Neon/420A Wheel Decoder in MS2Extra not working for me :(

mpdsn · Post by **mpdsn** » Sat Jun 02, 2007 7:27 pm

Has anyone else run into this problem? The ignition output seems to be all over the place. One of my VB921 has overheated due to high DC values.

My question is, is anyone currently successfully using MS2Extra on a car with the 420A/Neon wheel decoder? It seems that fuel calculations are fine.

I will be looking at the code to see if I can find out why I'm having an issue. It also appears that spark for one set of the wasted spark coils is timed differently than the other set.

Any suggestions would be appreciated.

UnaClocker · Post by **UnaClocker** » Sat Jun 02, 2007 7:30 pm

I have about 1,000 miles on my setup. Works great. Most important tip I can give you is DROP THE DWELL!! I had mine at 3.3 or so, like I had in MS1E.. Dern near cooked my VB's.. Turned out, 1.4 is all I needed. Yeah, 1.4 sound rediculously low, but it's plenty. Did you have your setup working with MS1E before upgrading, or are you troubleshooting a whole new setup?

md95 · Post by **md95** » Sat Jun 02, 2007 7:37 pm

MS2 Extra references 14.0 volts for dwell, which is why you need to lower your values, OR, use your MS1 Extra running dwell time, then use the dwell adjustment/battery voltage table to compensate. Use no correction (100%) at 14 volts, then more correction below 14, then less above 14.

mpdsn · Post by **mpdsn** » Sat Jun 02, 2007 8:25 pm

Thanks for the suggestions. I have tried dropping the dwell time to 1.4ms. Seems like it helped keep the spark timing a little closer at low RPMs on the scope and computer generated wheel pattern.

However, once I bumped the RPMs to about 5000 RPM the spark timing would jump between 9 deg BTDC (which is what I have it set to for testing )and somewhere around 24 deg BTDC.

I thought maybe it was my scope, but the MS1 works perfectly fine. Proper dwell and spark location.

It appears to me that the MS2Extra code seems to be selecting the wrong spark and dwell trigger teeth. The selected time may or may not be right, but I haven't went that far into it yet.

UnaClocker · Post by **UnaClocker** » Sat Jun 02, 2007 9:39 pm

So your doing this testing with a simulated output? Not the real thing? The MS1 reads the wheel much differently than the MS2 does, might give it a whirl on the real thing first.

mpdsn · Post by **mpdsn** » Sun Jun 03, 2007 3:27 pm

I think I found the problem. Thanks for all of the suggestions. I have revised the code and tested it on the scope. It appears to be working very much like the MS1. I also tested it on the car. No more VB921s to add to the RIP pile.

Post by **muythaibxr** » Sun Jun 03, 2007 3:32 pm

What did you change?

mpdsn · Post by **mpdsn** » Sun Jun 03, 2007 4:02 pm

The following changes were made to ms2_extra_ign_wheel.c:

Code: Select all

/* ----------------------  NEON MODE ------------------------*/
    if ((flash4.spk_mode & 0x1f) == 5) {
        //hardcoded for the moment, may want to store in table instead
        no_teeth = 16;
        last_tooth = no_teeth;
        no_triggers = 2;
/* Old wheel deg_per_tooth settings
        deg_per_tooth[0] =  600;
        deg_per_tooth[1] =  150;
        deg_per_tooth[2] =   50;
        deg_per_tooth[3] =  150;
        deg_per_tooth[4] =   50;
        deg_per_tooth[5] =  150;
        deg_per_tooth[6] =   50;
        deg_per_tooth[7] = 1150;
        deg_per_tooth[8] =   50;
        deg_per_tooth[9] =  150;
        deg_per_tooth[10] =  50;
        deg_per_tooth[11] = 150;
        deg_per_tooth[12] =  50;
        deg_per_tooth[13] = 150;
        deg_per_tooth[14] =  50;
        deg_per_tooth[15] = 600;
*/
// 2007-06-03 Revised deg_per_tooth settings (www.santos-racing.com)
        deg_per_tooth[0] =  600;
        deg_per_tooth[1] =   64;
        deg_per_tooth[2] =  136;
        deg_per_tooth[3] =   64;
        deg_per_tooth[4] =  136;
        deg_per_tooth[5] =   64;
        deg_per_tooth[6] =  136;
        deg_per_tooth[7] = 1064;
        deg_per_tooth[8] =  136;
        deg_per_tooth[9] =   64;
        deg_per_tooth[10] = 136;
        deg_per_tooth[11] =  64;
        deg_per_tooth[12] = 136;
        deg_per_tooth[13] =  64;
        deg_per_tooth[14] = 136;
        deg_per_tooth[15] = 600;
// End of revision 2007-06-03

  	trigger_teeth[0] = 10;
  	trigger_teeth[1] = 2;
        trig_angs[0] = -1110 + tmp_offset; // 69 BTDC (on next event) = 111ATDC
        trig_angs[1] = -1110 + tmp_offset; // 69 BTDC (on next event) = 111ATDC
        if (flash4.no_cyl !=4) {
            conf_err = 17;
        }

I have changed the 150 degree to 136 and the 50 to 64, as well as the long pulse from 1150 to 1064. They have also been resequenced to match the output from the crank position sensor. No more wondering spark output as per the oscilloscope. It is exact from 0 to approximately 43 degrees. Above 43 BTDC it is slightly off but less than 2 degrees by the time you reach 49.

The new degree numbers were obtained from the factory service manual in case anyone is wondering. I didn't just invent them.

A compiled S19 can be found here for those that are interested in testing it on their setup: http://www.santos-racing.com/support

If you do test this code and it works for you please post feedback, as it would be nice to have these changes implemented in the official code releases.

Post by **jsmcortina** » Sun Jun 03, 2007 4:05 pm

Good info! When I started writing that code there was debate about the tooth sizes, but if that's from a factory manual then great.
We'll put your numbers in the next alpha.

James

UnaClocker · Post by **UnaClocker** » Sun Jun 03, 2007 4:36 pm

Awesome, ya know, I saw the timing wandering around on my flywheel when I did my initial testing, but someone else had indicated it didn't wander at all. So I shut my mouth and ignored it. Glad to see it's going to be improved in the future, I'm surprised it worked as well as it did with the numbers being so far off. I don't fault Ken or James at all, I remember the big hunt for those numbers you just changed. Go figure the FSM had them.. ;)

md95 · Post by **md95** » Sun Jun 03, 2007 7:13 pm

Interesting..When i used a timing light, that timing mark was the most solid thing ive ever seen. It didnt budge a millimeter. This was even running off of the spark map, not fixed timing. I did NOT T-light the other ignition output though..this was just no. 1(4) cylinder. I really haven't even driven the 420a code on MS2 Extra, most of my spare time has been dealing with the 4g63 code

Take note though that my spark map is set at idle, so there is no timing wander in degrees. The bins are all set at 10 degrees, all around the idle rpm/kpa. If your map bins aren't adjusted smoothly, then your timing mark wont be either. If you set it for fixed timing, then there shouldnt be any jitter.

UnaClocker · Post by **UnaClocker** » Sat Jun 09, 2007 9:11 am

Took a look at the source included with the 0604 Alpha, and it looks like these wheel changes made it into that one. I haven't driven my MS2E'd car for a week (sharing one MS PCB between two cars at the moment), so I'll go ahead and load the latest alpha on it and have a look with the light.

UnaClocker · Post by **UnaClocker** » Sat Jun 09, 2007 10:30 am

WOAH.. What in the heck.. The car is UNDRIVEABLE with this new code. Constant misfiring and pinging. Barely even idles. Tried reburning the code, reburning the MSQ.. Then went back to the 0531 alpha and all was well again.
I checked the timing mark with my light, seems just fine. Re-checked with the 0531 alpha and it's very stable all the way up as well. My previous instability may have been due to my really bad dwell settings.
md95, any chance you can try the two codes on your 420a engine and see what you find.

Post by **jsmcortina** » Sat Jun 09, 2007 10:44 am

Two different Neon cranks ?!?! Bizarre!

EDIT: Unless some other code changes made it sick. Drop me an email jsm@jsm-net.demon.co.uk and I'll compile a version with the old pattern in there.

That is easy enough to allow for in the code.
Could you guys please strobe you timing with fixed angles from 10 to as high as you dare in 1 deg increments and check that it is a smooth response.

James

UnaClocker · Post by **UnaClocker** » Sat Jun 09, 2007 10:58 am

Ok, I'll shoot you an email. Was just thinking it'd be nice if I could get the gcc compiler to work on my machine, I'd have tried exactly that change.
My timing window only goes to 14 or maybe 16 degrees, so I can check that range. Just check it at idle, or up and down the rev range? I checked 10 degrees quite a ways up the RPM's.. About as high as I wanted to go with my head inside the engine bay. ;)

Post by **jsmcortina** » Sat Jun 09, 2007 11:19 am

Check as much as you can really. Because of the way the code works it is important to confirm various angles. If you bring the revs up a bit, also check the low angles down to maybe 5BTDC.

James

Post by **jsmcortina** » Sat Jun 09, 2007 12:18 pm

I sent UnaClocker a CVS snapshot with an option to choose "old" or "new" tooth pattern.

The car is still not driveable with this code, it's missing BAD..
The old Neon angles do work better for me, from the little bit of
testing I could do with the driveability issue.

Ok, so which alpha versions have you tried? I want to try to narrow things down to be sure I'm looking in the right place.
0531 = good
0601 = ??
0604 = bad (new crank angles)
0609(CVS) = bad (either crank angle)

Could you try out 0601 and confirm what I posted above it right then I can do some diffs on the code.

As a side note I was using 0604 (ish) on my LT1 last night and aside from the injector PWM issue, it seemed to idle ok.

James

UnaClocker · Post by **UnaClocker** » Sat Jun 09, 2007 12:23 pm

You're right on the alpha's. I haven't tried 0601, I'll load that one up and give it another whirl.

UnaClocker · Post by **UnaClocker** » Sat Jun 09, 2007 1:18 pm

Well great. For some reason, my car will not run correctly under any code version now. I'm done messing with it for the day, I apparently will be completely unable to debug any code, since I seem to have a hardware problem. Sorry.

Post by **jsmcortina** » Sat Jun 09, 2007 1:22 pm

That's weird. Best of luck in getting it working again. I guess you'll be stepping back to code and settings where you had it driveable.

James