muythaibxr's Xanga

Monday, August 06, 2007

Mon, 06 Aug 2007 17:37:50 GMT

Wow, been a while since I've posted here...

Julian and I are about to take down the Corolla for the final "build."

That means we'll be pulling the engine, and replacing the coolant line system, and going over all the external seals (rear main, front main, etc..) and checking to make sure the work we did before installing the engine the last time is still good. We'll also be replacing the flywheel and pressure plate bolts.

While the engine is out, we're going to tear out a lot of the interior, and pull out as much of the wiring as we can, then take the shell to a body shop, where hopefully they'll be able to fix it all up like new, and weld in a rollcage.

Once we get the shell back, we'll redo all the wiring (I'm going to set it up so I can easily switch between ms2 running ms2/extra, and ms3, as we'll be developing that soon hopefully).

Then finally we'll be able to race that car!

My RX7 is going to get some fixes soon too. I need to figure out where its oil leak is coming from, and fix a couple vacuum leaks, but other than that it starts up and runs alright.

Thursday, May 03, 2007

Thu, 03 May 2007 01:16:08 GMT

Probably the longest I've ever sat in front of a camera:

http://video.google.com/videoplay?docid=-3203891044384052237&hl=en

Monday, February 19, 2007

Mon, 19 Feb 2007 02:05:46 GMT

I added the load code from my last post, but have not tested yet... I'll probably do that tomorrow... Then I'll do the filtering code, and clean up a lot of the messes that are still in the code.

James and I spent the last month mainly doing bug fixes.

I also branched the code so that I could work on making a stable version to turn into MS2/extra v1.0, while James continues ripping apart code and rewriting it for v2.0.

I had my first tester other than myself try out my EAE code, and he said he got it working in a few minutes as well as his previous MAPdot accel setup that took him weeks to get to a usable state. In other words, I think we might finally have an Accel Enrichment algorithm that "just works."

I'll be testing the new load code in the ae86 as soon as I get it bench-tested, I'm thinking that once I get it working well, it'll make the ITB setup a lot more tunable... meaning I think I'll be able to get the benefits of Speed Density (good fuel economy, good low-load behavior) and Alpha-N (good throttle response, better tunability above 20% throttle position) at the same time.

That should be good for people with turbos AND really lumpy cams... they'll be able to run alpha-n but use the secondary table to cover conditions under boost..

similarly, I could run alpha-n as my primary load on the ITBs, but have a second that modifies the first depending on vacuum, OR run Speed Density as my primary, and then when that maxes out around 20% throttle, or around 90kPa, just have the secondary table more or less take over. I'm going to try both and see which one works the best.

I should probably make the EAE load point changeable too... right now it's tied to the primary fuel point... which might not be desirable.

Anyway, enough babbling...

Wednesday, January 24, 2007

Wed, 24 Jan 2007 04:19:02 GMT

Finished up the first cut at rotary trailing ignition code for ms2/extra on Sunday.

Next are helping James with distributor mode, new load calculation and representation in the tuning software, along with the addition of modifier tables (one table modifies another based on a second load calculation which is different from the first), and adding noise filtering code to the Input Capture ISR.

Should make ms2/extra a lot better!

Friday, December 08, 2006

Fri, 08 Dec 2006 23:07:23 GMT

Pics of my new workbench setup:

And another pic:

Pretty cool huh?

Monday, November 27, 2006

Mon, 27 Nov 2006 03:51:50 GMT

Not much news... I implemented staged injection, so it should be possible now to run a stock RX7, although without trailing ignition. I'll be implementing that after my move; it shouldn't be that difficult, but I need to concentrate on the move for this week.

I'll probably put together a Megasquirt with MS2 to test with the rx7 fairly soon as well.

I also need to make a couple more modifications to my EAE code. IT's responding as quickly as possible now, but I need to make one more slight modification to make it respond more aggressively to increase intermediate response a bit.

Anyway, just a quick update.

Monday, November 13, 2006

Mon, 13 Nov 2006 21:50:40 GMT

A new alpha release of ms2/extra will be coming soon.

I've done a lot of work in this one to refine fuel control:

1) calculate fuel in .01 microsecond units (instead of original code which was in microseconds). The ms2 site currently says the ms2 can do microsecond resolution, but because all fuel calcs were being done in microseconds, and with no floating point math, a lot of rounding error occurs during the calculations, resulting in actual pulse-width control of around 200-400 microseconds. With the change over to calculating fuel in .01 microsecond units, the rounding error happens in the .1 and .01 microsecond places, and then gets divided out when converted to microseconds for the actual pulse-widths to control the hardware at. The result is that the engine runs much smoother, and various enrichments occur in much finer increments than before.

2) Removed X-tau code and replaced it with what should be a more robust, tunable, and easier to tune wall-wetting model. Basically it's mostly the same math as x-tau (more simple), with more simple tuning parameters, and it works better to boot (probably because it's easier to understand the tuning parameters). This is the code that ensures when you move your foot, the right amount of extra fuel is injected (or not) to maintain a good AFR and good response.

James made some good changes as well, such as converting most of the ISR code to assembly from C (code is smaller and faster) and adding overrun fuel cut. Just the conversion of ISRs to asm alone is a huge task.

Anyway, just wanted to update here for people who don't regularly check the forums.

Monday, October 09, 2006

Mon, 09 Oct 2006 01:52:51 GMT

Julian and I took the ms2/extra code on an extended test-run today. Once I got the PID/Smith Predictor EGO settings worked out, and got the VE table close to tuned, the car ran great! So this car is running wasted-spark ignition using the every-tooth wasted-spark (or COP depending on your settings) wheel decoder that I wrote.

The latest alpha, and any general info you want can be found at the new MSx/extra forum:

http://www.msextra.com

Monday, September 11, 2006

Mon, 11 Sep 2006 20:26:59 GMT

I've seen a lot of people googling for "MS2/Extra" and coming up with this page...

While James (jsmcortina on msefi) and I are writing the ms2/extra code (I wrote the wasted-spark/COP ignition stuff, and am still adding to it, and will be writing other fuel/spark related features), I don't normally post detailed information here about it.

For more detailed info, and links to the latest alpha/beta code, go here:

http://www.msefi.com/viewforum.php?f=91

I did get some more testing done last night, and so far everything's working great.

Sunday, August 27, 2006

Sun, 27 Aug 2006 17:22:14 GMT

The ms2/extra code that I wrote the ignition code for started the car! My code works! I even got it to rev really quick and got it to be really responsive. I'll probably take it out for a spin sometime this evening to start tuning.