PDA

View Full Version : Couple of AutoProm questions



brianko
11-19-2012, 05:13 PM
Have spent several hours now with my new AutoProm and TP v5 tuning with an APYP bin ($6E). Very enjoyable, can't imagine tuning without one! A couple of times, I discovered that the bin image in the unit no longer verified with the TP bin, but I had made no changes to anything, just a datalog run. Anyone know what's up with that? Also, can someone explain why one would want to disable checksums? I just left all of the checksum options at their defaults, everything seemed to work fine.

EagleMark
11-19-2012, 06:31 PM
I got spoiled with the AutoProm and had to get a RoadRunner...

When you change a paremeter in bin the checksum will change, hence the need for AA in mask ID. This may be your issue. Also you need to disconnect and the bin in AutoProm will be fine, it will run forever without being hooked up. IIRC if you turn off TP or pull cable something goes wrong.

When you make a change to Scaler or Table and Commit it will beep meaning change is done, but I don't hear that in Constants so I upload again.

brianko
11-21-2012, 05:57 AM
Still a bit fuzzy on the checksum business. I know what a checksum is, but I don't see the reason behind disabling it (presumably by putting "AA" in the codemask field?). The checksum should be recalculated after every modification is saved to the bin file, correct?

EagleMark
11-21-2012, 08:10 AM
Yes, everytime you save a bin in TunerPro it will recalculate the checksum. If you change something then checksum changes, save, recalculate.

Have to disable it to emulate, set to AA. Technically I don't know why, ECM checks or something while changes are being made? :confused:

brianko
11-26-2012, 05:05 AM
Strangely enough, I've been emulating without disabling the checksum, ECM seemed to pick up the changes just fine. I'm going to try with the checksum disabled and see if there's a difference....

1project2many
11-26-2012, 06:21 AM
If the ecm does a checksum recalculation and your emulator calibration is in the process of being updated the ecm will detect an error. Usually the check engine light will flicker and a "phantom" code 55 will be stored but I've actually had the engine stall or stumble and even had a backfire once large enough to blow off a K&N filter. You're playing around in the Land of Undefined Behavior when you're updating the calibration while the ecm is checking for unaccounted changes. Safest answer is to set mask to $AA so ecm doesn't check.

Six_Shooter
11-26-2012, 06:32 AM
If the ecm does a checksum recalculation and your emulator calibration is in the process of being updated the ecm will detect an error. Usually the check engine light will flicker and a "phantom" code 55 will be stored but I've actually had the engine stall or stumble and even had a backfire once large enough to blow off a K&N filter. You're playing around in the Land of Undefined Behavior when you're updating the calibration while the ecm is checking for unaccounted changes. Safest answer is to set mask to $AA so ecm doesn't check.

^^^THIS and +1

I've emulated without disabling checksum, and thought the ECM was detecting the changes, and while it was taking to some changes it didn't take to all changes, once I realized I forgot about disabling checksum, and changed it, the emulated changes took immediate affect.

For something so simple to change I don't see why you're questioning this so much.

The way I see it, a change in the bin will cause a need for a recalculated checksum, if the values in the current bin don't add up to the current checksum, the ECM will see it as an error. Any change will (could) cause this, so why take the chance? The only time this doesn't need to be disabled is when there is a re-calc of the checksum, without being used while the changes are made, I.E. changes made in a bin editor without emulating, and chucksum re-calc'd before being uploaded to an emulator or programmed to an EPROM for use.

Most of my current bins that I use in my cars and my friend's cars I leave the checksum disabled, even when I program to an EEPROM.

brianko
11-26-2012, 06:55 AM
For something so simple to change I don't see why you're questioning this so much.

Guess it's my way, asking questions. Thanks for the explanation.

EagleMark
11-26-2012, 07:51 AM
I've forgotten to change the checksum back to mask ID and never had an issue with running differant, data differant or CEL.

1project2many
11-26-2012, 10:11 AM
Chances are never resetting the "engineering byte" (unofficial name for $AA) will never lead to problems. I try to remember to reset it on my own cals although it can take me years to decide I probably won't be doing any more tuning to a vehicle. When I'm tuning someone else's car I always set this back to the correct mask id. All I need is for someone to zap an eprom then call me with a whole list of strange behavior that I can't identify.

You know, I can't ever remember an unmolested factory eprom giving me a code 55. It's kind of like the oil pressure switch in the fuel pump circuit. In the thousands of vehicles I've worked on, I've only encountered one or two that actually had inoperative relays. But it's still nice to know the switch is there.