank pulsewidth taper can be in events.
Add master CAN poll/broadcast selector.

Minor/internal
Tweak to twin-trigger cam speed 2 cyl. (Should be no functional change.)
Remove RTC_TIMING, SDREAD_TIMING
Remove some debug lines.

The two main additions here are the Generic closed-loop control and GPS capture from IO-x. Both will likely need some refinement. The format of the GPS data is likely to change slightly in the next release, but that's pending a firmware update from Jean.

The firmware zip is available from:
http://www.msextra.com/doc/ms3/files/be ... beta11.zip

James

[jsmcortina]

Another note on GPS. Presently this supports the JBperf IO-x GPS only. That has had GPS since 2010 and users have been asking for it to be supported since then. The GPS data isn't used within the MS3 at this time. It can be recorded to a regular serial datalog along with the other data. There's also a new "interleaved" SDcard log format that will log GPS frames at their slow rate while still recording normal data at a fast rate. As this was only added last week, TunerStudio doesn't yet support extracting it.

Just like other GPS data that is within a datalog, the MegaLogViewer scatterplots can display them.

ms3-gps-route.png

Or, the datalog can be edited in a spreadsheet to give just the GPS data fields and then rendered in GPS map overlay software.

James

[nismoautoxr]

I started getting this with alpha10 and still get it with beta11. I get a differance report each time I connect with TS. Heres the screenshot. Not sure what it is .The first time I got it I used the controller settings , the next time I got it I used TS settings.

[jsmcortina]

Please post your TunerStudioAppDebug.txt file, that might shed some light on this.

I disable Difference Reports and CurrentTune, so I won't see that.

James

[nathaninwa]

Thanks James for this release. Im doing some tinkering on the car today and will load this midmorning or so.

[nismoautoxr]

here is the debug file James.

[nuvolarossa]

tried the demo msq.
When enabling dwell table, it shows as y axis "ignload INVALID".
It's just a label issue or it will affect table operation too?

[nathaninwa]

Enabled coil dwell control, and drove around town for about hour today, no issues. Thanks James.

[nismoautoxr]

James ...I installed the TS beta update today that Phil put up and this afternoon when I got in the car and connected with TS, it did not give me a differance report . I will hook up to it when I drive it tomorrow and see if I get a diff report and let you know .

[nismoautoxr]

well I really dont know how to explain it but since I updated ts to 2.5.07 I am no longer getting the differance report so guess it doesnt matter .

[jsmcortina]

tried the demo msq.
When enabling dwell table, it shows as y axis "ignload INVALID".
It's just a label issue or it will affect table operation too?

I think that's just a label issue - investigating.

James

[nismoautoxr]

James,
Did you make any changes to spark cut on the tractin control reactions between alpha8 and beta 11? Reason I ask is .. in alpha 8 it was very on or off...even with 1% cut it was like a total loss of power, with beta 11 it is very smooth and reacts more like I would expect .

[jsmcortina]

James,
Did you make any changes to spark cut on the tractin control reactions between alpha8 and beta 11? Reason I ask is .. in alpha 8 it was very on or off...even with 1% cut it was like a total loss of power, with beta 11 it is very smooth and reacts more like I would expect .

A quick skim through the ChangeLog didn't remind me of a specific change, but I did re-organise that code a few days ago to clarify the logical path and fix the "stuck" tc_retard.

James

[hardline]

Glad you got the VVT operating window fixed. Started the car on Beta11 and knew immediately that it works properly. With my settings with Alpha9, where the window didn't resize I had my cam timing set in absolute values. Upon going to the operating window, the VVT went full retard on the exhaust cam. Sounded like a stuck open EGR Valve. Glad I was expecting it or it might have thrown me for a loop.

[nismoautoxr]

James,
Did you make any changes to spark cut on the tractin control reactions between alpha8 and beta 11? Reason I ask is .. in alpha 8 it was very on or off...even with 1% cut it was like a total loss of power, with beta 11 it is very smooth and reacts more like I would expect .

A quick skim through the ChangeLog didn't remind me of a specific change, but I did re-organise that code a few days ago to clarify the logical path and fix the "stuck" tc_retard.

well so far that is working pretty well sir.

James

[darsondriss]

I think something has messed up in the QR25 trigger code, I think it was modified a version or two back. I updated to beta 11 earlier today and was unable to get the car to start. I reverted to alpha 9 with the same tune and it fires right up. I repeated the process with the same results. Ive attached a msq and datalog. If i had to guess I would say its not getting spark, I can smell fuel.

[jsmcortina]

I think something has messed up in the QR25 trigger code, I think it was modified a version or two back. I updated to beta 11 earlier today and was unable to get the car to start. I reverted to alpha 9 with the same tune and it fires right up. I repeated the process with the same results. Ive attached a msq and datalog. If i had to guess I would say its not getting spark, I can smell fuel.

Yes, I did change it, from the reference data it _looked_ as though QR25 was misaligned.

Can you confirm that spkA is wired to the coilpack on cylinder 1 ?

James

[elaw]

I just upgraded from alpha 8 to beta 11 and got the "neverending difference report" thing too, with a dialog that looks *almost* exactly like what nismoautoxr posted.

Here's what I realized: if I clicked "send tunerstudio settings" I was seemingly stuck in a loop forever. If I clicked "use controller settings", the two values it was upset about would be divided by 10 each time.

This I think is from the second time through:

Beta11problem.png

And this is from the third time:

Beta11problem2.png

I suppose it's possible that if I did it enough times that the value reached zero the problem would stop. But I stopped it by going into "generic closed-loop a" and "generic closed-loop b" and turning them on (they started out off) then off again.

Edit: this is with TS 2.5.07, and my first beta 11 .msq (before the problem went away) is attached.

[nismoautoxr]

Odd stuff there. I also came from a variant of alpha8 but I started getting the report with 10 but I was only on 10 for a couple hours before flashing in what was then test code for beta 11 so its looking like its something from 8 to post alpha 10 . I dont get it anymore . It just sortof automagically disappeared but I also did not take note to the actual figures that were in the yellow boxes nor did I take note of what they were the next diff report.

[elaw]

Hey just something really trivial I noticed... in the dwell table, the values can go to zero. But on the ignition options screen, the values (both cranking and running) won't go below 1.0.

Personally I'd vote for 0.0 or maybe 0.1 as the minimum - you never know what weird coil might show up that requires very little dwell. I usually run mine at 1.8, and experimenting with the dwell table was able to go as low as 0.3 before the running of the engine seemed to deteriorate.