MS2/Extra 2.0 beta 16 (20071122)

mops · Post by **mops** » Mon Dec 17, 2007 3:17 pm

hassmaschine wrote:
Philip Lochner wrote:My MS2/Extra Beta 2.0 causes the car to stumble every time (possibly not ALWAYS) I do a burn.

Is that normal?
it does this with all versions of MS, as far as I know.

it's not as bad if you have less cylinders with dizzy pickup. stumble is there beacue while ecu is burning it cant count teeth on a tooth wheels, so it has to resync after burn...

Philip Lochner · Post by **Philip Lochner** » Tue Dec 18, 2007 10:42 am

hassmaschine wrote:
Philip Lochner wrote:My MS2/Extra Beta 2.0 causes the car to stumble every time (possibly not ALWAYS) I do a burn.

Is that normal?
it does this with all versions of MS, as far as I know.

The reason I ask, is that none of the B&G versions I've worked with thus far (since MS2/2.35) has done this - at least not in such a way that it can be felt even if under power. MS2/ExtraBeta2.0 is my first exposure to Extra fw and I considered that the Extra community may consider features (like RPM based IAC control) more of a priority than a smooth burn - I do

Its not a big deal in my mind, just must make a mental note to be on overrun when burning.

msoultan · Post by **msoultan** » Tue Dec 18, 2007 10:52 am

I have no idea how it compares to b&g code, but it's done it with all the extra versions. Welcome to the club!

Mike

Post by **muythaibxr** » Tue Dec 18, 2007 2:57 pm

There are a few reasons for the stumble:

1) Our paging code. We stuff more features in than there is RAM to hold everything at once, so most of the time, we read variables from flash, and only load settings into RAM when necessary, and in 1k chunks.
2) We burn a whole page at once instead of doing the "trickle" burn that the standard code does.

The standard code actually burns from the main loop a little at a time to try to make sure that it doesn't keep interrupts from running. However, on wheels with large tooth counts, at higher rpms, you can still get a burn stutter.

We figured people would put up with the stumble if it meant getting features they wanted.

Ken

UnaClocker · Post by **UnaClocker** » Tue Dec 18, 2007 5:18 pm

muythaibxr wrote:We figured people would put up with the stumble if it meant getting features they wanted.

Ken

Heck yeah. Really, once you have the car dialed in, you are hardly ever doing burns anymore. And when you get used to the slight studder, you simply learn to not hit burn till your decellerating or at least at a stable cruise.
Might be worth nothing this in the MS2Extra manual though, so people don't have to constantly ask if they did something wrong. Just copy your above post into the manual..

md95 · Post by **md95** » Wed Dec 19, 2007 2:22 pm

Question: What would happen if the MS lost sync from the wheel for a split second, then re-synced? Would it behave in the same manner as doing a data burn(stumble)?? I ask this because every once in a while, i'll get that familiar stumble when just drving, mostly on the highway after a while, and it behaves just like a data burn..The ASE light goes on real quick, then goes out(i just happened to be staring at the MT display when it happened)..I know when you do a data burn, it forces a wheel re-sync, correct? I know this is the same behavior, as when doing data burns i've noticed ASE goes active..just one other issue i've been trying to figure out since original install..

Post by **muythaibxr** » Wed Dec 19, 2007 3:10 pm

Yes, if you lose sync, rpm will briefly drop to 0, causing the ASE indicator to blink on.

Ken

md95 · Post by **md95** » Wed Dec 19, 2007 6:47 pm

muythaibxr wrote:Yes, if you lose sync, rpm will briefly drop to 0, causing the ASE indicator to blink on.

Ken

Thats what i thought...Well that's great, for some reason the 4/2 CAS is loosing sync randomly.

Post by **muythaibxr** » Thu Dec 20, 2007 8:53 am

This can be caused by many things though, most of them not code related.

Of course we'll check into it, but you should go through all your wiring. The ground to the chassis bothers me a bit, as well as grounding the LC1 heater to the same place as the ECU.

Ken

md95 · Post by **md95** » Thu Dec 20, 2007 9:34 am

muythaibxr wrote:This can be caused by many things though, most of them not code related.

Of course we'll check into it, but you should go through all your wiring. The ground to the chassis bothers me a bit, as well as grounding the LC1 heater to the same place as the ECU.

Ken

Well, yesterday i redid all the grounds, brought in 2-14 gauge twisted pair grounds right from the neg batt cable, directly to the MS ground(does not share ECU grounds anymore). 12 volt power is directly from the battery through a ignition switched relay, with a stereo noise filter right before the MS db37 on the switched 12v wire. Power/ground wiring can't get any better than that. I have a brand new starter, and maybe 6 month old battery. Alternator is good, charging 14-14.4 volts.

I don't want to say it's code related either, which is why i haven't said too much about for this long..I wanted to take the time to rule out any other issue that could cause this. At this point, it seems like i have...only other thing would be the CAS itself... I only really notice this happen after a while of driving, mostly 25-30 minutes of driving, and thereafter. On long drives, it happens maybe every 15 minuts, or maybe once in an hour. No rhyme or reason, it's totally random. It's more pronounced at higher rpm, with more occurance(3k rpm+). It's never happened at WOT. If you do end up testing this out, it will have to run for quite a while..If you need an msq, let me know..i'm pretty sure i'm on beta 15 yet though.

EDIT: heres the msq file attached.. If you/James want to check it out, i'd greatly appreciate it.

msoultan · Post by **msoultan** » Thu Dec 20, 2007 1:37 pm

I thought I'd do a complete post of all the INI bug fixes I've made to my INI so far:

- Unset TRIGLOG (don't know if this is a bugfix, but other docs have said to disable it?)

- Updated tpsThresh and mapThresh with corrections made by Lance (correct scale and limits)
tpsThresh = scalar, U08, 584, "%/s", 0.10000, 0.00000, 0.00, 25.5, 1 ; * ( 1 byte)
mapThresh = scalar, U08, 585, "kPa/s", 1.00000, 0.00000, 0.00, 32767.0, 0 ; * ( 1 byte) threshold for MAE
changed to:
tpsThresh = scalar, U08, 584, "%/s", 1.00000, 0.00000, 0.00, 250.0, 0 ; * ( 1 byte)
mapThresh = scalar, U08, 585, "kPa/s", 1.00000, 0.00000, 0.00, 255.0, 0 ; * ( 1 byte) threshold for MAE

- Increased TPSOXLIMIT so that a higher value can be used (sometimes TPS goes over 100)
TPSOXLimit = scalar, S16, 590, "%", 0.10000, 0.00000, 0.00, 100.00, 1 ; * ( 2 bytes)
changed to:
TPSOXLimit = scalar, S16, 590, "%", 0.10000, 0.00000, 0.00, 255.00, 1 ; * ( 2 bytes)

- Enabled EGO step count and Proportional gain for simple EGO algorithm:
field = "Ignition Events per Step", egoCount, { egoType == 1 || egoType == 2}
changed to:
field = "Ignition Events per Step", egoCount, { egoType == 1 || egoType == 2 || egoType > 2 && egoAlgorithm == 0}
and
field = "PID Proportional Gain", egoKP, { egoType > 2 && egoAlgorithm > 1 }
changed to:
field = "PID Proportional Gain", egoKP, { egoType > 2 }
goType > 2 }

I have verified that both the step count and the proportional gain are being utilized in the simple mode in MS2Extra.

Hope that helps!
Mike

Post by **muythaibxr** » Fri Dec 21, 2007 8:56 am

md95 wrote: EDIT: heres the msq file attached.. If you/James want to check it out, i'd greatly appreciate it.

James or I will check it out during the upcoming week.

It sounds a bit like the rollover code is broken again.

Ken

md95 · Post by **md95** » Sat Dec 22, 2007 8:43 pm

muythaibxr wrote:
md95 wrote: EDIT: heres the msq file attached.. If you/James want to check it out, i'd greatly appreciate it.
James or I will check it out during the upcoming week.

It sounds a bit like the rollover code is broken again.

Ken

After more testing, apparently the Megasquirt doesnt like sudden TPS signal drops. After driving around all day today doing some last minute shopping(fun!), i was getting tons of what i thought was lost syncs..so many stumbles..I finally got the computer out and went driving around..When the stumbles occured, the accel light was going nuts, and the throttle position gauge would quickly blip from -10% to normal..Opened the hood, then quickly unplugged and plugged in the TPS connector on the sensor, and bam, replicated the behavior. The sync light never flinched at the issue, no resets, so maybe the code is not broken? I just know for a fact when this happened before on the highway the ASE light had flinched on and off, but this time it didnt....

I have a nice 3 hour drive coming up, each way, all in one day, so that will be a nice test for any lost wheel syncs. I'll report back with any findings, until then, dont waste your guy's time just yet...in the meantime, i'm getting a new TPS and making sure the wiring/connections to it are perfect. I just found it very strange how the MS reacted to a total loss of tps signal...it wasn't happy!

UnaClocker · Post by **UnaClocker** » Sat Dec 22, 2007 9:28 pm

Try driving for a bit with the thing fully unplugged.

md95 · Post by **md95** » Sat Dec 22, 2007 10:02 pm

UnaClocker wrote:Try driving for a bit with the thing fully unplugged.

Yep i did that, and it ran perfect...Pretty much went from doing almost all the time, to unplugging it, then never did it again.

Post by **muythaibxr** » Sun Dec 23, 2007 12:14 am

That's good news. Ignition issues can be a pain to debug!

Ken

md95 · Post by **md95** » Mon Dec 24, 2007 11:17 pm

TPS issue is fixed, but still loosing wheel sync..only really noticed at around 3k on the highway today..will report back tomorrow night to confirm 100%..

Will the 0-255 second counter revert to 0 again during a wheel re-sync? Or is that only on a reset? Also noticed after a few re-syncs, the IAC stepper had opened up a little bit and backed off steps. Had a bit higher idle when i got off the highway...the TPS issue definately did not do that.. I'll have about 6 hours of good highway drive time to get some decent data anyway.

Post by **muythaibxr** » Tue Dec 25, 2007 9:03 am

The 0-255 counter should not go back to zero unless the whole MS is resetting. Losing wheel sync should not cause that.

Ken

md95 · Post by **md95** » Tue Dec 25, 2007 9:11 am

muythaibxr wrote:The 0-255 counter should not go back to zero unless the whole MS is resetting. Losing wheel sync should not cause that.

Ken

Ok, just wanted to know, as i'll take note of that on the drive..I should be able to get some decent datalogs..could be a bit long, but i should catch the issue.

md95 · Post by **md95** » Tue Dec 25, 2007 10:46 pm

Ok, back from the trip..It's definitely doing something, what seems to be lost wheel syncs. It's not the TPS issue, the datalog and gauges dont show any odd behavior like before. Also, at the issue, the IAC backs off steps, and then goes back to full warm steps, and EGO correction is delayed as well, just like when you do a full warm start, or data burn. The ASE light does flash on and off like I had noticed before.

Here is the datalog. I marked it with the spacebar in 3 places, with the issue a second or two before the marks. I'm not sure really what to look for at the data, when it concerns a lost wheel sync, so more decerning eyes will have to have a look. Nothing odd that i can see, except for sharp drops in some perameters at the lost sync, esp the rpm.

EDIT: the datalog is over 1mb in size! wont let me attach it, so i'll post a link to it..

http://www.geocities.com/mitsurst95/misc/1.xls

Megasquirt Support Forum (MSEXTRA)

MS2/Extra 2.0 beta 16 (20071122)

INI Changes rollup