TUSB2077A: TUSB2077A show only PORTDIS

Hello Bawornsit,

What is the state of the UP_DM / UP_DP lines?  Can you confirm that voltage levels and clock to the hub.  Please provide the schematic if possible.

Regards,

JMMN

This is external power reduced to 5V.

This is LDO regulator connected from +5V USB upstream power

I connected +5V_BUSIN0 from USB_upstream power  and +5V_BUS from 5V regulator  together and provided to Power distribution switch (TPS2044).

There are a diode on the +5V_UPBUS to prevent reverse power from 5V regulator that shown on the next picture.

Hi,

I can't tell from the schemaitcs, but please confirm that DP at the hub is always connected to DP, at the connector and DM at the hub to DM at the connector as well.  The downstream facing ports need 15K pulldowns, 1.5K is too strong.  The pulldowns should be on the hub side of the 27 ohm series resistor.  Also, I would try running the hub without the EEPROM to start.

Regards,

JMMN

DP and DM upstream are connected to chip pin as well with 27.2 Ohm and DP0PUR is connected as well with 1.5K Ohm.

For downstream ports, this is my fault. however, I will change to 15K later.

Can you check the line states of upstream DP and DM?  What does the host report?  What operating system are you running on the host?

Regards,

JMMN

My host is running on windows 10. I see. sometimes, it show device error or not recognized. Mostly, not recognized. I tested the cable connectivity is going well.

Only PORTDIS is turning on. this mean the chip works properly? but, it can't connect to my PC only.

The issue is not clear to understand for me.

Can you install usbview.exe or USB device tree viewer and share the report?

This is a report after I connect the hub to my PC. I have compared to before connecting. They are the same.

Please check the line states of the upstream DP and DM, and confirm if the clock is running.

Regards,

JMMN

I use 48Mhz oscillator to drive the chip and I checked it with oscilloscope that show 48Mhz as well. DM and DP running approx at 8Mhz at Vpp=270mV.

Sometimes, rarely,  I see the USB device detected in Device Manager with request failed message.

I found some thing. PUR pin on the chip did't pull up with 3.3 volt that is why my PC can't detect TUSB2077A. I active high to BUSPWR pin therefore, I need to connect 3.3V bus to pull up.

Now, my PC detect the chip but, "Device Descriptor Request Failed" message is remain appeared.

The PUR should go high even in BUSPWR mode as long as power is provided.  I noticed that the clock input to the hub had very low Vpp, can you confirm the loading on the circuit is correct?

Regards,

JMMN

I tested the clock it's defected. I replace a new one then it works. Unfortunately, I got suspended state instead.

I tried to Disable/Enable in Device Manager then PORTCFG is now on but, PORTDIS is still on and no power on all downstream port.

Ok, good news on getting the clock issue resolved.  It is normal for a USB host to put a hub in suspend if nothing is connected.  Can you confirm if connecting anything to a downstream port helps?  Also, please send the usbview or usb device tree viewer screenshot to help determine the port states.

Regards,

JMMN

I just do an experiment by removing the eeprom. Connect EXTMEM and GANGED to 3.3V. It sill failed to enumeration.

The PORTDIS show for a second after connect upstream then off and SUSPENDED light on instead.

Please put a scope on the upstream DP / DM lines and confirm if they are toggling.  You can trigger on DP going high when power is applied.

Regards,

JMMN

I probe the D+ D- signal then I capture this during I am plugging in the usb cable to my PC. I think this look normal.

Signal levels look correct, please do a single capture on DP going high and zoom out to the 100 ms range.

In this test, I supply the chip with external source. Immediately I plug the cable, the voltage little bit drop that is where the trigger point.

I shifted more to the left for more detail.

Ok, so purple is DP and blue is DM, correct?

It looks like DP is high before the cable is connected, this can make it suspend.  Can you power on the board / connect the cable and toggle reset and see if enumeration completes as expected?

Regards,

JMMN

I have tried to use transistor to switch on when usb connected and see what happen on oscilloscope. It look the same as the last picture, 3.3V pullup before the signal coming.

then I tried to connect usb and wait for finishing initialization then pull 3.3V off and put back again instead so this is what happen.

My PC still can't enumerate.

Everything looks clean coming from the USB host, I cannot see anything that would prevent the USB hub from responding.  Can you ohm out the connections from the host to the hub to make sure everything is connected as expected?  And confirm that the clock is running on the hub? 

Regards,

JMMN

Thank you for every reply JMMN. Now my usb hub work properly. I just do a mandatory check all pull-up resistor.

1. pull-up 3V3 with 1.5K resistor on D+ upstream.

2. pull-down 15k all downstream port.

3. pull-up 3V3 every OCR with 10K or 15K resistor

4. check every pin on the chip are not accidentally connect each other.