Page 3 of 5

Posted: Mon Oct 12, 2009 12:09 am
by Craig Moates
Hi Folks, I'm not exactly sure where Calvin stores the serial number in the Ostrich EEPROM, I've left that sort of thing to him in general. This can be done using the 'H'+'W'+n+MSB_LSB+B1+..+Bn+CS command that is published for my hardware. It can also be read back using the 'H'+'R'+n+MSB+LSB+CS command. But, you've got to be careful on what segment is utilized. It's contained on the AVR, not the FTDI. I haven't seen a situation really where AVR EEPROM gets corrupted unless brownout or major voltage spiking is going on, but that should all be isolated and protected on the Ostrich really. And even if it did, simply re-applying the serial to the AVR EEPROM is the right answer I'd think, assuming that this doesn't cause an administration headache. Still, it shouldn't happen. There are also additional timing protections and such to avoid that sort of thing. Calvin, where is it being stored at in EEPROM? A good area would be, for instance, 0x0080-0x0087 or something like that maybe. There's lots of flexibility. Some hardware ID info is in there as well. Knowing where eCtune stores it or expects to see it would be helpful I think. True, using other software could clobber that area. However, I don't think anyone else uses this approach of locking Ostriches via serial to software licensing. I'd typically recommend that folks use the Ostrich for tuning, and then install a chip for the final calibration. Running off of an emulator, Ostrich or otherwise, on a relatively permanent basis is cool and all, but it is a more complex device than a chip. Calvin, could you let me know what strategy you are using with respect to tagging the Ostriches? I suppose it's possible that a reserved area is being used, but I thought we went through this at some point. The same strategy of using the EEPROM area to uniquely identify hardware is being used on several of my other devices without issue, but those are typically set up to be protected as read-only. If the software is constantly being updated and refreshing the serialization of the hardware, that could cause issues, particularly if the serialization routine isn't verified back afterward. Battery voltage should certainly have nothing to do with this whatsoever. If there is a bad Ostrich battery in the field, I have yet to see it. I've seen many folks suspect the battery, take it out, measure it, mess up the circuitry in doing so, and chase their tails in that regard, but Ostrich batteries just haven't showed up to the game where they are a problem in anything but an imagined scenario. If the vendor ID is changed, such as can happen when resetting the Ostrich with software such as TunerPro, EmUtility, OstrichReset, Neptune, or even as an option within eCtune, then indeed the Ostrich will self-erase by design. This should only occur at the time a vendor ID byte change request is executed though. If we need to take this offline, feel free to moderate as needed. I'll be glad to help you get to the bottom of something Ostrich-side or Software-side if you think there is something problematic. If there's anything I can do, I certainly will. Hope this helps, Craig

Posted: Mon Oct 12, 2009 4:31 pm
by calvin
i have done tests here.. resetting vendor in ectune doesn't change serial.. i'll double check everything.... but suddenly unpaired? i use the serial set option in the newer ostrich firmware

Posted: Mon Oct 12, 2009 11:54 pm
by calvin
ok... so far what we know.. ostrichReset app did not remove the serial. The updated version removes the serial number eCtune.. eCtune doesn't have code to write a new serial if the serial is not blank(new ostrich) Craig moates and I are working on it very hard. I'm gonna test tunerRT and crome if they reset the serial. @Craig: The reset in ectune set the vendor to 0. The serial doesn't get wipe. Only the rom area is wiped. I'm using the command supplied for vendor and serial for ostrich 1 and 2 OstrichReset update 10/13/2009 removes the serial. I'll try emuUtil and other software.

Posted: Tue Oct 13, 2009 12:15 am
by calvin
emuUtil: Reset Vendor to 3 but serial remains the same. TunerPRO: Reset vendor to 0 but serial remains the same. Crome Free: If vendor is ectune. Crome can't upload. Crome does not change the vendorID nor serial number. Ok now we have to debug futher.

Posted: Sat Oct 24, 2009 9:16 pm
by danz
when i had the issue, i was in the middle of doing a pass at the drag strip, and had been making changes to the tune all evening. the car almost appeared to shut down for a millisecond when it happened. ive never used my ostrich with any other software since it was paired with ectune. just to give some more information. thank you calvin and craig for looking into this.

Posted: Sun Oct 25, 2009 2:45 pm
by Synoptic
what about a bad ostrch 28pin cable ? some pins getting shorted or disconnected could corrupt ?? Vibrations in car, etc... loose socket ??

Posted: Sun Oct 25, 2009 5:16 pm
by CRX2
We checked the cable, and replaced it just too be sure....

Posted: Sun Oct 25, 2009 10:38 pm
by blowngsr1
i have had this problem also around 4 or 5 times over the last year. I have had to get a new serial from calvin every time.

Posted: Mon Oct 26, 2009 11:57 am
by calvin
wes? i'll check today. don't register we have to find out what happens i'm busy with craig. i can confirm that ectune doesn't write a new serial or does not remove the serial... something else must be going on

Posted: Mon Oct 26, 2009 3:44 pm
by blowngsr1
yes it is Wes I will wait on you to get back with me.

Posted: Thu Nov 05, 2009 8:29 am
by calvin
ran a few test with wes.. no couldn't reproduce. but i have a solution i'll release shortly.

Posted: Mon Nov 09, 2009 1:55 am
by blowngsr1
hey i was thinking. and i had to reformat my laptop so I switched to xp from vista. i wonder since i had ectune on vista and xp if that would explain the 2 vendor ids we had? well just a thought. Wes BTW it seems to be working fine for now. ill keep everyone posted if it screws me again.

Posted: Tue Nov 17, 2009 6:13 pm
by Craig Moates
From the FW standpoint, I wouldn't think that would matter. Calvin might be able to comment though. Like Calvin said, we've got a strategy which should put some provisions in place on both sides to guard against serial number corruption. Even if we can't reproduce the issue, we can add countermeasures just in case.

Posted: Wed Nov 18, 2009 7:51 am
by calvin
we will add countermeasures. But it's unexplainable!:) i guess we have these things in life.

Posted: Mon Nov 23, 2009 3:48 pm
by cyanide
Bump. Just had this problem after trying ectune for the second time. The Ostrich was not connected to an ECU at any point of time after being paired with ectune. I tested my new license with Ostrich, shut down my laptop and tried to open up ectune an hour ago. The Ostrich was unpaired and I had to request another license. I'm waiting for the email with the changed Ostrich license. Anything I messed up? I think I did a force-update after pairing, so perhaps that's what did it in my case...