MS2/extra beta 7 (20070812)

mattd860 · Post by **mattd860** » Thu Aug 16, 2007 6:29 pm

Ken, Based on what I think you said earlier, 0% timing mask fo Map Sampling is what is the previous codes are set to? So the BETA 6 code is set to 0%. Or is it 50% which is the default setting for BETA #7? I want this to be exactly as it was before because currently my car is running better than ever but have the Beta 6 installed. Thanks.

wes kiser · Post by **wes kiser** » Thu Aug 16, 2007 7:20 pm

Just loaded the code and drove. I haven't tried a new beta for the last 1.5 months (car has been down). Car runs fine, definately went at least 20% rich in most spots (I haven't tried code since the map sampling on tach was introduced). The map is much more stable, and the car feels "smoother." I will see if this holds after I lean it out though, as rich typically feels very smooth and happy. This is the first time since I have tried any code the full boost, high rpm stuff has changed, and it is went from 11.5 to 9.5:1 (good thing it didn't go the other way, as it might be new head gasket time).

Post by **muythaibxr** » Thu Aug 16, 2007 9:17 pm

I'm not surprised it went rich if that's the last time you tried any new code.

The default setting in beta 7 is 50%, the default setting in beta6 would've been 0%

According to a few people, you might have to retune anyway if you switch to beta 7, so I'm going to dig around and see if I messed something up somewhere. I did not touch anything that should cause retune to be necessary when using a 0% setting, but I should be able to see a difference like that on the bench.

However, at the moment, I'm still doing work for my "day job" (after midnight no less), and have been almost as swamped the whole week at work, so I probably won't have time to do anything testing wise until Saturday.

At that point, I'll take a look to see if I can find what's making people run 20% richer, and fix as many of the other remaining bugs as I have time to fix.

Ken

wes kiser · Post by **wes kiser** » Fri Aug 17, 2007 12:59 pm

I also noticed a weird thing in a datalog. It only happend once in a .5 hour drive.

Mapdot spiked to -7xx, and PW spike to 11.5 at the same time. There were no other inputs that moved. I would post a log, but there really isn't anything there but that (and it is semi-irriating to post the log from my laptop).

It only happend once, and never again. Just curious if anyone else has seen anything like that. I am not running map anything for accel enrichment (TPS only, especially for new code evaluation), and even so can't think of a circumstance where a negative mapdot should result in more PW than I see at 20 psi. At the time I was cruising at 55 kpa, and 2400 rpm. Steady throttle, and everything. I never felt it, as it was far too brief.

Post by **muythaibxr** » Fri Aug 17, 2007 1:17 pm

I've seen weird things like that happen, but traced them down to serial comms issues with my USB->serial converter. Sometimes you can get weird comm errors on regular serial ports too.

Ken

UnaClocker · Post by **UnaClocker** » Fri Aug 17, 2007 7:49 pm

Installed this beta. Instead of wasting gas trying to retune, I just scaled my whole VE table down 15%, and am increasing fuel where needed/as needed. I was able to raise my map lag factor quite a bit. I have the map mask set to 35% (picked that number at random). Seems ok. Did a couple full throttle sprints to see how my Beta 6 missing problem was doing. I had a couple misses, not as bad as I had with Beta 6.. I'll keep fiddling with my settings until someone else comes down with the same problem.. Might be something else, somehow. Fuel pump, or VB921's maybe..
Maybe I didn't have enough time with the correct settings for my AIS in Beta 6.. But this Beta 7 starts like a stock ECM. Fires right up, idles down promptly and just purrs.

Post by **muythaibxr** » Sat Aug 18, 2007 9:30 am

Alright, I went and compared beta6 to beta7 pulse-widths under vacuum on the bench, and at every vacuum I tested, the pulse-width was still right... Meaning under identical conditions, the pulse-width matched. The only other thing I can think of is that EGO is doing something different. That's a bit more difficult to test on the bench though.

I made some changes to that code, but they should only have been whitespace changes.

If anyone who's seeing a difference from beta 6 to beta 7 that's not affected by changing the map sample time mask can help me, I'd like you to turn off EGO by changing the "correct above" RPM to something higher than your redline, then drive around, and see how far off your beta6 VE table is on beta 7.

I want to try to determine if the problem you're seeing is just something to do with MAP sampling (which means you just have to retune) or if I somehow broke EGO when changing whitespace to make it more readable.

Ken

Post by **muythaibxr** » Sat Aug 18, 2007 9:59 am

Also, I'm going to be on vacation this entire upcoming week, so if I'm not responding on the forums, that's why. When I'm back, I should have time to fix any remaining bugs.

It's likely the only fix I'll get in today is overboost protection.

Ken

wes kiser · Post by **wes kiser** » Sat Aug 18, 2007 7:15 pm

Sorry, can't help you there. I don't use EGO for anything but data. I could enable it to see how it works (but then I wouldn't have anything useful to compare it to).

I have mine about re-tuned now. My peak boost, peak ve number dropped for a 117 to a 109. Some of the lower load portions exibited much larger percentage changes (like from 65's to 55's). It seams to run better in general with this code though.

Post by **muythaibxr** » Sat Aug 18, 2007 8:24 pm

Yeah, considering how long it's been since you updated, I'm not surprised that there's some retune for you, I was mainly asking about EGO for Ballistic.

Thanks for letting me know how it went for you though, seems like we're finally getting this code stable! I'll fix up dwell for low-tooth-count wheels, and I think this code will be solid after that!

Ballistic · Post by **Ballistic** » Sun Aug 19, 2007 4:04 am

Ken, I don't know if this is information that helps or not, but indicated AFRs were too rich as soon as the sensor warmed up and began displaying values. Instead of the usual 13.5:1, the AFR was closer to 11:1. This is with a coolant temp below EGO correction enable.

Is that sufficient, or would it also be helpful to perform the other test?

Post by **muythaibxr** » Sun Aug 19, 2007 4:18 am

that should be enough.

That said, as I said before, my bench tests showed the exact same pulse-width from beta 6 to beta 7, so as of right now I'm thinking that you're seeing a combination of factors, mainly influenced by the MAP sampling code.

Ken

wes kiser · Post by **wes kiser** » Mon Aug 20, 2007 7:23 pm

Ken, just out of curiousity, have you don't bench tests comparing pw's for different numbers of squirts?

I started out with my 2 squirt alternating tune, and it was very rich. Over the course of a couple of days I leaned it out. Then just for grins I went back to trying to tune for 4 alternating, and all of a sudden it wanted the same VE table 4 alternating used to require. I know from past experience the low load values normally change in this situation because it is opening time dominated, but the high load values changed substantially (and changed back to where they used to be with the same number of squirts).

When I tuned for 4-alternating before, the boosted regions of my VE table never needed to budge. Keep in mind I run 900cc injectors on a 2.3 liter, so this isn't necessarily and "easy" change.

This really doesn't make sense to me, just thought I would point it out.

mtlcafan79 · Post by **mtlcafan79** » Tue Aug 21, 2007 9:16 am

My truck exhibits the same behavior going from four alt to two alt.

Post by **muythaibxr** » Sun Aug 26, 2007 3:02 pm

I did bench test that, between beta 6 and beta 7, and didn't see any change there, so if something changed to cause this, it happened before beta 6.

I vaguely remember changing something that looked wrong in the ini file.

I'll have a closer look this week now that I'm back from vacation.

Ken