Print Page | Close Window

Just lost two controllers... can't upload firmware

Printed From: Aurora
Category: Whatever Else
Forum Name: Off Topic
Forum Discription: The name says it all
URL: http://www.aurorashow.com/forum/forum_posts.asp?TID=995
Printed Date: 23 May 2024 at 11:40pm
Software Version: Web Wiz Forums 9.06 - http://www.webwizforums.com


Topic: Just lost two controllers... can't upload firmware
Posted By: LightsOnLogan
Subject: Just lost two controllers... can't upload firmware
Date Posted: 25 Dec 2009 at 8:19pm
Two of my ACx8s just became unresponsive.  They will not discover, and I cannot connect to them with the firmware updater via I/O 1 jumper to force bootload.  I'm out of spares, so I'm hoping someone can tell me how to restore these to working condition.

It would happen on Christmas day, wouldn't it???



-------------
http://www.aurorashow.com/">



Replies:
Posted By: deweycooter
Date Posted: 25 Dec 2009 at 8:27pm
PM sent

-------------
http://www.deweycooter.com/wiki/index.php/Aurora - Aurora Lights Wiki
http://www.aurorashow.com/">


Posted By: Jonathan
Date Posted: 25 Dec 2009 at 8:44pm
Ha!  Just saw this on D-Light and responded with the exact same thing.  Well, I hope the 'hail mary' trick works for you.


-------------
~Jonathan


Posted By: LightsOnLogan
Date Posted: 25 Dec 2009 at 8:53pm
Didn't work... if anyone knows the EEPROM address where the controller ID is stored on an ACx8, please let me know.  I suspect from testing that part of the issue is a lost ID and I could slap the PIC in a programmer and rewrite the ID into the EEPROM address.

I noticed that I have other ACx8s on the network which are still "working" but will not discover in HWUtil.  I suspect they will be dropping out soon.



-------------
http://www.aurorashow.com/">


Posted By: LightsOnLogan
Date Posted: 25 Dec 2009 at 9:09pm
My wife just handed out some candy canes and explained that we'll be working on it.  Everyone seems to be enjoying it even with 16 of 48 mini trees missing, so I'm going to keep it running instead of switching to the tech difficulties message.

One car had a 2 year old who said... "I can fix it!  I got tools!"



-------------
http://www.aurorashow.com/">


Posted By: Jonathan
Date Posted: 25 Dec 2009 at 10:01pm
Time to grab as many extension cords and power strips that you have and plug the trees into whatever else works.  It will look painfully strange but people are more interested in the lights doing anything than they are in having them doing them in the right order.

-------------
~Jonathan


Posted By: LightsOnLogan
Date Posted: 25 Dec 2009 at 10:14pm
It's not the ID... I figured out the EEPROM address for the ID by reading a good PIC and the non-responsive one still has its ID in that address.  I wonder what makes these become unresponsive?? 
 
Note: I already PIC swapped one earlier this year which exhibited the same inability to discover and accept firmware.  That makes 3 PICs down, all ACx8s.  There are others in the yard that are still running but will not discover.  I'm wondering how many hours it will be before they vanish too.
 
I thought about retiring the ACx8s this year but instead decided to put that on next year's budget.  I'm regretting that decision now.


-------------
http://www.aurorashow.com/">


Posted By: Jonathan
Date Posted: 25 Dec 2009 at 10:18pm
It's happening to the ACx16's too.  I think it has less to do with the controller and more to do with firmware but that's just my gut feeling. Since I'm not an expert I can't comment with any real authority. 


-------------
~Jonathan


Posted By: ChrisL1976
Date Posted: 25 Dec 2009 at 10:32pm
you guys got me worried now.  I checked my controllers and lights before the show.  All controllers got discovered but no lights on one full controllers.  Had to unplug it before it would kick the lights back on.  

-------------
Chris

www.lightsonsixth.com


Posted By: LightsOnLogan
Date Posted: 25 Dec 2009 at 10:40pm
I'd agree, except these two ACx8s were rolled back to 1.06 after I had problems with 2.0.  I ran 1.06 for 3 years without incident.  Of course, that is unless 2.0 did something that a 1.06 rollback doesn't reset.  That wouldn't surprise me given some of the oddities going on with ACx16 1.17 and roll down/up procedures that I've read about. 
 
I kept my ACx16s on 1.13 this year (I'm thankful I did that now). 
 


-------------
http://www.aurorashow.com/">


Posted By: LightsOnLogan
Date Posted: 25 Dec 2009 at 10:43pm
Originally posted by ChrisL1976

you guys got me worried now.  I checked my controllers and lights before the show.  All controllers got discovered but no lights on one full controllers.  Had to unplug it before it would kick the lights back on.  
 
I had to power cycle these exact same two controllers a couple nights ago because they were dark (it fixed it a couple nights ago).  When I saw it tonight the first thing I did was go to the breaker box figuring they were just due again.  It didn't work tonight and you know the rest of the story that started this post.  That's not what you want to hear I'm sure, but you have reason to worry.


-------------
http://www.aurorashow.com/">


Posted By: Jonathan
Date Posted: 25 Dec 2009 at 11:02pm
Trying to remember... I believe Firmware Updater v2.2.2 corrected an issue where the header wasn't being read properly.  I tried to find the post on D-Light about it but was unsuccessful.  So I suppose that an update done before v2.2.2 *could* have been corrupted.


-------------
~Jonathan


Posted By: LightsOnLogan
Date Posted: 25 Dec 2009 at 11:15pm
I just checked... the 1.06 downgrade was done with 2.2.0.  Maybe these firmware writes are going out of bounds and corrupting the bootloader?  Nah... that shouldn't be possible.  The bootloader on these things are supposed to protect from that happening.


-------------
http://www.aurorashow.com/">


Posted By: Jonathan
Date Posted: 25 Dec 2009 at 11:46pm
It'll be interesting to see Darryl's thoughts on this.  Seems to be happening across the board.

-------------
~Jonathan


Posted By: ChrisL1976
Date Posted: 26 Dec 2009 at 6:19am
Jonathan...is this happening on non-upgraded acx16's as well or only ones that went to 1.17?  That particular controllers has been 1.13 since last year since its never acted up, I didn't mess with it.    I've had controllers not get discovered occasionally, but that just took a few trys and they popped up.  Never one get discovered and not light up.    

-------------
Chris

www.lightsonsixth.com


Posted By: deweycooter
Date Posted: 26 Dec 2009 at 10:11am
Mine happened on a new ACx16 - purchased this season.


-------------
http://www.deweycooter.com/wiki/index.php/Aurora - Aurora Lights Wiki
http://www.aurorashow.com/">


Posted By: Jonathan
Date Posted: 26 Dec 2009 at 11:16am
Originally posted by ChrisL1976

Jonathan...is this happening on non-upgraded acx16's as well or only ones that went to 1.17?    

Hard to say for sure.  All my controllers were running various beta firmware at one point in time and I didn't leave any controller untouched.  Currently I'm running a mix 0f 1.15Beta and 1.17, and the DCx16's are split between 1.11 and 2.0 because I can't get the strobe twinkle mode to engage on 2.0.


-------------
~Jonathan


Posted By: LightChristmas
Date Posted: 26 Dec 2009 at 1:02pm
Have 8 ACx8s running V2.0, two Acx16 running 1.14 and two ACx16 running 1.15 under Aurora 1.0.68 since Thanksgiving with no problems. Is it possible that later versions of Aurora could be messing with them?

-------------
http://www.aurorashow.com/">


Posted By: LightsOnLogan
Date Posted: 26 Dec 2009 at 3:33pm
Just got the ACx8s working even though they still will not accept firmware.  I isolated the controller and resent the baud rate (Auto) to the controller and it came to life.

This doesn't make sense...
1. The controller would respond to the "All" events (FF) even though it had a corrupted baud rate, yet it wouldn't respond to its own address until I resent the baud???  That's just plain bizarre!
2. The controller now discovers and responds again, but it will not take firmware (I wanted to reflash to be on the safe side but couldn't get it to take the firmware)

I did notice that if I power cycle an ACx8 while the show is running it does dump a bunch of garbage onto the RS-485 which freezes the show for over a second.



-------------
http://www.aurorashow.com/">


Posted By: LightsOnLogan
Date Posted: 26 Dec 2009 at 4:18pm
Originally posted by LightChristmas

Is it possible that later versions of Aurora could be messing with them?
 
I can guarantee it is not Aurora.  I just repatched in the network and went to run a test show prior to restarting the schedule.  #40 had dropped out again during the time nothing was on or even plugged in to a network.  Once again, resending the baud rate brought it back to life.
 
I really wish I could get the firmware to flash... I would actually like to be able to leave the house and know things are going to keep running!


-------------
http://www.aurorashow.com/">


Posted By: deweycooter
Date Posted: 26 Dec 2009 at 8:50pm
That's interesting - I'm interested in seeing if I can resurrect my ACx16 too...


-------------
http://www.deweycooter.com/wiki/index.php/Aurora - Aurora Lights Wiki
http://www.aurorashow.com/">


Posted By: LightsOnLogan
Date Posted: 26 Dec 2009 at 9:02pm
Another possibility here... could the RS-485 chips be experiencing thermal issues leading to data corruption on those controllers (and subsequent crashes)?

I'm trying to isolate it now... I did some PIC swapping.  We'll see if it follows the PIC or not.




-------------
http://www.aurorashow.com/">


Posted By: LightsOnLogan
Date Posted: 27 Dec 2009 at 11:55am
The problem DID NOT follow the PIC microcontroller!  #40 and #43 still decided to lose it even after PIC replacement.

I'm now moving around RS-485 COM chips to see what happens.  I did notice that they were the old Intersil ones that some have reported problems with instead of the Texas Instruments (now Maxim?) ones.

I need more lights on the mini trees... I can't see them in the daylight unless I'm right on top of them!


-------------
http://www.aurorashow.com/">


Posted By: LightsOnLogan
Date Posted: 27 Dec 2009 at 7:47pm
As I noted in the D-Light forums, it seems that replacing the RS-485 transceiver chips stopped the EEPROM corruption. 
 
I hope that was it anyway... no falures for seven and a half hours.
 
If you are having issues, you might want to check your transceiver chips.
 


-------------
http://www.aurorashow.com/">



Print Page | Close Window

Bulletin Board Software by Web Wiz Forums® version 9.06 - http://www.webwizforums.com
Copyright ©2001-2007 Web Wiz - http://www.webwizguide.com