Powering the EvalBot via the ICDI USB port [ISSUES

[Chris Muncy]

From the e2e TI forums at e2e.ti.com/support/microcontrollers/stellaris_arm_cortex-m3_microcontroller/f/471/p/66996/247240.aspx?PageIndex=2

Hi all,

having assembled the evalbot and after a few moments of fun I was trying to get it connected to my PC.
I used the USB cable and the ICDI plug on the board.
As a first step I managed to retrieve memory entries via the uC/Probe, so everything worked fine as long as the batteries were still plugged in.
Without the batteries there was no life sign at all.
During the root cause analysis I realized that the board was assembled in a way not to allow powering by the USB cable plugged into ICDI plug although described otherwise in the docs.

D6 is not assembled and D7 is replaced by an 0R resistor.

What is the reason for this change?
Would you expect a problem with the voltage level if I did assemble D6 and D7 instead of the 0R as originally intended?

Btw the figure B-1 EVALBOT Component Location on page 23 of the user's manual (EKB-UCOS3-EVM-UM-00) does not fit to the schematic in the same doc and does obviously not reflect the actual board version.

Thanks
Remo

Hi Remo,

Unfortunately, you have discovered an error in our build process.

The way the board is currently shipped, users can use a USB cable to power the board only through the MCU's USB connector.

If this operation is unacceptable, you can move the 0-ohm resistor from the D7 to the D6 position. The board can then be powered from the ICDI USB cable.
The diodes (D6,D7) are not installed because the voltage drop when the motors start into a load can be enough to sag USB VBUS (to U3 FT2232) out of spec and the debug connection will be lost.
Thanks for pointing this out so that we can get it corrected.
Regards,
Sue

Hi Sue,

thanks for the very quick response.
I appreciate your frank reply.
Its not a problem for me at all, optionally I could follow either your proposal to switch the resistor to D6 position
or use on the D6/D7 location a pinheader with jumper for manual selection of power source.

Br
Remo

So keep that in mind. As the evalbot currently ships, you *must* have batteries in it in order to talk to it via the ICDI port unless one of the above mention work arounds is performed.

[jackthevendicator]

Oct 25, 2010 16:00:10 GMT -5 Chris Muncy said:

From the
So keep that in mind. As the evalbot currently ships, you *must* have batteries in it in order to talk to it via the ICDI port unless one of the above mention work arounds is performed.

Maybe an high current low drop schottky diode (like those in switching power supplies) could do the trick?

[jwhildeb]

My bot, which arrived earlier this week, has the resistor in the correct place (D6) and works fine on ICDI USB power. Guess they sorted things out

[chemuduguntar]

I ended up using two USB cables, one plugged into the ICD port and the other into the USB Device port.

[yaleman]

If you have one without the fix and don't have a use for the USB Host port, you can plug a cable into the ICDI port, and one from the usb host to the usb device port, this seems to make it work.

[senatorpenguin]

Jan 29, 2011 1:11:00 GMT -5 yaleman said:

If you have one without the fix and don't have a use for the USB Host port, you can plug a cable into the ICDI port, and one from the usb host to the usb device port, this seems to make it work.

That shouldn't work, you probably either have a fixed version where nothing is necessary, or you're simply running off of battery power. I don't have mine in front of me, but without the fix, it looks like the 5V host power will attempt to power the board through the power multiplexer (which picks the higher voltage between USB and the 4.5V batteries), fail, and switch back to the batteries as fast as possible. If you haven't enabled the USB host power supply in software with the EN_VBUS pin, nothing will happen at all.