TUSB2077 USB Hub, Crystal only runs in Reset

Hello Chuck,

Can you share your schematics.

How long the RESET signal is low after the 3.3-V rail reaches its 90%?

Regards.

Hello Elias,

Yes I can send you a copy of the schematic. I don't see how to add it here in the text box. I will need to get a picture of the reset tomorrow. I put the schematic in my profile file folder.

Chuck Still

Hello Elias,

I have uploaded the schematic and some pictures to my profile files tab. Not sure if that is the best place to put them. Can you access them there? Or should I put them someplace else? I have been working on this but still no resolution. The three pictures are; First is the 3.3V power  (top trace) and reset (bottom trace). The other two are the 3.3v power rail and the xtal pin 45. I have two chips on the card one works and one is failing to keep the crystal running. The failing trace shows that the crystal starts up but then abruptly stops 16msec after it started.

Chuck Still

Chuck,

We can access the files there.

In the future when you are composing a post, you should have three "tabs" (Compose  |  Options  |  Preview) above the post you are replying to:

This is not that intuitive, but if you click on the "Options" tab, you should be able to attach files to a specific thread.

Hello Chuck,

Which of the two hub is the non-working?

Do you have pull-ups on every OVERCUR# input?

IS the SUSPEND output active of the non-working hub?

Regards.

Hello Elias,

The working chip is marked U1200  and the non-working chip is U1201. Yes we have pull-ups on all the OVERCUR# pins. 3.01K ohms, one resistor for each pin. I have attached the schematic page with these and the current limit chips.

The SUSPEND output is active high on the non-working chip. Also the PORTDIS output is active low on the chip as well. These signals are in the opposite states on U1200 (the working chip).

What would put the TUSB2077 in a suspend state 100msec after power-on? I have a scope trace that shows the SUSPEND changing from low to high just over a 100msec after power is applied.

Thank you for your help,

Chuck

Chuck,

Your USB host is not enumerating the TUSB2077 (the non-working one).

Check the  serial impedance to be 33 ohm on the DP/DM lines.

Place a 22pF caps to ground on the DP/DM lines.

Double-check that the RESET# signal is low min 1ms and max 3 ms after the power rails reach its 90%.

Are the downstream devices on the TUSB2077 powered-up before the hub is powered?

Regards.

Elias,

I have checked the 33 ohm serial termination resistors and they are correct. I have not placed 22pf caps on the DP/DM lines for the host port connection on the TUSB2077 yet. It was my impression from the USB documents that the caps and 15K ohm pull down resistors were to be on the host end of the USB connection. I switched circuit boards as a check that I was not chasing some other problem. The reset on this board is correctly delaying 2msec after power is up above 3v. This USB hub and all downsteam endpoint devices power up at the same time. They are all on the same power source. Tthe ITX board is also on the same power supply  that is the host end of the USB connection.

I see that if I ground the DP line (pin 3) the TUSB2077 comes out of suspend and starts to run. I don't see any USB traffic on the DP/DM pins when the chip is out of suspend. Not sure what the protocol is with the host system to establish a connection.

Chuck Still

Chuck,

Have you discarded ESD damage and PCB quality issues?

If you have 70 boards which never fails, we should assume your design is correct.

Have you tested your passing units several times?

Regards.