Possible Carrier errata? USB device mode.

Diode D5 on the PicoZed Carrier Card is connected to UDB_VBUS_UART in the USB UART circuit, which is not the USB interface you are attempting to use. D15, which is connected to the USB_VBUS, is a circuit protection device that should not interfere with normal operation.

Please do not make duplicate posts of the same question. I will delete the duplicate post.]

-Gary

Oops. I meant D15..



Yes, I know it's a protection device, but I'm not convinced that it isn't interfering. Electrically, the carrier board is capable of meeting the datasheet specified configurations for host and OTG mode, but it is not for device mode.



Can you point me to any examples of a Picozed + CC setup working in device mode? I'm really really stuck.



Also sorry about the double post... The captcha images and I were not getting along yesterday.

Have a look at this evaluation board that microchip makes for the USB3320.


http://ww1.microchip.com/downloads/en/DeviceDoc/evb3320.pdf

This board is configured "Out of the box" to operate in device mode, and they have it connected just as the datasheet recommends, with 10k ohms of series resistance and a 2.2uF bypass cap on the usb connector side.


Furthermore, if you read page 3 of the user manual here:

http://ww1.microchip.com/downloads/en/DeviceDoc/evb3320user.pdf


It clearly states that, "The USB controller must detect VBUS when a USB cable is attached in device mode"


D15 on the Carrier Card will prevent VBUS detection and prevent the phy from operating properly in device mode.

Hi Troy,

I do not think that D15 should prevent VBUS detection, but you could always short the device to test the theory.

I am still investigating, but I think the more likely source of your issue is that the USB_OTG_CPEN signal on the PZCC FMC Carrier card, driven by the CPEN pin of the 3320 on the PicoZed SOM, is being asserted. This will cause U15 to drive USB_VBUS, as it should in Host or OTG mode which is the default for the PicoZed. I suspect that your PetaLinux driver will need to write to the 3320 to de-assert CPEN for USB device mode to operate correctly.

-Gary

trjeppsdl,

The Symbol on the schematic might be a little confusing, but this is not a DIODE in the traditional sense.  It is a voltage protection circuit.  The diode is there as a voltage sense / protection gate.

The voltage from pin 1 to pin 3 is free to flow (0.16ohm typical R).  However, if the voltage from pins 1 to 3 becomes larger than 5.45V (min) - 5.75V (max) the Zener component will "short" to ground up to it's MAX rating (~10A until the part is damaged).  This ensures that if there is a transient or voltage spike (typically tested using a human body model http://en.wikipedia.org/wiki/Human-body_model) the sensitive 5V circuits are not damaged.  This can be typical depending on the environment when plugging and unplugging cables.  If you look at the end of the USB cables, you can even see that the outside pins (power and ground) are longer.  They are designed that way so as to engage the protection circuit before the data lines are even connected.

If you check the datasheet (http://www.te.com/commerce/DocumentDelivery/DDEController?Action=showdoc&DocId=Specification+Or+Standard%7FSCD26730%7FF%7Fpdf%7FEnglish%7FENG_SS_SCD26730_F.pdf%7FRF1489-000) for the part, you can see what I mean on bottom of page 2.

From the sounds of your issue, I agree with Gary that you probably do not have the CPEN pin configured properly.

--Dan