Inconsistent SerialUSB to IDE connection

LeafLabs Garden » Support » Maple Hardware Support

Inconsistent SerialUSB to IDE connection

(6 posts) (5 voices)

Started 4 years ago by RickRantilla
Latest reply from samtal

RickRantilla
Member

Arrghhhh! I'm using IDE 0.0.12 on a Windows XP system connected via USB to a rev 5 Maple. The interface seems to change from day to day.

I always select under "tools" "board" to "Flash" and "serial Port" "com3". The first few days I'd just hit "upload" and the program would upload as expected. When I press "Serial Monitor" I'd get a window on "com10" that would receive data from my application, as, for example in the "hello world" stub. If I went to "tools" "serial port" there would typically be this new port, com10, listed and checked.

Then I started getting "Couldn't find the DFU device: [lEAF:0003]" error, so I would use the "reset" while uploading. This required hitting reset at the right time so I also use the "Perpetual bootloader" scheme (pressing reset and the "but" button on the board). This got more awkward as I installed the Maple board inside my device.

Now I no longer get the second com10 port after a successful upload, so I can't see what my application is doing anymore.

I just reloaded "hello world" and can't get any feedback (no com10). I also changed it to include "blink" (so I'm sure my new program is there). The only serial port is com3 and it doesn't work as a monitor.

Any suggestions?
thanks in advance...

Posted 4 years ago #
LarryP
Member

@RickRantilla,

First off, I feel your pain -- and it shouldn't have to be so !@#$%^ hard!

If memory serves, one of MicroSoft Windows "features" is that it's pretty *stupid* about USB devices, and give even less in terms of decode-able error messages. One manifestation of M$ USB "features" is that the same device (e.g. Maple + cable) is recognized differently, DEPENDING ON WHICH USB CONNECTOR ON THE WINDOWS HOST IT IS PLUGGED INTO. So, if you get it working again (good luck; honestly, that's what it takes), note exactly which physical connector you have the cable plugged into on your PC and *ALWAYS* use that same USB connector, whenever you work on your Maple board. (And if at all possible, don't use that connector for anything else.) This may not solve all of the problems (remember, we're talking about M$ here), but it may help to get M$'s excuse for operating systems to work more nearly sanely.

My fingers are crossed for you! except when I'm typing ;-)

Posted 4 years ago #
RickRantilla
Member

Thanks, Larry. I'm using an old Dell portable with XP because I never could get the IDE running on my Windows 7 machines.

A different limitation (defect) seems to have turned into a feature for me. It appears you can't power the Maple from an external power supply and program it at the same time. So I hooked up a switch to the jumpers for USB vs. External power. I can now flip this external switch at just the right time while I'm compiling and uploading a revised program and somehow Windows XP "usually" hooks up both com ports. On to the next problem...

Posted 4 years ago #
poslathian
LeafLabs

Rick, we definitely feel your pain on the windows front. Using a single MCU to be both the processor you are programming as well as the programming interface turned out to be a very tricky setup, and while we have been able to smooth things over quite a bit in Mac and linux, Windows has always really really hated all the switching back and forth between "programming" mode and "serial port" mode. Not to mention all the naming issues and that crap. What is really needed is a custom windows driver to behave the right way and handle both the serial port and DFU modes in one driver. In our future products we are switching to a two-chip setup, where the programming interface chip is independent and always on, one permanent serial port.

However, should definitely be able to power and program at the same time. Just switch the jumper to EXT power (or BAT) and plug in the USB anyway.

Posted 4 years ago #
cyberia
Member

http://forums.leaflabs.com/topic.php?id=73

this fixes windows 7 driver signing so that you can use maple on your better computer

that may help

Posted 4 years ago #
samtal
Member

RickRantilla

Don't you feel you are left there alone with your pain. It is shared by many.
The USB probalem you describe has been reported many times in the past, with no real cure from LeafLabs (except blaiming it on Windows, when I guess 90% of us use Winodws XP).
I have several Maples r5, and occasunaly I struggle with it up to a level that I just want to give up the Maple, and use a proffesional development board.
My problem is that I work remotely on the board, thus I can not press the reset.
My solution is as follows:
I use a SiLabs USB to serial board to convert from the 5V usb to the 3.3 V Maple.
(I also have one Prolific USB to UART, and others use FTD or similar).
That board has a RTS pin, which I connect to the Maple reset pin.
Each time I upload program to the board, I close the port, and re-open it as soon as I see that the IDE starts to search for the DFU.
This generates a pulse that resets the board and help loading the program.
The drawback is that each time I open the port, the board will reset, but I can live with that.
One more thing you should try -
In the IDE 'File', 'Preferences' change (extend) the 'Delay after reset' parameter. This will give you larger time latitude in the reset.
Good luck
samtal

Posted 3 years ago #

RSS feed for this topic

Reply

You must log in to post.