Open-source USB interface: Audio Widget

Well Börge - not a major problm though - I googled up the elf in less than a few seconds. This is what I get trying to load it:

C:\Program Files\Atmel\Flip 3.4.3\bin>prog d:\projects\audio-widget-nik-20111023
.elf
C:\Program Files\Atmel\Flip 3.4.3\bin>batchisp -device at32uc3a3256 -hardware us
b -operation erase f memory flash blankcheck loadbuffer d:\projects\audio-widget
-nik-20111023.elf program verify start reset 0
Running batchisp 1.2.5 on Sat Nov 12 12:50:40 2011

AT32UC3A3256 - USB - USB/DFU

Device selection....................... PASS
Hardware selection..................... PASS
Opening port........................... PASS
Reading Bootloader version............. PASS 1.0.3
Erasing................................ FAIL Could not read from USB device.
(A)bort, (R)etry, (I)gnore ? r
Erasing................................ FAIL Could not read from USB device.
(A)bort, (R)etry, (I)gnore ? i
Selecting FLASH........................ PASS
Blank checking......................... FAIL Could not read from USB device.
(A)bort, (R)etry, (I)gnore ? i
Parsing ELF file....................... PASS d:\projects\audio-widget-nik-201
11023.elf
WARNING: The user program and the bootloader overlap!
Programming memory..................... FAIL Could not read from USB device.
(A)bort, (R)etry, (I)gnore ? i
Verifying memory....................... FAIL Could not read from USB device.
(A)bort, (R)etry, (I)gnore ? i
Starting Application................... PASS RESET 0
Summary: Total 12 Passed 7 Failed 5
C:\Program Files\Atmel\Flip 3.4.3\bin>

---

Now this is how I tried (maybe there are some steps I am not aware of...)

1: Started up the pc with the widget hanging on USB
2: Searched fir the elf and googled to find it - downloaded it
3: Hold the progbutton whilst pressing reset and then released it
4: started CMD and fired the prog with file and location as in the text above

Brds
 
Last edited:
Hi Turbon,

something is wrong in your system which doesn't depend on which .elf you're using. I have seen similar errors only twice before.

Once was when the ESD diodes at the USB data pins were not assembled. This happened on a single board of the very first batch of AB-1.1 boards as they came from the manufacturer. The board worked for a while and then the USB died. (It so happens that this is the board on the AB-1.1 homepage...)

The other time it happened I couldn't erase. I saw no obvious reason why, and put the board aside, since I had a few more to pick from.

Reinstalling the bootloader from avr32studio, and various other light debugging, with the JTAG ICE mkII did NOT solve my problems back when I had them.

I believe your problem is on the device side, since I used the very same PC setup (without reboot) to program other devices after the errors occured.


Børge
 
The ESD diodes you say Börge...

I had no idea that they where diodes :) OK, one of them at least seems to be a gonner - could I substitute them with whatever diodes as a test?

I will give you a picture when I figure out how to do it Börge... It might be educational - if this works - everything should ;-)
 
Last edited:
Yes - more happy soldering then ;-)

I tried to load the blinky flashy elf again that worked yesterday without luck so I'll start chasing esd diodes as a start - mcu - no problem...

I had a few peeps at the shematics but couln't find the ESD diodes...

I'll have a look at the bags provided by George - maybe I'll find a partnumber or any usable reference there.

This will work in due time - I'm sure about that.

This seems to be what I need:
http://www.google.com/url?sa=t&rct=j&q=pgb1010603mr&source=web&cd=1&ved=0CCcQFjAA&url=http%3A%2F%2Fwww.littelfuse.com%2Fdata%2Fen%2FData_Sheets%2FPGB106032.pdf&ei=aHy-Tv7gEayQ4gSzpIChBA&usg=AFQjCNHZffRolS7BtanawEYi141BnAau0A
 
Last edited:
Now that seemed as magic to me Börge! Thanks - but I will probably just dremmel away the pins and tidy up with my soldering station. I always check afterwards at 10x so that everything looks nice. I will this time add some lead to all pads before putting a new mcu in place (if needed...)

About the ESD diodes - are they polarized? The symbol seems to have 2 anodes...

About the MCU - Mouser gave these choices - which one do I need?:
http://se.mouser.com/_/?Keyword=at32uc3a3256&FS=True&bdprt=1&cartid=26255041


Brgds
 
Last edited:
:) another one - I feel chatty today...

---

Now if I am brave enough and ain't worried about the mcu anymore I could theoretically do without the ESD diodes - their purpose is to protect the mcu from overvoltage - right?

Another question - I have seen different references from pictures, schemas and text about how the jumpers should be connected. I find it a bit confusing and would like to see a settlement on this.

Brgds
 
Last edited:
Hi Turbon,

I totally understand that you're frustrated. I solder these things with a dedicated soldering microscope and Metcal irons. For non-destructive removal I use the ChipQuick metal. Beware that it needs extensive metal cleanup, should you choose to use it. The AB-1.1 batch was stencil printed and reflow soldered by the manufacturer.

George and I both want to sell our stuff. The main difference (at present) is that George sells kits and I sell assembled+tested units. That's why my stuff is more expensive. And it's also why George can get naked boards out faster and in smaller quantities.

I still have a pile of AB-1.1s here, and I'm happy to sell them for $120+shipping to people who prefer to put their soldering hours into modding analog filters and PSU, not TQPF144's.

The problems you are having are rare, but unfortunately not unique. The only fix I know is to replace the MCU. With the AB-1.1 it is also possible to slam in a new module and keep your analog mods.

It will be interesting to hear what Atmel has to say about possible causes, so please ask them and fill us in. I'm traveling this week, so my bandwidth is somewhat reduced.


Børge
 
For playback you need one OUT and one IN for rate feedback. As we are also interested in capture, another IN endpoint. So we have been looking for devices with 3 ISO EP's.

Alex

Have you looked at the ST Cortex-M3 or M4? I haven't worked with it myself, but the datasheet looks interesting. It has USB FS and HS that supports 3 or 5 IN+OUT endpoints respectively. The M4 runs at 168MHz too so should have some horsepower to spare for DSP duty perhaps? Comes in a nice 64-LQFP package too.

STM32F405RG - STMicroelectronics
http://www.st.com/internet/com/TECH...AL_LITERATURE/REFERENCE_MANUAL/DM00031020.pdf

OJG
 
A quick update on my post #324.

I have now replaced the original LDO feeds to the dac and XO with Lifepo4 batteries.

For inquisitive minds, the lifepos are connected do a DPDT switch that connects to/disconnectes from the circuit to charg them in parallel. There is a shottky diode between the charger and the battery powering the XO to make sure that even on a full charge the battery voltage does not exceed 3.3V + 5%. (Thanks John Kenny and Dan Baldwin for the idea)

The sound is substantially improved in a number of respects. Better decay of notes, more realistic brushwork on cymbals and skins. Yet, I have the impression that I have lost some dynamics and bass extension relative to the LDO supply. In fact, I think I have to increase the volume 4db relative to the previous iteration. I am not 100% of this last point, but A/Bing is difficult...

The batteries feed the circuit directly with no additional capacitor beyond the original one. I wonder if putting some 460uF at the point where the lifepo feeding the DAC connects to the board would help with dynamics.
 
Alright, answers seems to be scarse in here - bad sales possibilities on this one perhaps - everything sold out?... Right - I take a photo of the thingie and ask Atmel then.

Brgds
Turbon,

I am sorry I did not participate in this tread I was unaware you were having problems.

ESD: protection against hi voltage power surges from the PC. These can come from defective computer supplies or static discharge. There is only a small risk if you remove both devices.

CPU: 556-AT32UC3A3256ALUT

Solder: 0.020 no clean, rosin core, (lead free for ROHS)
 
Nah, never mind George - sorry for the outburst. The problem was a blob of lead under the usb connector - the now non existing ESD diodes were probably allright.
I had to desolder the connector and lost some leads both on the connector and the board in the progress so some wire helped a lot. Now thanks a lot for the part number of the mcu - I believe I need a new one.
About the board - IF one looks from the front of the usb - is from left to right pin no4 connected anywhere? I can't see it is on the board but in the layout picture there seems to be a line. So with pin 4 floating - no esd diodes I have so far got the board to present itself as unknown to my pc. The pc recognice a happening on the usb port when I press the program-reset buttons but it is not recognised as an atmel dfu device.

Brgds and thanks for your input and soothing words Börge and George.

This will eventually work :)