TUSB2046B: TUSB ENUMERATION IS NOT HAPPENING REGULARLY AND NOT GETTING THE OSCILLATOR OUTPUT ALSO

Hi,

We are using the USB HUB controller"TUSB2046BIRHBR" in our design,i am getting below observations.

1.USB Enumeration is not happening every power on time properly.

2.some of the boards USB enumeration is happening initial power on only,but later USB HUB down stream ports are not getting detected and driver also not loaded on linux prompt.

3.Some of the boards i am getting the crystal output during power on,later after multiple times power and off then i an not getting crystal output correctly and also usb enumeration not happening every time properly.

4.Here i am attaching our designed schematics for reference.

5.Right now i have shorted all over current(oc) pins,all down stream ports are activated at a time.

6.could you please check and suggest what might be missing and why i am facing this issue,some boards are working and some boards are not working with same circuitry.

Thanks & Regards

K.Manikumar

Hello,

It looks like the downstream ports are missing the 15K pulldowns, this may be causing the detection issues.

Can you confirm if when the clock stops if the TUSB2046B is in Suspend mode by checking suspend status on pin 32?

Regards,

JMMN

Downstream ports are pulled high with 10k in current design,may i know whether downstream ports are to be pulldowns or pull up??

Intermittently clock is coming some times and some times not coming, means TI usb hub driver not loaded by linux prompt.

could you please suggest the root cause for this.

Thanks & regards

K.Manikumar

Hi,

We are facing an issue with TUSB2046BIRHBR device:

1.With clock out and with out clock output at USB HUB, when i try to measure suspend pin it is always low.

2.Currently,all downstream ports are pulled high and all over current pins are shorted.(ganged mode).

3.In some of the boards,USB hub is working fine and in some boards initial power on USB hub enumeration happening properly,but later after second time power on,USB hub downstream ports are not getting detected, even TUSB2046 driver not getting loaded in linux prompt and clock output also not coming.

4.Please let us know how to fix this issue.

Thanks & regards

K.Manikumar

The downstream ports of the hub (DP/DM) require 15K pulldowns, they are not in the current design.

Please make sure that the OVERCURz pins are high.

If the hub has a thermal pad, please make sure it is connected to ground with enough solder.

The part number of the clock source in the schematic is for a resonator, not a crystal.  The TUSB2046B is designed to work with a 6 MHz crystal or 48 MHz clock source.

Regards,

JMMN

Hi,

Could you please suggest the exact 6MHz crystal part number/make for this TUSB2046 for self powered operation...Above 6MHz resonator part No(CSTCR6M00G55B-R0/Murata) we kept by taking from TI reference design circuit only..

Thanks & Regards

K.Manikumar

Hello,

Please see section 8.3.2 of the datasheet for information on acceptable crystals: 

https://www.ti.com/lit/ds/symlink/tusb2046i.pdf

I do not see any references to the Murata resonator in any of the TUSB2046B documents, please let me know where it is located so it can be corrected.

Regards,

JMMN

Hi,

Surprisingly it worked with murata 6mhz resonator in most of the boards,only few boards i could not see any usb hub controller TUSB2046 getting detection in linux.

may i know what might be the issue.

Thanks & regards

K.Manikumar

Hi,

In EVM430-FR6047 reference schematics,6MHz used is resonator from murata..is there any impact by using 6Mhz resonator instead of crystal.

Thanks & Regards

K.Manikumar

Hi,

1.We tried by putting 15K pull downs at downstream ports(DM/DP) lines,still the behaviour is same,means USB HUB not getting enumerated

2.Somes times if i try to measure the VBUS voltages at downstream ports connected to USB connector,i can measure only 2.5V not more than that,but it should be 5v in working boards/working behaviour.

may i know what could be root cause for this issue.

Thanks & regards

K.Manikumar

What percentage of the boards are not working?

If there is not enough voltage at VBUS, there is likely an issue with the TPS2042 device, the TUSB2046B does not handle the VBUS directly.

Can you confirm if on the non-working boards does the TUSB2046B stay connected?  Is it just the downstream devices that do not work?

That particular design was not done by our team and a resonator doesn't match our hub requirements but USB Full Speed is pretty forgiving with timing so the resonator may function well enough.

Regards,

JMMN

Hi,

Out of 12 boards,

1.6 Boards are working very well and 03 boards are working when we try to connect downstream ports then only we are seeing TI TUSB2046 driver in linux

2.03 boards are worked initial power on time,later when we try to check multiple times,then we are facing issue.TUSB2046 driver loading and USB hub down stream ports also not getting detected.>

3.In the non working boards,when i try to measure VBUs voltage for hub getting detected,i am getting VBUS voltage 5V.but i couldn not see TUSB2046 detection in linux prompt.

4.We did the X ray for both working and non working,but both are same,i have not seen any difference.

May i know what might be wrong

Thanks & regards

K.Manikumar

Hi,

1.We have connected down stream ports as per below.

2.We have tested by adding 15k pull downs at DM/DP lines of down stream ports,but the behaviour is same and shorted all over current pins and pulled to high.

3.What might be missing here and could you pls suggest anything we are missing here.

Thanks & Regards

K.Manikumar

Can you check that the 3.3V to the hub is stable, and when the hub is connected you should be able to see DP go high and then low when it receives a reset for the USB host.  Also, please check that the clock starts.  If DP never goes high or the clock never starts, it could be an assembly issue.

Regards,

JMMN

Hi,

We have replaced new USB hub in one of the board,we have measured voltages also

1.We are getting 3.3V continuously to USB hub

2.DP line is pulled to 3.3V and we are seeing only 1.5V to USB hub controller and clock also not started.

3.Is there any 15k pull down resistors are required on host side DP & DM lines(up stream usb)??

4.What could be the reasons that we may not seeing any TI driver and down stream ports also in the linux???

This type of behaviour we are seeing only 3 boards out of 12 boards and surprisingly during initial power on time these 03 boards worked very well,but on the second time power on onwards USB hub was not detected and clock also not started??

May i know what might be the reasons??

Thanks & regards

K.Manikumar

Hi,

We have observed following during observations,need your help on this.

Case 1:When all down stream ports connected(All over current pins are shorted together)

         i.Out of 10 times power on/off,we have observed 2 times USB Hub controller enumeration not happened.

        ii.Measured voltage at DP line of up stream during un enumerated time,there we are getting 0.025V,but during enumeration time,we are getting 3.00V

Case 2:When connected only 02 down stream ports,one down stream not connected means removed connection:

           i.Out of 10 times power on/off,we have observed 02 times USB hub enumeration problem.

Case 3:Nothing connected to USB down stream ports

We have observed, only initial power on time TUSB2046 driver getting detected on the linux prompt.

But in all above cases only during initial time only we have observed TUSB2046 driver on linux prompt,next power on time we have not observed TUSB2046 driver.

We have connected ganged port as per below.

USB HUB supply is stable always and why we are observing DP line low for the next power on time??

If we want to connect all the down stream ports as per port basis and all down stream ports are getting detected by linux,what are the changes we need to incorporate in out design??

What might be the reasons for USB hub getting enumerated for some time and not enumerated for some times??

Could you please help us on this??

Thanks & regards

K.Manikumar

1. 3.3V is power to the hub?

2. Where is the 1.5V measured?

3. Most USB hosts integrated the 15K pulldowns - check with the vendor.

4. Does the hub respond after a reset (toggle reset pin to the hub low then high again)?

It seems like the hub is going to suspend and not exiting, can you check the level on the SUSPEND output.

Hi,

1.Yes 3.3v is power to USB Hub controller.

2.During 10 power cycles,we have observed that 2 times USB HUB DP line voltage is 0.025V only,so at that time USB enumeration is not happening.

3.Suspend status is always low only.

4.Hub reset is not toggling only 2 times out of 10 times.

5.Is all the down stream ports connected all together when all OC pins were shorted??When we connected all downstream ports together then only we are seeing all down stream ports related driver info on the linux prompt.

6.But with out connecting all down stream ports together,how we need to connect the downstream for getting detected??

could you pls tell us the reason for the above mis behaviour and now we have shorted all oc pins of USB hub.means10,14,18,22 are shorted,we are seeing all down stream ports only when all down stream connected.

May i know the reason for this issue??

Thanks & regards

K.Manikumar

Hi,

As you know,we are using TUSB2046BIRHBR(USB hub controller) from TI

Continuation to the above chat discussion,we are facing following issues.

1.Out of 10 times power ON/OFF,8 times we are seeing TUSB2046 driver in linux and downstream ports we are seeing when we connected all at a time.

Remaining 2 times we are not seeing TUSB2046 and downstream ports also,DP line voltage measured at that time is 0.0.25V only,but in working condition we have measured 3.01V

2.Could you please confirm and tell us the reasons for the USB hub not getting enumerated for remaining 2 times??Is there any particular reason that USB Hub(TUSB2046BIRHBR) not getting detected in linux/DP line voltage getting low??

Thanks & Regards

K.Manikumar

When all OVERCUR are shorted together - the OVERCUR are all high, correct?

How long of a power on reset pulse is there on the RESET/ pin?

Which Linux build are you using?  Have you tried connecting another USB hub?

Regards,

JMMN

Hi,

Yes all over current pins are high,

Linux build we are using is 4.19.

When i try to probe reset pin,it is always high from power ON time..

Yes we have replaced with new hub in one of the board,but we are seeing 8 times better enumeration and 2 times we are not seeing enumeration and at that time DP voltage also 0.025V instead of >3V.

What could be the reasons for the hub not working in remaining 2 times & why we need to connect all down stream ports together to see in linux??

Thanks & Regards

K.Manikumar

Hello,

The hub must have a power on reset.  Reset should be low for at least 100us to 1ms after the hub is powered.  Can you add a capacitor to the pin to provide this reset and confirm if it improves the behavior?

Regards,

JMMN

Hi,

We have added 0.1uF capacitor from USB_HUB_RESET pin to ground(from resistor end to Ground), but there is no change in behaviour.(means 2 to 3 times we are getting enumeration problem out of 10 times)

Why we are getting usb hub enumeration problem during power cycles,could you pls help us on this??

Thanks & Regards

K.Manikumar

Can you provide scope plots of DP and DM on the upstream port of the hub when enumeration fails?