Axon 2 bootloader issues

[scottad73]

I have the Axon 2 MCU and using the AVR Studio 4.  I was able to successfully upload code to the MCU but for an unknown read am no longer able to.  I verified via hyperterminal that the connection was good.  When I initiate the bootloader, then power the MCU the bootloader doesn't come out of the idle state.  The code being sent successfully compiled.
What could be going wrong, hardware or software, and how do I remedy this?

[Admin]

When I initiate the bootloader, then power the MCU the bootloader doesn't come out of the idle state.

You mean the / symbol just keeps spinning after turning on the Axon?

[scottad73]

That is correct.

[Admin]

Try restarting your PC, I find Windows occasionally hangs a port and refuses to reopen it.

If still no luck, I suspect a connection problem with the USB on the Axon. Look to see if the pins are well-soldered or not. I'd actually recommend reheating the pins with a soldering iron to make sure the solder is melted properly.

Check the tx/rx pins with an oscope or a multimeter with a frequency function while attempting to use the bootloader.

[scottad73]

I reheated the pins and that worked!  Thank you!!

[scottad73]

And it stopped working...   the bootloader only worked 3 times.  Went to lunch and when I got back nothing...  I double checked the battery voltage and the USB connection.  I will use an o-scope at the university to test further.

[scottad73]

I am still waiting on a response for this problem, it has been over 10 days.  This is a college project not some hobby that I can wait on.  I am very frustrated.   If we can not resolve this issue I would like to have my money refunded and I will return your product.

[Admin]

Oh, I was waiting for the oscope results you mentioned.

Trace the rx/tx pins, and check on the ATmega640, the small black CP2102 USB chip, and of course the USB connector pins.

[scottad73]

Ok,  I finally got to use the oscope and wouldn't you know it, I got signals from both rx/tx pins.
So I compiled the code again and it uploaded to the MCU just fine.
Hopefully it stays this way.
 

[Admin]

It could be a 'ghost short'. Its when it usually works, and always works when you measure it, but fails under random unknown conditions. The most annoying electrical problem ever to debug . . .

If it happens again, heat up the solder on the ATmega640 pins that link to Rx and Tx and see what happens.

[scottad73]

another, new?, problem with Axon 2.  The computer doesn't see the Axon 2 anymore!  If I plug in the battery the MCU turns on and runs the last loaded program.  I have tested the com port and it is working fine.  I tested the voltage, I no long have access to an o-scope, to the usb pins and they are 1.0V, 0.07V, and 4.76V.  Pins were read from the center pin to the outside pin, ignoring the ground pins.  While testing these pins I noticed that the CP210x USB to UART Bridge is very hot. I don't know if that is normal.  Is there any other diagnostic that I can do?

Thanks

[Admin]

the usb pins and they are 1.0V, 0.07V, and 4.76V.  Pins were read from the center pin to the outside pin, ignoring the ground pins.  While testing these pins I noticed that the CP210x USB to UART Bridge is very hot.

Those voltages are about right, although the 4.76V is a bit too low.

The fact that it gets hot, and the voltage is low, means you have a short bringing massive current through the USB chip. Did you use the 3.3V output or do anything out of the ordinary?

[scottad73]

I did connect the power to the 3.3V output momentarily.  Didn't think that was related to the USB.  Can I fix it anyway?  I've been using the bootloader, so now I'm guessing that I need  to purchase a programmer?

[Admin]

I did connect the power to the 3.3V output momentarily.

Yeap, you fried the USB

You have two options . . . replace the CP2102 chip, a difficult task requiring a rework station. I can help you through it if you choose this option.

Or never using USB again, so you'll need an ISP hardware programmer to program from now on . . .

Don't feel too bad, I think 3 other people also fried the USB with similar mistakes, if I remember right.

[scottad73]

Argh!  I think that purchasing a rework station is out of my budget.  In fact a ISP hardware programmer is as well,  do you know were I might find directions on building my own programmer?

[Admin]

I've never made one, couldn't recommend any . . .


This is cheaper and still fully functional:
http://www.pololu.com/catalog/product/1300

[madsci1016]

I recommend that programmer, it's easier to use then an official Atmel ISPmk2, as i just found out.

[scottad73]

bought the pololu USB AVR programmer and will update after I play with it.