MS2/E Beta 5 issue with MAP sensor?

Jon k · Post by **Jon k** » Fri Jul 20, 2007 9:40 am

I flashed the Beta 5 to this MS2 unit - the map reading with no hook up is 12.2 kPa - this is with the ECU sitting on the stim, with the hose for MAP hooked to atmosphere - it reacts if I blow or suck on it, but obviously it's no where near 99 kpa or so.

I swapped an MS1/Extra CPU in the very same unit, MAP sensor reads perfectly. What is wrong? I am supposed to take this unit to the owners house and tune his car this afternoon.

It did this on Beta 2 or Beta 3 also, I posted a datalog that had crazy low kPa readings and now I think it's an issue

Megatune is setup for the standard 2.5 bar sensor - MS1/E works showing 99 kPa ambient, but MS2/E is extremely off.

Switched the MS2/E chip into another MS running MS1/E, (the one from my running car, known working) and the map is reading 12 kPa area. Whats up? Am I not setting something right?

Flashed the released 1.02 or 1.0.2 or whatever, and the map sensor read 99 - 100 kPa ambient pressure. So, something is up with a setting or something when I am using the betas. Does anyone have any information? I need the latest beta because I need to use > 255 degree trigger tooth

Post by **muythaibxr** » Fri Jul 20, 2007 10:31 am

Try reflashing. That is one of the typical symptoms of a bad flash.

I'm also not seeing that behavior.

Ken

Jon k · Post by **Jon k** » Fri Jul 20, 2007 10:37 am

muythaibxr wrote:Try reflashing. That is one of the typical symptoms of a bad flash.

I'm also not seeing that behavior.

Ken

Ken,

Flashed it 3 maybe 4 times.

I flashed the 1.02 release, and it works showing correct kPa reading on 2 difference ECUs, 2 different MS2 CPUs. If I flop my MS1/E CPU into either box, they work. If I flash the CPUs to beta 5 or such, I get the kPa issue.

Post by **muythaibxr** » Fri Jul 20, 2007 10:39 am

OK, again, I have not seen this issue occur on my MS2, and there are more than a couple people using beta 5 in their cars who have not reported any issues.

Did you go back and make sure that the fuel algorithm didn't get messed up in the upgrade? Is it doing that just after reflash or only after you load your settings?

Ken

Jon k · Post by **Jon k** » Fri Jul 20, 2007 10:42 am

muythaibxr wrote:OK, again, I have not seen this issue occur on my MS2, and there are more than a couple people using beta 5 in their cars who have not reported any issues.

Did you go back and make sure that the fuel algorithm didn't get messed up in the upgrade? Is it doing that just after reflash or only after you load your settings?

Ken

This is on a fresh flash - the one CPU is brand new never flashed until today, the other had a previous release on it that I flashed over, but no settings are even being loaded its plain after a fresh flash (multiple flashes).

Jon k · Post by **Jon k** » Fri Jul 20, 2007 10:51 am

Ken --

Sorry to have jumped the gun.

I flashed it literally 8 times now, kind of in a frenzy of "Take this freaking firmware, now, damnit!" and it's working.

I don't know what the problem is... but I have never had luck of flashing once, twice, or three times with MS2/E ever. It's always required 4, 5, 6 flashings.

Post by **muythaibxr** » Fri Jul 20, 2007 11:00 am

Your computer, or the serial port you're using with it might have problems. I've never had problem flashing an ms2...

The people who have had trouble have slower or older computers though... How fast is your computer? how old? what kind of serial setup is it (I forget).

In any case, it's difficult, if not impossible for me to debug the problem as I don't have a computer that's having the problem.

Ken

Jon k · Post by **Jon k** » Fri Jul 20, 2007 11:06 am

muythaibxr wrote:Your computer, or the serial port you're using with it might have problems. I've never had problem flashing an ms2...

The people who have had trouble have slower or older computers though... How fast is your computer? how old? what kind of serial setup is it (I forget).

In any case, it's difficult, if not impossible for me to debug the problem as I don't have a computer that's having the problem.

Ken

This is a modern laptop - its a Centrino 1.4 GHz Dell 600m with a 9 pin serial port. It's setup correctly and has no trouble datalogging, flashing MS1/E, etc., etc., etc. I really do not think it's a computer issue, as I am in the profession and know this computer is well capable of transmitting data at 115,200 bps - it's formatted running WinXP Pro with NOTHING other than MT/MS software installed.

rb26dett · Post by **rb26dett** » Fri Jul 20, 2007 11:12 am

its funny you should say that. i remember scratching my head when seeing the cpu hit 100% during a flash at some point. is the updated loader still a cpu hog like that? is that necessary to do what it needs to do due to the nature of serial comms? or could it be buffered in a suitable way? or is it that the firmwares flashing code could be more robust as discussed with eric, dave, bruce, and al a month or 4 back? just curious about something that cant need much in the way of actual cycles hogging many of them in a way that i assume means its using them for waiting in loops for responses and such. seems like if thats the case, something is far from optimal, and that at some points other programs running on the system are going to be stealing a time slice to do their thing possibly at a time when the downloader (why isnt it called uploader?) needs to have its stuff in the registers out of ram.

can you help me understand this, or is it not even the case anymore?

fred.

Post by **muythaibxr** » Fri Jul 20, 2007 11:17 am

This is a modern laptop - its a Centrino 1.4 GHz Dell 600m with a 9 pin serial port. It's setup correctly and has no trouble datalogging, flashing MS1/E, etc., etc., etc. I really do not think it's a computer issue, as I am in the profession and know this computer is well capable of transmitting data at 115,200 bps - it's formatted running WinXP Pro with NOTHING other than MT/MS software installed.

You would be surprised, even some modern laptops have trouble sometimes with 115200 speeds.

Have you tried fiddling with the various time delay settings on the downloader program?

Ken

rb26dett · Post by **rb26dett** » Fri Jul 20, 2007 11:20 am

thing is, its not a throughput thing, its likely a latency thing. hence my questions above.

fred.

Post by **muythaibxr** » Fri Jul 20, 2007 11:39 am

I know it's a latency thing... which is why I suggested setting the various delays a bit longer. Some of the delays are set by default to 0, and won't flash properly without making those delays significantly longer.

I was not part of any discussion on the reflash code, etc... so I'm not sure what you're talking about there. The only time I've seen ms2dl.exe hit 100% CPU usage is when it first loads the s19 file ... it loads the whole file into an internal data structure before it starts sending data over the serial port.

I can't imagine anything else in there that'd cause 100% CPU usage with the commandline version of the program. The GUI version might have something else GUI-related going on though.

Ken

rb26dett · Post by **rb26dett** » Fri Jul 20, 2007 2:06 pm

sorry, that latency comment was directed (poorly) to jon to explain why fast maybe not enough... i wasnt trying to tell daddy how to suck eggs

as for the 100% thing, it could have been some old version or b&g version. not sure what i was using when i saw that. i did see it though. i guess its even possible (probable?) that it was on linux with a wine lib, and that was causing the cpu consumption. not sure.

fred.