GregC

I am having trouble with the emulation function in TunerPro RT with the Autoprom hardware. I am helping tune a 93 Z28 (DA3 mask) and we have had several issues with emulation since we first started using this stuff. At this time, we cannot get emulation to work while connected to the ALDL.

On some of our early test drives with emulation enabled, everything seemed normal early in the test drive, but after 2-3 miles it seemed that "online" changes were not affecting the performance of the engine (i.e. reducing VE values with no change in INT/BLM). We noticed that after the test drive, the Autoprom was quite warm since we have to locate it under the hood during test drives (93 ECM is underhood).

Does high temperature cause issues with the Autoprom? When not emulating, we are able to read ALDL data and we can read/write chips with no problem. Does anyone have any suggestions that we can try in an effort to get this working?

Even without the emulation, these products have enabled us to tune the car to a level better than it has ever been since in the current owner's possesion. I am greatly appreciative of all of the archived information on this site as it has greatly reduced our learning curve.

Craig Moates

This is a bug in TunerPro (assuming that is what you're using). Try downloading a beta version here:

http://www.tunerpro.net/beta/Bailey/...ProRT_v400.exe

GregC

We are currently using TunerPro RT v4.13.

I mentioned this emulation + ALDL issue in another thread where some guys using the $58 mask were having similar problems. Mark replied that there is no issue with this on the 93 Z28 $DA3 mask. Here's a link to the other thread.

https://www.thirdgen.org/forums/diy-...-searched.html

The beta version is listed as v4.00. Do we need to "roll back" from v4.13 to this v4.00 beta software that I just downloaded from the link you provided? Sorry if I'm making this more complicated than it needs to be.

junkcltr

iTrader: (1)

I have had several cases of that happening over the past two weeks with TunerPro v4.13 beta (comms fix beta). I run on Win98. From what I have found it appears to be due to a memory leak. Sometimes the leak cause TP emulation problems, sometimes it doesn't. TP usually crashes soon after this problem occurs. The memory leak causes all kinds of problems. It is somehow related to the fonts that are used Anyway, Mark is aware of the problem. It occurs in both the RT and non-RT versions.

Very good question. Most electronics are rated for a certain temperature range.

Both of those guys work very hard on the stuff they put out there. The great thing about the forum is people willing to share info. Everyone benefits from it. There are some people that like to "keep an edge" by not telling a lot, but in the end a group will figure it out and eventually surpass those types of people. So yeah, this is a great forum with lots of good (and bad) info.

Mangus

Wrong link. Get the latest here:

Index of /beta/Builds

Craig Moates

Oops, sorry for that bad link.

On the emulation issue, it does sound like a software issue. One thing you can try is periodically re-upload the binary. This can get around software issues which fail to perform incremental updates. But if you're getting the audibles, then the updates should be occurring.

The AutoProm should be OK in the temperature ranges you're going. The NVRAM and micro are both industrial rated. Logic chips should also be OK. The plastic case will probably soften up before it gets hot ehough to be a problem. Might put some shielding in there if you're concerned. Watch out for rain though, it's NOT weatherproof!

GregC

Thanks Mark - we will try this update later in the week and I'll post our results.

That makes me feel better about our application with the Autoprom. The 93 Z28 is a little different than thirdgens in that the ECM is located underhood in the RH rear corner of the engine compartment. When trying to emulate, we had the Autoprom placed below the ECM in a "well" formed between the strut tower and the firewall. The case was not soft as you mentioned, so I believe we can proceed without fear. We thought about placing the Autoprom in an insulated container if heat is a problem, but hopefully this won't be required.

junkcltr

iTrader: (1)

I use that method all the time on the bench. It works. I had an interesting failure tonight. I was editing an ADS file using TP GUI and TP crashed. I went to open the ADS file but it came up blank. I opened up Explorer and the file size was 0 bytes. It wiped it out. Strange. Never had that happen before.

With the good yields and quality of todays manufacturing. Even the commercial parts are usually industrial temp. grade or better. That helps.

Mangus

The crash occured after the file was opened (and wiped) and before any data was written to the file. You seem to find rather uncommon issues!

junkcltr

iTrader: (1)

I had TP running. I stopped the ALDL connection. I opened the ALDL ADS file setup box and clicked on the "ALDL sensor values". I entered a new item and then did a "save and close". TP just locked up. I killed TP and re-opened. I would just hang at the splash screen. I killed it again. Opened TP and it hung at the splash screen. Killed again, then opened MS-Excel (to adjust OS memory). Ran TP and it came up fine. Tried to open monitor window and all was "no item", and all of the senor and data bit stuff was blank in the ADS GUI window. Went to MS-Explorer to look at the ADS as text. Open it up and it is blank. Close it and look at the file size.........0 bytes. ?????

I still get failures from Acrobat "fonts sever failure". I get "no fonts" from MSYS when trying to run EMACS sometimes. TP crashes with the fatal execution error box all the time. They always seem like random crashes. The Win98 Start menu goes blank sometimes.

I have been thinking of throwing a copy of MS-Visual Studio C++ 4.0 on it so that it will run the debugger in info when it crashes. I could give you a stack trace in cases where it crashed. Would that help? It is stange, sometimes it will crash every three minutes when moving around between windows and sometimes it takes an hour. It does get to be a pain when it does it every 3-5 minutes (and will not retain the Monitor window items). Either way, it is definitely usable. It makes code & cal changes a lot easier to see with using the ALDL data for debugging.


EDIT: I really wish I didn't come across this weird stuff.

Mangus

Yeah, a call stack would be helpful, but based on what you're seeing, how often you're seeing it, and the symtoms (i.e. fonts, etc), I'm nearly certain there's something else going on related specifically to your system.

I still have the emails you sent some time ago, and as soon as I get back to TunerPro work (I'm currently working on the scanning side of things, to be released as ScannerPro and eventually integrated into what will be TunerPro v5), I'll investigate further.

I wish you weren't seeing such weird stuff too!

Z69

I had a similar problem with 98 that went away when I got a new video card.

I had the emulation & no update problem with the previous to v4.13 also.
I upgraded in hopes of eliminating it....
Now that I think about it. I started using an xdf that had the different catagories about the same time. If that makes any diff.

Mangus

Try the build linked above and let me know if you still have trouble. There is a known issue in the .400 build where changes made with the table editor toolbox are not updated.

junkcltr

iTrader: (1)

Ha ha, darn S3 video card. It used to have fits with older versions of TP.

Strange about the XDF file difference. I had been working on the $58 stuff with a $58 XDF from NWSTP and a $58 ADS file from somewhere (don't remember) and I rarely got the font changed, but got the usual fatal error crashes.
I have been writing routines for the $8D code for the past few nights with the stock $8D ADS file and super_8d_2 XDF file. The font problem change happens quite a bit now followed by a crash.

I will load up VC++ 4.0 this weekend.

junkcltr

iTrader: (1)

I wouldn't be surprised. Compaq doesn't exist anymore for a reason. I missed out on some auction laptops this Spring. Maybe I can pick one up this Fall.

GregC

We did some testing last night with the new beta software from the link that Mark provided. The new software seems to have improved our situation considerably! We were able to successfully emulate and read/log ALDL data simultaneously.

We did have a problem with incremental updates. When I tried to make an incremental update to the lower VE table, as soon as I hit the "Update" key the SES light would turn on and stay on. To correct the problem, we had to turn the key off and re-upload the bin just like you mentioned. After this, the SES light would go out and the vehicle resumed normal operation during the next emulation session.

On incremental updates, is there a limit to the number of cells that can be changed at one time? The first time the problem happened, I was trying to change 6 cells in the VE table. After we re-uploaded the bin, I tried just changing 1 cell and had the same problem.

Even with this issue, the emulation is a big time saver. It means that we can stop the car, make changes, re-upload, and be back on the road without ever opening the hood to change chips!

We were also using a new laptop that my friend purchased yesterday because we were tired of borrowing one! The new laptop is a Gateway with Windows XP and S3 video. Is there a problem with TunerPro and the S3 video adapter? Additional specs on the laptop are listed in the link below:

Gateway - 370 Notebook - MX3228

We placed the Autoprom in a small flexible "sandwich" cooler underhood. It seemed to help keep the unit noticibly cooler. I figured it was better to be safe than sorry.

Thanks to everyone for all the help.

Craig Moates

Did you disable checksums by uploading AA to 08? It's an option under the Tools/Preferences menu in TunerPro. I think it gets set back to 'unchecked' when you install a new version, so just a thought.

junkcltr

iTrader: (1)

Not that I know of. I once had a problem with a Linux install with an S3 Virge card but that has nothing to do with the thread. Video cards are mostly just an opinion thing......much like which is the best ECM to use for a given application. I like ATI stuff now......once liked NVIDIA.......and once liked MATROX. In the end, they are all good.

Good to hear you have it working. The SES coming on during emulation could be caused by many, many things. Start off by trying what Craig said to do.


Mark,
I went to the HP/Compaq site and saw that they have the "latest" year-2001 drivers for my machine there. I will give them a try this weekend (check what I have first) and see if it makes a difference.

Mangus

There is no limit to the number of cells you can update at once in a table. If the number of cells is over a certain number (I don't recall off the top of my head), the entire table is updated. This was because the Romulator didn't like to do scattered updates (the Ostrich & AutoProm have no problem with it).

I think Craig has the right suggestion here - make sure you've disabled the checksum by uploading 0xAA to the mask ID byte. That sounds like it's your issue.

Regarding the S3 issue - it's not likely to be the graphics card itself. It's usually a driver issue. When crashes happen, if the OS tells you that it's in some other module (not TunerPro), it's most likely an OS/driver issue.

GregC

This advice was right on the money! We made this simple change and incremental updates were no problem. I am not exactly sure what checksums are and strictly went by the advice of you guys' experience. I will do some more research on that topic and would be appreciative of any known links that explain the function of checksums and how/why this operation had to be disabled.

BTW Mark - I tried to copy the 2000 rpm row from the lower VE table and paste it into the 2000 rpm row in the upper VE table and could not get this to work. Is it only possible to copy & paste within the same table?

Now the fun really begins. We are really trying to dial in our VE tables since the camshaft has been changed by the previous owner (specs unknown). I am starting to see how some folks on this board have burned over a hundred chips trying to get the "perfect" tune. This LT1 exhibits the "Split BLM" trait even in the higher RPM ranges, so I haven't quite figured out how to attack that problem . I'll post tuning questions in another topic if I can't figure it out searching archived information.

I really appreciate the expert help and these really cool tuning tools. This stuff is quite addictive once you get a sniff of success!

Z69

There's a little piece of code in the program that adds up all the code on the chip from point A to B. This total value is then checked against a stored value in the cal area on the prom. If they do not match, then you go into limp mode. When your emulating, they will never match after you do the first edit.
So if you put AA in, the checksum routine will not be done and the two values are never calc'd or compared as designed from GM.
This is to prevent a damaged bin from trying to operate the engine.

junkcltr

iTrader: (1)

True for TunerPro. False for TunerCat. I found this out while writing code for my emulator to work with both TP and TC. TC keeps track of the checksum at all times and stores it. TC plays with some other stuff on the emulator too.

Mangus

TunerPro used to update the checksum with any emulation updates, but I found that doing so was unreliable (there was a period of time between the change and matching checksum update which still caused a hiccup). I subsequently took out the feature.

TC actually uploads AA to the mask ID byte for you, and stores the original mask ID in emulation EEPROM. TC restores the mask ID byte before doing a complete download.

junkcltr

iTrader: (1)

Way off topic.......but is there a chance that TP could support MonteCarSlow's NVRAM module?

He has a TCL script that has modular functions and it is prerry straight forward in terms of protocol. Basically, ALDL data transfers.

EDIT: While I am off topic. Is there a way to chop up ALDL log files? I have a few that are long and want cut certain pieces out of. But the file is binary......so I couldn't figure out where a line started and ended.

GregC

If I understand this correctly, checking the "disable checksum" option in TP preferences (which loads the value AA into byte 08) only applies to emulation mode. The checksum value is still calculated and placed in memory when I burn my bin file onto a chip. Is this correct?

Mangus

Correct.

GregC

Mark

I have noticed that with this beta version (v4.14), we have issues regarding copying & pasting values in the tables. When copying & pasting a from a compare bin to the current bin, the number of changed cells during the paste doesn't match the number of copied cells. In some cases, the copy & paste fails completly. Can you check this and report your results?

Mangus

Index of /beta/Builds

GregC

That fixed the problem - thanks for the update.

The Autoprom & TunerPro RT are great products. It blows me away knowing that you guys are just doing this stuff on the side in addition to your regular "day" jobs. Thanks for all the support.