Miata Turbo Forum - Boost cars, acquire cats.

Miata Turbo Forum - Boost cars, acquire cats. (https://www.miataturbo.net/)
-   MS Labs - Miata Accessories (https://www.miataturbo.net/ms-labs-miata-accessories-89/)
-   -   Custom firmware modifications (https://www.miataturbo.net/ms-labs-miata-accessories-89/custom-firmware-modifications-55596/)

Reverant 02-13-2011 02:50 PM

Custom firmware modifications
 
After driving my car more than a full year on MS2, there are several issues that I wanted to correct, since I am a programmer and I have the source code, I fixed them myself. :)

Issue number one: You are driving and coast down. As you approach the idle region, the MS enters closed loop idle, and if you actually don't put the tranny in neutral or press the clutch in, the MS will try to maintain your idle speed, thus keeping the car moving and increasing the idle valve duty cycle to really high numbers...put it into neutral and you find yourself idling at 2500rpm or more.

So I connected the clutch & neutral switch to the PE1 input (switch tables) and changed the firmware to prevent the MS from entering closed loop idle if you are not in neutral and not pressing the clutch.

Watch for the "TRANS ENG" indicator. When it goes green, it means that either the tranny is in neutral or the clutch is pressed, thus allowing the code to enter closed loop idle.



Issue number two: When you heatsoak like I do in city traffic, you need to apply non-linear IAT corrections in low load/low rpm areas. Unfortunately the stock MS2 3.1.0 firmware applies this correction under all circumstances, whereas the older MS1 firmware did provide a way to taper down as the RPMs rise. So once again, I modified the code to do this.

Watch the lower right "Gair" gauge. As I increase the RPMs, the Gair is reduced to normal (ideal gas law) Gair value.



I've been testing the first modification for quite some time and it works really well, I did a first test of the second just today and though it does work, the non-linear corrections still need to be there beyond my hardcoded 2100rpm...I'll try 2500rpm and see where that takes me.

18psi 02-13-2011 02:55 PM

OMG I can't wait:)
You are truly awesome Dimitris.
If you want to test anything on any US cars I'll be your guinea pig

richyvrlimited 02-13-2011 04:37 PM

You deserve a bowdown Reverant. I would love to be intelligent to 'do' things like that.

Will you be making the code available for us numptys to use?

Braineack 02-13-2011 04:44 PM

I never had #1 issue after I simply lowered the RPM in which PID is active.

#2 is cool.

Reverant 02-13-2011 04:51 PM

Code will be made available once its tested out ok. #1 requires hardware mods (connect clutch and neutral switch to PE1), #2 does not. I had to delete some features (MAF, N2O, EAE) to make room for #2 as the processor flash memory is full.

Reverant 02-18-2011 08:03 AM

More custom firmware awesomeness...

Different idle speeds if you have the A/C engaged! So you can idle at 800rpm without the A/C on, switch on the A/C and it will go to 1000rpm. It works on the bench, I need to test it on the car.

This is getting addictive!

richyvrlimited 02-19-2011 06:29 PM

Good we need a clever bastard like yourself to keep the MSII relevant.

The fact it's semi MX5 based mods is all the better

Asx 02-28-2011 07:11 PM

Any progress? I'm extremely interested in the Gair compensation. This is /the/ reason I regret not getting the 12 tooth crank trigger and running a MSI

Reverant 03-01-2011 03:18 AM

It works all right. I plan on removing anything that seems to be out of place for a miata (several trigger wheels, nitrous, MAF) to make room for more features.

richyvrlimited 03-01-2011 03:40 AM

Ahhh I (seriously) daydearm about being able to do stuff like that.

Keep up the great work Reverant :)

skidude 03-01-2011 11:51 AM

That's really cool, I'm just jealous I'll never be able to use it with MS3.

Zaphod 03-01-2011 01:37 PM

I am really interested to see and use this with my DIYPNP...

musanovic 03-01-2011 01:50 PM

:bowdown::bowdown:
Way to go D.

Reverant 03-01-2011 05:00 PM

A little patience goes a long way. I want to give these to a few locals here to test them out as well first...it will be a lot easier to debug locally than over the interwebs.

18psi 03-01-2011 05:43 PM

Any ETA? (not being pushy, just curious:D)

Reverant 03-01-2011 06:01 PM

Said two first mods are already done. These two (clutch/neutral and IAT corr) are already on my car.

I *LOVE* the clutch/neutral modification. I can let the car in gear in slooowly moving traffic in 2nd/3rd gear and it moves without touch the throttle, and at the same time the CL Idle doesn't activate which means you won't find yourself with a 2k idle once you go in neutral. Also if you frequentuly find yourself with a 1500/1800rpm idle without knowing why...this will probably fix it. I used to have this issue, and I've NEVER had it eversince I did this mod. Hardware mod is required but its easy.

I really haven't stress tested the IAT corrections as I can't find a suitable place to WOT test it during the day where I move to the shop and back (roads are busy). The IAT correction value may need to be readjusted so I would like to be able to provide a suitable value.

I still haven't modified my own ECU for the A/C system (yes, a hardware modification is required). I will once a find some free time (lol), but it does work (I tested it on a spare ECU I had laying around). I want to completely change the way the current A/C scheme works in light of this.

ampz 03-06-2011 05:51 PM

These 3 functions are the ones IMO that will make an MS not just better but great in the miata.

The car drives good as it is, but the a/c and stop/go cruise do remind you that it is a modified ecu. Tuning gets close to ironing it out, but you can still tell.


Originally Posted by Reverant (Post 695924)
Said two first mods are already done. These two (clutch/neutral and IAT corr) are already on my car.

I *LOVE* the clutch/neutral modification. I can let the car in gear in slooowly moving traffic in 2nd/3rd gear and it moves without touch the throttle, and at the same time the CL Idle doesn't activate which means you won't find yourself with a 2k idle once you go in neutral. Also if you frequentuly find yourself with a 1500/1800rpm idle without knowing why...this will probably fix it. I used to have this issue, and I've NEVER had it eversince I did this mod. Hardware mod is required but its easy.

I really haven't stress tested the IAT corrections as I can't find a suitable place to WOT test it during the day where I move to the shop and back (roads are busy). The IAT correction value may need to be readjusted so I would like to be able to provide a suitable value.

I still haven't modified my own ECU for the A/C system (yes, a hardware modification is required). I will once a find some free time (lol), but it does work (I tested it on a spare ECU I had laying around). I want to completely change the way the current A/C scheme works in light of this.


Reverant 03-07-2011 05:37 PM

I did the A/C mod on my own ECU today, after swapping the MS3. Watch how the "target RPM" gauge on the lower left changes from 800rpm to 1000rpm when I press the A/C button.


Braineack 03-07-2011 05:48 PM

I want this and the AIT corrections mod. You have this in a MS3 firmware package?

Reverant 03-07-2011 05:53 PM

MS3? Aaaah, the great mystery that is the MS3. Well, you see, the MS3 firmware isn't open source yet....I begged for the source on the msextra forum, so that I could add a second "thermo" fan option, but alas, I felt like shouting in the desert.

richyvrlimited 03-07-2011 05:58 PM

Yeah the lack of response from Ken and James on your thread was rather telling, such a shame :(

Braineack 03-07-2011 06:10 PM

booo. Well maybe you can share Ken the code and have him implement it? I'd really like the AIT corrections RPM reduction and the A/C idle target. I asked for the A/C idle target before the code was implemented, but they never implemented it. I also wanted a second "learned" DC% with A/C when using closed loop idle.

Reverant 03-07-2011 06:15 PM

It should be dead simple on the MS3. "If target RPM<1000 when A/C requested, set target RPM=1000". The MS3 knows what the A/C is, the MS2 doesn't, which is why it needs a bit more work. I'm sure Ken can do it in less than 5 minutes for the MS3, he doesn't need my code.

y8s 03-07-2011 10:09 PM

And he probably will do it for 1.1... or he will get a lot of chat messages from me to do it in 1.11.

Reverant 08-22-2011 09:06 AM

More firmware mods...exploring the stock 90-93 ECU has opened a lot potential!

PWM idle duty correction Vs system voltage. Switching on your fans, headlights, power windows, heated rear window and your RPM dips? Notice how the "Idle PWM %" gauge changes when "Battery Voltage" gauge changes.



And something "a few" people might be interested in...A/C idle up!!! I've assigned an input on the MS2 to read the A/C button. When it sees that input going low, it changes the target RPM to 1000rpm (if actually less than 1000rpm of course) and INSTANTLY adds 12% duty cycle to the idle valve.



Can't wait to try it on the car...sadly the gf's car battery died this morning, so I swapped in the Miata's battery so she could go to work.

Braineack 08-22-2011 09:13 AM

funny. i add exactly 12% on my ms3x. can't remember the delay off hand, but it's the perfect number.


the idle voltage correction would be cool to have.

Techsalvager 08-22-2011 09:23 AM

in order to implement this stuff, what stuff has been removed?

richyvrlimited 08-22-2011 09:29 AM

EAE IIRC.

Great job though Reverant!

Reverant 08-22-2011 09:29 AM


Originally Posted by Techsalvager (Post 762177)
in order to implement this stuff, what stuff has been removed?

Lots. MAF, nitrous, all non-miata wheel patterns, EAE.

Greg G 08-22-2011 09:44 AM

Oooooooh. Nice. Perhaps you could include the antistall feature from the mario code? I've found it's pretty useful..and has taken such a load off my mind knowing that it's there :)

This will be available for the DIYPNP guys right? :)

Reverant 08-22-2011 10:08 AM

There is no source code for the last few mario_b firmware modifications...

gslender 11-02-2011 12:27 AM


Originally Posted by Reverant (Post 689365)
Code will be made available once its tested out ok. #1 requires hardware mods (connect clutch and neutral switch to PE1), #2 does not. I had to delete some features (MAF, N2O, EAE) to make room for #2 as the processor flash memory is full.

Unclear why stuff had to be removed... So far I've not seen any need to delete anything to make things fit, so what am I missing here?

G

yunvmyegt 11-13-2011 06:20 PM

what would removing everything we dont use do aside from make room? would it just burn faster for the most part, not make it able to actually run faster? im not sure how laggy it is in the first place but i thaught id ask atleast. i just notice ms2 when i hit burn when running makes a tiny hicup, where ms1 never did, like when stopping tuning in TS. maybe its just that im still on a serial cable who knows..lol...

richyvrlimited 11-14-2011 03:38 AM


Originally Posted by yunvmyegt (Post 795140)
what would removing everything we dont use do aside from make room? would it just burn faster for the most part, not make it able to actually run faster? im not sure how laggy it is in the first place but i thaught id ask atleast. i just notice ms2 when i hit burn when running makes a tiny hicup, where ms1 never did, like when stopping tuning in TS. maybe its just that im still on a serial cable who knows..lol...

It'd hiccup on MS1 too, you just didn't notice/remember.

Only MS3 'fixes' the hiccup.

Reverant 11-14-2011 04:05 AM

The hiccup is inevitable due to the way the flash memory is deleted/written on the HC9S12. You can't just change 1 byte or 2 of the flash memory, you have to first delete and then write an entire page of memory - data which is used while the ECU is running, hence the hickup. This limitation is not there with the MS3.


All times are GMT -4. The time now is 11:13 PM.


© 2024 MH Sub I, LLC dba Internet Brands