TUSB2046I: Getting "Power Surge" warning when plugging USB hub into laptop (Windows 10)

p.s. - I'm not getting 6MHz when I probe on R6.  Page 11 of the datasheet recommends 1.5K for R6.  Is this correct ?

You may get overcurrrent, can you send schematic for review?

Best

Brian

I included the schematic in my original post.  It's not overcurrent, the crystal is not functioning.  The 6MHz crystal is not oscillating and I suspect that the 1.5K series resistor is the reason why.  Please advise ...

I changed R6 from 1.5K to 100-ohms, and still no oscillation.  I changed R6 to zero-ohms, and no oscillation.  There is 3.3V on VCC.

Still troubleshooting. Please let me know if you have any ideas.

My POR circuit is releasing RESET high 50msec after 3.3V turns on.  Is this too long ?

For some reason, the crystal stopped working (it still works if I hold the device in RESET so I know the crystal isn't damaged).  I tried increasing the POR time from 500usec to 1.1msec by increasing C23 from 10nF to 22nF, and this didn't solve the problem.  What is the POR time range required for this device ?

did you try 1uf or 2uf for C23?

Best

Brian

I tried C23=1uF and R21=15K which resulted in a 50msec delay between VCC=3.3V and RESET going high.  Is 50msec too long ?  Does the device have a window for when the RESET needs to go high and, if so, what is this window ?

According to Joel Jimenez ....  The TUSB2046B requires a power on reset pulse of between 100us to 3ms after 3.3-V (Vcc) reaches its 90%. We recommend an external RC circuit (15K pull-up resistor and 1 uF cap), please use this circuit to try to solve the issue. Anything less than 100 us is too short for the device and anything longer than 3ms can interfere with USB enumeration (mainly an issue for embedded systems).

Is this correct ?  If so, how is this window (100usec - 3msec) being met with 15K/1uF ?  Like I said earlier, I'm getting 50msec with these values.

See (+) TUSB2046 reset condition - Interface forum - Interface - TI E2E support forums

Simulation even predicts around 50msec ....

While waiting for a response, I tried C23=2uF and R21=15K, and the oscillator doesn't work (it still works if I hold reset low).

I just found another piece of information from a different post that may shed some light on the problem

It turns out that pin 32 (SUSPEND) is HIGH on my device, which explains why the crystal isn't oscillating.  Apparently, the USB hub needs to be connected to a host in order for the crystal to turn on, but when I connect my USB hub to a host, I get the "Power Surge" warning, so I need to figure out how to achieve a successful connection between my laptop and my USB hub.

Any recommendations?

sorry, I didn't realize your upstream port was not connected.

Can you send full schematic to me for review?

Best

Brian

Brian,

The schematic is in my original post.  I removed R74 & R76, got a USB cable, cut off the connector at one end of the cable, spliced the wires, soldered the green wire to U9 pin 1 and the white wire to U9 pin 2 and the black wire to GND.  I left the red wire disconnected because the USB hub and all of the downstream peripherals are self-powered.  I plugged the USB-A connector on the other end of the USB cable into a laptop (so, yes, my upstream port is connected).

When I plug the USB cable into my laptop, I can see that the Generic USB Hub pops up in the Device Manager (which tells me that at least part of it is working), but I also get a pop-up stating that there is a Power surge on the USB port.  See image below :

When I plug the USB cable into my laptop, I expect to measure the 6MHz crystal oscillating on my oscilloscope (it doesn't).  Shouldn't the 6MHz clock be oscillating?

Finally, one question that I haven't received an answer yet is ....

Does the device have a window for when the RESET needs to go high and, if so, what is this window (i.e. min/max POR delay time) ?

Thanks,

~Allen

Allen:

  Normal condition, OVERCURR# should be high. since these pins are grounded in the design, it means OVERCURR event is happened.

Best

Brian

Good catch Brian !  I lifted the OVRCUR# pins and wired all of them to 3.3V, and the "Power Surge" error goes away.

However, I'm still expecting the 6MHz crystal to oscillate but it still only oscillates when I pull RESET low.  Are there any other conditions that are required for the 6MHz oscillator to turn on ?  If there is no USB peripheral connected downstream, would this disable the 6MHz oscillator ?

I thought if you connect upstream port to host,  6MHz oscillator will work, it's not right?

Best

Brian

any update?

I have a catch-22 problem.  I don't have any functioning downstream USB ports at this time (separate troubleshooting tasks ahead).  I can successfully make an upstream USB connection (Generic USB Hub shows up in Device Manager), but the 6MHz oscillator doesn't come on.  However, if I force RESET low, the oscillator runs (so I at least know that the oscillator isn't broken).  I need to determine if the oscillator isn't running because there are no downstream USB ports.

If you have an eval board (I don't), could you test the 6MHz oscillator with an upstream port connected and no downstream ports and let me know if the oscillator is active or not ?

Thanks.

let me check if I have TUSB2046 EVM.

Best

Brian

Any luck finding an eval board ?

BTW - I never got an answer to this question ...

Does the device have a window for when the RESET needs to go high and, if so, what is this window (i.e. min/max POR delay time) ?

If you could find out the answer and share, it will allow me to finalize my RESET circuit.

Thanks,

~Allen

max reset window is around 1ms.

Best

Brian

What is the minimum power-on reset time ?

What is the maximum power-on reset time ?

I understand the "window" to be the difference between the max & min power-on reset times (so I'm not sure what you mean by "max reset window").

Please clarify ...

Thanks,

~Allen

Hi Allen:

  it's not really a window, 

  If Reset keep low at 1ms during power up , it should work. This seems min  reset time,  not  max reset time.

Best

Brian

Okay, so just to confirm, I can make RESET go high any time after 1msec and the device will work, correct ?   That is, 50msec won't be a problem, 1second won't be a problem, 10 seconds won't be a problem, etc.).  Is this true ?

well, over 1sec is too long.

Best

Brian