MS2/Extra 2.0 beta 18 (20080105)

Post by **muythaibxr** » Sat Jan 05, 2008 11:07 am

Changed/fixed in this release are:

- Several .ini fixes related to Accel, EGO, Table Switching, etc...
- part of above was to add TPS and MAP accel and decel indicators to the bottom of main MT window.
- TpsThresh and MapThresh for normal AE are now ints... their max's are 200%/sec and 1000 kPa/sec respectively.
- All C files were re-indented using gvim's autoindent feature.
- Fix display of load for dual table
- Change normal AE so that MAPdot threshold can trigger decel, and make all indicators for normal AE work properly. Make code in general more aware of MAPdot vs TPSdot accel

If I missed anything that James did, I'm sure he'll post that here as well.

The beta is located at:

http://www.msextra.com/ms2extra/files/b ... 080105.zip

Ken

lammbn · Post by **lammbn** » Sun Jan 06, 2008 11:49 pm

alright now, i hope this isn't a stupid question, but after two hours of searching i have to ask. i am getting ready to buy my megasqurit unit right now, money in hand and burning a hole in my pocket, but i want to make sure i have everything straight before i do. can never be too prepared ya know!

does MS2-Extra code run the dual tables or do i need this beta version to do so? if so why do all the beta versions have updates/fixes to this feature? this is a very critical part of what i want to do with my system, as well as other MS2 exclusive features. i need this to run fuel on one bank and octane boost (ie. toulene) on the other at high load/rpm.

also, just a thought, i cant seem to find anywhere a definitive list of all the differences, updates, and/or fixes that these beta versions have in relation to the general MS2-Extra code or eachother. is this right under my nose or does it just not exist? it really needs to be put somewhere more obvious if it is available.

thanks for listening to my newb questions.

460stang · Post by **460stang** » Mon Jan 07, 2008 9:44 pm

http://www.msextra.com/ms2extra/MS2-Ext ... m#features

There is your table.

MS2 extra V2.0 will probably be in Release code by the time you get your MS purchased and installed, V1.0 will do the dual table code but V2.0 does pretty much everything better.

NOTE: This is probably the wrong place for this. Starting your own topic would have been a good idea.

md95 · Post by **md95** » Fri Jan 11, 2008 7:28 pm

I loaded up Beta 18 tonight, to do some more testing with the 420a code..all went well, but it took a good 30 minutes to download the s19 to the chip? Is this normal? Or was i just having some odd issue with my computer or something.?? BTW, this was going from a brand new MS2 card with standard code on it, dont know if that makes a difference. It loaded just fine, just took forever. I only did a limited "garage" test, and everything seemed as normal anyway.

Ballistic · Post by **Ballistic** » Sat Jan 12, 2008 8:01 am

I loaded up B18 this week, my first testing since ~B9 or so. On trying closed loop stepper IAC I seem to have the same problem as described here.

Asteris wrote:
Small remark. When Idle set at Closed loop IAC, cranking position steps are Grey. Motor goes to the value taken from PWM cranking position (you can change that if you select PWM closed-loop), but it allows only 1-100 values, where stepper motor may need higher value.

Also Idle Valve closed Duty or Steps (%), doesn't affect operation, motor moves between Idle Open Duty or Steps (%) and Min Duty or Max Steps for PID (%).
As i remarked before, no cranking IAC step available when at closed loop IAC. Had to tweek ini file, because revs were up to 2500 for a couple of seconds.

I've not looked into it in depth or tweaked the ini, but do have 2500+ RPM on startup for a few seconds before it settles down. This is my first exposure to the closed loop IAC control so I'm not certain of the settings.

Post by **muythaibxr** » Sat Jan 12, 2008 9:39 am

Yeah, closed loop with the IAC is going to be interesting anyway since those valves move so slow.

I might have to mess with the code a bit to make the algorithm more responsive on those valves, but it may not ever be as good as the PWM valves.

Ken

Ballistic · Post by **Ballistic** » Sat Jan 12, 2008 11:33 am

I don't think it's the PID control, but rather a crank value that is near wide open. RPM during cranking was higher also. I saw 65 on the IAC steps gauge before the idle slowed down. Normal warmed up running values are ~250+.

Asteris · Post by **Asteris** » Sat Jan 12, 2008 2:17 pm

PWM and IAC closed loop, share the same variable for cranking position, but there is no IAC cranking steps to set in Megatune, so do it manually. Find pwmidlecranking inside the megasquirt.ini file and set it to the steps you want.

Ken, please fix it at next beta.

Asteris · Post by **Asteris** » Sat Jan 12, 2008 2:29 pm

muythaibxr wrote:Yeah, closed loop with the IAC is going to be interesting anyway since those valves move so slow.

I might have to mess with the code a bit to make the algorithm more responsive on those valves, but it may not ever be as good as the PWM valves.

Ken

Why do you keep saying that IAC valves are slow? My valve takes less than 0.5 secs to fully retract from closed (160 steps), when MS is first powered up. And on the stock ECU, it handles fine the fan on/off change, it instantly raises the revs at 1000 from 850 as soon as the A/C button is pressed and does two or three idle corrections within a second, when needed.

With MS, when lets say, idling drops to 750, it takes 2-3 secs to try and raise it and if it overshoots to 1100, it then takes another 2-3 secs to try and lower it. Is there any delay variable that i'm missing?

Post by **muythaibxr** » Sat Jan 12, 2008 2:32 pm

Maybe your valve is faster then, the ones I've tested aren't. Either that or a couple steps on the valve makes a huge difference in idle speed.

So I take it you're actually using PID closed loop with your stepper then?

Ken

Post by **muythaibxr** » Sat Jan 12, 2008 2:35 pm

Asteris wrote:PWM and IAC closed loop, share the same variable for cranking position, but there is no IAC cranking steps to set in Megatune, so do it manually. Find pwmidlecranking inside the megasquirt.ini file and set it to the steps you want.

Ken, please fix it at next beta.

Didn't realize it was broken.

Asteris · Post by **Asteris** » Sat Jan 12, 2008 2:48 pm

muythaibxr wrote:Maybe your valve is faster then, the ones I've tested aren't. Either that or a couple steps on the valve makes a huge difference in idle speed.

So I take it you're actually using PID closed loop with your stepper then?

Ken

Yes, and today i finally managed to avoid sinking of the revs below target idle, when i let off the gas pedal. If only i could make it react faster when fan cuts in. Perhaps stock ECU senses voltage drop and acts faster.

PID at 30, 45, 5, start retracting steps at 170, 135 steps cranking, 135 open, 155 closed steps.

135 steps is 1200 rpm, 155 is 800. Each step is about 20-25 rpm. Magneti Marelli valve (GM straight 4pin style), used at Fiats, Alfa Romeo, Peugeot ans Citroens, which use Marelli ECUs . I''ll see if i can find a spare one to send you for testing, if you need one.

Post by **muythaibxr** » Sat Jan 12, 2008 4:18 pm

Hrmm, each step makes a large difference in rpm then... maybe that is acceptable then.

Ken

Ballistic · Post by **Ballistic** » Sat Jan 12, 2008 4:48 pm

Here's some additional data:

I am using an unmodified GM LS1 throttle body.

Total number of steps from fully extended and seated in the housing to fully retracted and bottomed out is right at 345. I have the retract value set to 355 to cover for any missed steps (this was more an issue with earlier code variants than with current versions).

Cranking value is 235.

From 0* to 180* my step table ranges from 200 to 260 steps.

I forget what the default step time is, but at 2-3 ms this would provide a fully extended to fully retracted time of 0.690-1.035s. I haven't timed it, but this seems reasonable based on my observations.

Minimum step size set to 3 and (based on what I recall) this makes a noticeable difference in idle speed.

Ballistic · Post by **Ballistic** » Sat Jan 12, 2008 4:52 pm

Also, I too noted the "seconds to respond" issue. It did a fair job of obtaining the correct RPM . . . eventually.

Post by **muythaibxr** » Sat Jan 12, 2008 4:58 pm

if it's taking too long to respond, increasing P might help; this is what I meant by it being slow though, and I have a few ideas that might help it speed up the response a bit with stepper motors.

Ken

Ballistic · Post by **Ballistic** » Sat Jan 12, 2008 7:54 pm

It's not that the change applied was not large enough, but that too much time elapses between corrections.

Post by **muythaibxr** » Sun Jan 13, 2008 9:17 am

Yeah, that's what I mean, it won't correct while the valve is still moving... I'll have to check to make sure that the "IAC_moving" variable isn't being left in the moving state too long, that's the only thing that would cause what you're seeing.

Ken

turbo355 · Post by **turbo355** » Tue Jan 15, 2008 12:48 am

I have a question. While using my jimstim in CAS4/1 mode i cant seem to get the Rev limiter to work at all for spark it does retard the ignition and it will cut fuel but the leds dont miss a beat while in spark cut no matter how the spark cut count is setup.

Post by **jsmcortina** » Tue Jan 15, 2008 6:42 am

turbo355 wrote: i cant seem to get the Rev limiter to work at all.

Please post your MSQ.

James