Updating replicator firmware

[Paul Breed(unrocket)]

Has anyone successfully updated their replicator firmware to version
5.2?
I can't seem to get the firmware update to work.
IT keeps giving me the :

The Firmware update did not succeed Check the console for details,
You can click upload to try again"

Message with absolutely no information on the console.

Replicator G can communicate just fine with the maker bot it can jog,
turn leds on off etc...
It just can't update the firmware...

[Chip Turner]

I managed to by power cycling rather than hitting the (very difficult
to press) reset button. Basically power it off, wait a few seconds,
hit upload, then power it on. I tried resetting a few times to no
avail but the power cycle got it on the first try.

> --
> You received this message because you are subscribed to the Google Groups "MakerBot Operators" group.
> To post to this group, send email to make...@googlegroups.com.
> To unsubscribe from this group, send email to makerbot+u...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/makerbot?hl=en.
>

--
Chip Turner - ctu...@pattern.net

[Paul Breed(unrocket)]

Still no luck.
If I click upload before power is on then it fails immediately.
I need to wait til after power up when the computer makes the new
hardware attached noise and then it still fails.
(It just takes longer to fail ie 15 or 20 seconds rather than
immediate failure.)

> > For more options, visit this group athttp://groups.google.com/group/makerbot?hl=en.
>
> --
> Chip Turner - ctur...@pattern.net

[Dan Newman]

On 17 Mar 2012 , at 9:00 AM, Paul Breed(unrocket) wrote:

> Still no luck.

Make sure you are using RepG 34 and not 33. RepG 33 had a bug whereby
it couldn't upload firmware unless you put a copy of avrdude in the incorrect
location it was looking for.

Dan

[Paul Breed(unrocket)]

I'm using version 34.
Possible it still has the same bug?
It acts like its not even trying.

Is there an alternate way to upgrade using the avrdude or arduino
tools?

[Dan Newman]

On 17 Mar 2012 , at 9:35 AM, Paul Breed(unrocket) wrote:

> I'm using version 34.
> Possible it still has the same bug?
> It acts like its not even trying.

On a mac, I looked in the console logs and saw the problem with
invoking avrdude. (Applications > Utilities > Console).
On a Linux box, you'd probably want to look at syslog or one
of the files /var/log or /var/adm (I don't use Linux much these
days). On a Windows machine, it might be in the event viewer?

At any rate, if you look at the logs you may see an obvious
problem. If you instead see a lot of timeout errors, then it's
a different issue -- avrdude not managing to communicate with
the machine. (I'm here assuming that avrdude is used for
a mightyboard.)

Dan

[Paul Breed(unrocket)]

Tried it on an XP laptop this morning, still no joy.
I had assumed that console meant the console window at the bottom of
the repg screen.
I tried the event viewer and nothing looks like avrdude or replicator
stuff in there...

Paul

[Paul Breed(unrocket)]

Has anyone gotten the firmware upgrade to work from a windows 7 or
windows XP machine using the stock repg-34 install?

Paul

On Mar 17, 9:53 am, "Paul Breed(unrocket)"

[Benjamin Rockhold]

How's your reset-button timing?

I hold the reset down, and release and press upload at the same time.

My first method is timing with the power switch, pressing upload after exactly one second after powering the machine on. Whenever the computer I am using is slower to detect and initiate the serial connection, then I use the normal reset button.
However, I am using MacOS and Linux, so I suppose some performance may differ.

[John D'Ausilio]

I don't think you'll ever get it to work with power-cycling ...

what happens is when the processor first powers up the bootloader
waits 2 or 3 seconds to see if the host is going to send a firmware
update ... if an update doesn't show up in those first seconds it goes
ahead and starts the currently loaded firmware, if you hit the upload
button during those first seconds it will accept the new firmware and
restart

[John D'Ausilio]

sorry, hit send too fast ;)

when you power-cycle then windows has to detect and reattach, and by
that time it's out of the bootloader and running the loaded firmware

you need to, in rapid succession, release the reset button and press
the upload button in repg

[Paul Breed(unrocket)]

I never got it to work with the repg tools.
So I did the following:

I wrote a c program that records all the command line parameters....
I replaced the rpg avrdude with this...

giving me:

avrdude.exe -CC:\Users\Paul\Documents\ReplicatorG
\replicatorg-0034\tools\avrdude.conf -cstk500v1 -PCOM12 -b57600 -D -
Uflash:w:C:\Users\Paul\.replicatorg\firmware\Mighty-mb40-v5.2.hex:i -
pm1280

I then played aroudn on the command line where the timing of the code
start was much more precise....

I worked on this for an hour or so with no luck.

Then I wen back and reread the messages here, I had been power cycling
based on the advice here...
So I tried the command line version with the hit reset advice and it
finally worked.
That's six or so hours of my life I'd like to have back...



Paul

[John D'Ausilio]

wow, I just looked at the Mighty Board schematic ...

the USB power only powers a little atmega8, which appears to be a
usb<->serial converter
I've not looked at the initialization code in the firmware but I
wouldn't be surprised if (and the behaviour indicates that) the
atmega8 doesn't get initialized until the main 2560 wakes up.
so power-cycling will not work for loading the firmware

[Whosawhatsis]

You don't need to hit connect in RepG before uploading firmware, and the USB chip is powered by USB, so it can be recognized when the bot is off. At least that's how it works with older versions, I haven't examined the mightyboard schematic. Power cycling should work if you can get the timing right.

With Gen3, there's a footprint (unpopulated by default) for a 2-pin header. If you put a header there, you can use a jumper to enable auto-reset. I find this makes things much easier. I know Gen4 doesn't have anything like this because a trace is cut on the arduino mega to disable auto-reset, but it might be worthwhile for the users of these newer electronics to work out a modification to allow it to be temporarily re-enabled in a similar way.

[Benjamin Rockhold]

Power cycling works fine for loading the firmware. However, it is not as effective under windows as it is under MacOS and Linux. The time it takes for Windows to load up the driver is longer than the processor to exit the boot loader phase.

[Benjamin Rockhold]

The Mightyboard does not receive power from the USB port, and the serial connection is only initialized after the 8u2 boots up.