Yes, it becomes pretty obvious when tables aren't set up correctly.
Yes, it becomes pretty obvious when tables aren't set up correctly.
I no longer monitor the forum, please either post your question or create a support ticket.
Thank you! I just BBF'd my Cummins CMB controller successfully with Feb. 12, 2012 Firmware. From the beginning of the flash (when it starts to check data file) to the end of flash (when you turn the key off) took an elapsed time of 7:14. Much appreciated! It sure is nice not to rely on the laptop and preserve the battery life of the laptop when making multiple changes fine tuning, etc..
-Andrew
Last edited by FUBAR; February 13th, 2012 at 05:32 PM.
06 5.9L, EFILive / Silver Bullet Tuning
06 LBZ, EFILive
BBx seems to be very stable so far!! Woohoo! Now when can us Cummins rednecks BBL?? Just kidding, you guys are doing a standup job!
06 5.9L, EFILive / Silver Bullet Tuning
06 LBZ, EFILive
Oh memories......
I remember when the E38 was brand new and BBL was only just starting. 10 min cal flashes from the laptop, setting up the laptop under the drivers seat to log long trips. The V2 and EFILive has come a long, long way in the past few years. To achieve both product enhancement and new calibrations is no small task. You guys have done a superb job and it shows when comparing with the competition.
Simon.
It applies to all controllers but more so to CAN controllers and even more so to Cummins controllers. Seems like anyone who wrote code for the various controllers had their own idea about how to implement the "keep alive" strategy. There's quite a few differences between not just manufacturer's but between controllers from the same manufacturer.
Anyway, just for everyone's benefit, the update I made was to redesign the keep-alive message strategy used by FlashScan/AutoCal. The keep alive message strategy is used to keep the various on-board modules in a particular state (e.g. logging, reading, flashing etc). If the keep alive messages stop (eg FlashScan is unplugged from the vehicle), then the controllers will (should) revert to their normal operating state. One of the reasons why reading/flashing may fail is controllers returning to their normal state because either the keep alive messages stopped or were not timely enough or some module (usually aftermarket devices) on the communications bus ignored the keep alive messages.
Previously FlashScan/AutoCal used a mix of synchronous and asynchronous keep-alive message strategies required for various controllers. It was manually configured for each controller and relied on known operating states of the target controller.
There is now just one type of keep alive message strategy for all controllers that works seamlessly with any controller in any operating condition and is fully integrated into FlashScan's/AutoCal's real time operating system.
It is more robust and it interacts with other modules on the bus in a predictable manner.
Regards
Paul
Before asking for help, please read this.
My head hurts after reading that....but many thanks! Let me guess...was this something specific that the only problem child that didn't like the strategy or message was the Cummins CMB controller? And when you say now there is a "more robust keep alive message strategy" in place, does that mean you didn't revert back to the previous firmwares, but in the new zip file changed to a whole new strategy? I think I'm reading your mail right... I have to say, it's been 100% reliable.
Thanks,
-Andrew
06 5.9L, EFILive / Silver Bullet Tuning
06 LBZ, EFILive
What happened was a mismatch between the test systems and the development systems over the past week while we were testing RC9 prior to its release.
One test system had already tested the CMB/CMC as functioning correctly, another test system found a fault with the E39 full-flash. Fixing the fault for the E39 full-flash caused the CMB (which had already been tested and passed) to fail in the RC9 release. That's how the CMB BBF failure got through our testing procedures. To prevent that from happening, every time a fault is found during testing we have to fix the fault and restart the testing process. The testing process is long and time consuming. It takes 3 people up to 3 days each just to test the various FlashScan/AutoCal read/flash/log, BBx and pass-thru combinations. I made the decision not to restart the testing because I didn't think that the E39 fix would cause any failures - how wrong I can be sometimes... :(
Once I realized what had happened, and saw that the keep-alive strategy for E39 and CMB was just too incompatible to implement with the old strategy, I bit the bullet and redesigned the whole thing based on what we had figured out over the years about the GM VPW, GM CAN and Cummins CAN controllers.
So to answer your question, yes this is a completely new update/strategy but I think it will end up being better and more reliable than the old system.
Regards
Paul
Before asking for help, please read this.
So much goes on behind 'closed doors' that most and myself included probably take for granted for the most part than we realize. That being said, I and probably can speak for others, appreciate the efforts and achievements the EFILive Development Team has done/doing!
-Andrew
P.S., When I sent you the BBx trace file, was I correct in how I saved it, or did I miss something. If you can't remember it's alright.
06 5.9L, EFILive / Silver Bullet Tuning
06 LBZ, EFILive
The file was empty, so it didn't help much. But before I was about to ask you for another version I tried the CMB BBF and got the same result as you so it was pretty obvious there was a problem - and it didn't take long to figure out what had happened.
The worst thing is when I can't reproduce the problem on our test system, that's when I really need the logs. If I couldn't have reproduced it here then I would have been pestering you get the trace file(s) again
Regards
Paul
Before asking for help, please read this.