Some questions about TUSB9260.

I'm trying to load firmware with FlashBurner. It says programming succeeded.

But there is no changes. The FlashBurner window always looks like below:

What must I do to make it work right?

Hello Roman,

      The TUSB926x utility will display the “Programming Succeeded” message after the TUSB926x device has acknowledged the reception of the firmware date; anyhow, the device has no way to determine whether or not that data was properly burned into the externally attached Flash memory. Because of this I would assume the issue will be either on the communication between the TUSB926x device and the Flash memory, or in the Flash memory itself.

•  Can you please provide us your implementation’s schematic?
•  What’s the brand and model number of the Flash device you are using?

 Regards,

Jorge

Hi

The Flash was SST25VF010A from Microchip Technology. Today I bought AT25FS010N from Atmel, which was recommended in TUSB9260 AppNote. But the problem still not solved. The device has 3 same blocks with TUSB9260 and they all not working.

 schematic:

4075.bridge_sch.pdf

Roman,

                 Everything looks fine on your schematic and now you are using the recommended Flash memory (supporting the necessary opcodes) the programming process should be working.

                 Using the same FlashBurner and firmware versions you are using I have been able to successfully program the Flash memory at my end by using our EVM board, so I’m pretty sure this is an implementation issue.

                 Can you please verify you have signal activity on the SI(Pin 5) and SCK(Pin 6) terminals at your Flash memory while trying to burn the FW?

                 Regards,

                Jorge.

Yes, there is an activity on SI, SO and SCK pins.

Roman,

                Can you please also verify your reset timing? I would like to make sure both the 3.3 V and 1.1 V supplies are up before the reset is de-asserted.

                Regards,

                Jorge.

Hi all,

Do you resolved this problem?

Because I have the same problem. I can't program the EEPROM correctly.

Thanks and have a good day.

Hello Matthieu,

Could you please provide more details about the issue you are experiencing?

Are you using our EVM or your own implementation?

What’s the EEPROM brand and model number you are using?

Which OS are you using?

Are you also getting the programming succeeded message?

Is there signal activity on the SI and SCK terminals at your Flash memory while trying to burn the FW?

Regards,

Jorge.

Hello,

First, please note that I have open another post for my problem so maybe I can close it to avoid double post (http://e2e.ti.com/support/interface/digital_interface/f/130/t/113825.aspx).

For my issue, I use my own implementation with the schematic joined on the other post.

EEPROM is an ST M25P10.

OS is Windows XP.

And yes I have the programming succeeded message. But for programming the flash, I open a switch on Data Out (like your jumper on EVM) of the flash to see the TUSB9260 on the flash burner tool and after I need to close the switch and launch the programming. If not, it doesn't work but I think it's normal ?

For the activity on the Flash, I'll check it tomorrow.

Many thanks

Best Regards

Hello,

We have activity on the flash.

The signals CS,SPI-Din,SPI-Dout and SPI-clk seems to be correct.

After programming, the device is recognized by the OS but not installed correctly.

Thanks,

Best regards

Hello,

When the EEPROM is programmed with firmware 0_97, the Led "nFirm_OK" is blinking and the Led "nHS_FS_Connect" is ON.

The led "nHDD_Act" blink but very slowly.

I have changed the EEPROM : New M25P05A.

The problem is not resolved. Do you have any ideas?

Thanks,

Best regards

Hello Matthieu,

Could you please provide more details about what you mean when saying the device is recognized by the OS but not installed correctly?

Regards,

Jorge

Hello,

When I plug the USB cable on my PC, a message appears "new device detected USB mass storage" but after I have the message "an error occurs during the installation of your device" like one USB key which doesn't work..

On the device manager I have this :

 Regards

Matthieu Baque

Well, the firmware seems to have been properly burned into the flash as the LEDs are blinking as expected and Windows has started to enumerate the device.

What's the error code shown by Device Manager? You can get it by double click on the yellow banged device.

Regards,

Jorge

Error code is 10, the device can't start

Regards

Matthieu Baque

Did you customized the VID & PID numbers or any of the descriptors options?

I'm asking because there are some cases where the descriptor options are changed without changing the serial number, and that would result on a Windows registry conflict that won't let Windows to properly enumerate the device.

I would suggest to do the following:

1.- Unplug your device.

2.- Go to regedit and remove any registry keys starting with "VID_0451&PID_9261", or "VID_0451&PID_9260", or "VID_xxxx&PID_yyyy" (being xxxx the VID and yyyy the PID numbers you are using for your device) at "HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Enum\USB".

3.- Remove any registry keys starting with "VID_0451&PID_9261", or "VID_0451&PID_9260", or "VID_xxxx&PID_yyyy" (being xxxx the VID and yyyy the PID numbers you are using for your device) at "HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Enum\HID".

4.- Reboot your system.

5.- Plug the device in again.

Please let me know if that solved the issue.

Regards,

Jorge

Hi Jorge,

I tried your solution but not solved the issue.

I remove the register key "VID_0451&PID_9260" in "HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Enum\USB".

But I don't have a register key "VID_0451&PID_9260" in "HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Enum\HID", is it normal?

When I replug the USB cable on my PC, the same message appears "new device detected USB mass storage" but after I have the message "an error occurs during the installation of your device" like one USB key which doesn't work..

Regards,

Matthieu

Hi Jorge,

Is it normal that after the device is programming, the led HDD_Act turns ON and OFF very slowly (with dimming) ?

It do the same thing if I put a SATA disk or not.

For programming the flash, I open a switch on Data Out (like your jumper on EVM) of the flash to see the TUSB9260 on the flash burner tool and after I need to close the switch and launch the programming. Could you confirm to me it's normal or not?

Regards,

Matthieu.

Hi Jorge,

Sorry, but I get back to you because I have no news about my problem on the TUSB9260 component.
Is it possible that the firmware is not correct? Could I have the 0.95 firmware?

Thanks and have a good day.

Matthieu

Hello Matthieu,

The FW seems to be working fine, so this doesn't seems to be a programming issue.

HDD_Act dimming on/off means the device is suspended. The OS might suspend the device if enum fails and that's why we believe this is an enumeration issue on your system.

Have you tried plugging in your device on a different system? One with a fresh installation of Windows?

Regards,

Jorge.

Hi Jorge,

Yes, I have tried plugging in your device on four different system (3 with window XP and 1 with window Seven).

The problem is the same, the installation driver is not good.

I don't think that the problem is due to the USB signals because the detection of the device in Flash burner software is OK and the programming of flash is OK.

When I see your schematic of the demo board and my schematic, the only difference is on the SATA Tx line, on your demo board you swapped TX and nTX but on my schematic I didn't swapped these signals.

Do you think it can be a problem?

In the datasheet you say it can be done but is it the same firmware for both case ?

Thanks and have a good day.

Bruno

Hello Bruno,

   Yes, that should be the problem. The SATA polarity is not auto-sensed. You can either change the polarity in your hardware or use the attached firmware with a different polarity setting.

2744.FW_097_No_Polarity_swap.zip

Regards,

Jorge

Hi Jorge,

I have downloaded the new firmware and it works.

Just for my knowledge : how do you detect the swapp on these pins without hard disk connected?

Thanks for your help and have a good day.

Matthieu

Matthieu,

I'm glad to hear your implementation is now working. We are actually not detecting the polarity swap; the polarity is currently defined and hard-coded in the firmware.

Regards,

Jorge

Jorge Llamas (SWAT) said:

Roman,

 

                Can you please also verify your reset timing? I would like to make sure both the 3.3 V and 1.1 V supplies are up before the reset is de-asserted.

 

                Regards,

                Jorge.

Hi, Jorge.

I checked the reset line and it's working as you said.

Hello Roman,

  Are there any changes on your device after trying to program the SPI Flash (Different LED blinking pattern, different enumeration by Device manager on Windows, etc)?

  Can you please verify the polarity swap on the SATA terminals? (http://e2e.ti.com/support/interface/digital_interface/f/130/p/93314/410784.aspx#410784)

Regards,

Jorge.

Hello Jorge,

Can you provide firmware that has polarity switched on both TX and RX? Thanks,

Turi

Hello Turi,

      Please find attached a license wrapper containing the Firmware binary with both Tx and Rx lines swapped.

Regards,

Jorge.

1373.TUSB926x_FW097_TxRx_Polarity_Swap.zip

Hi Jorge,

Thanks for the swap firmware. I have another question regarding the use of a crystal instead of a crystal oscillator. Can I use a 25MHz crystal (NOT a crystal oscillator) if FSEL0=3.3V and FSEL1=0? This combination is shown as an example in the Implementation Guide, but  Table 3 lists the crystal options as "reserved". Thanks,

-t

Are you using the TUSB9260 or TUSB9261?

If you are using the 9260, we recommend you switch to the 9261 which fixes many errata in the 9260.

Hello Turi,

Perhaps you are seeing the wrong datasheet.

The TUSB9260 does not support 25MHz input frequency, the combination FREQSEL[1:0] = 01 is reserved.

The TUSB9261 does support 25MHz input frequency and it is correct to set FREQSEL[1:0] = 01 for 25MHz.

Regards.

Hello,

I'm using a TUSB9260 (and I can't switch to the TUSB9261 because it isn't pin to pin compatible and my production has already started) and it works with a crystal oscillator (a quartz not just an oscillator) at 25Mhz with FREQSEL[1:0] = 01. Does it mean that my design couldn't work properly in some conditions?

Regards

Matthieu Baque

There are jitter issues with the TUSB9260 in 25MHz.

It is recommended to use input frequency of 40MHz just as is stated on the datasheet.

Regards.

And could you tell me what is the consequence of these issues (sata link failed, speed reduced).

I use the TUSB9260 in USB2.0 instead of USB 3.0 maybe these issues occur just in SuperSpeed?

Regards

Matthieu Baque

The problem occurs with SSC enable on the Rj (random jitter) component of RX/TX, you can see the consequences on section 5.2 of datasheet (http://www.ti.com/lit/ds/symlink/tusb9260.pdf).

The firmware of the TUSB9260 does not support disable SSC, TUSB9261 does.

If you are going to use only USB2.0 host, then this jitter issue may not apply.

Regards.

Hi Elias,

Is it possible to confirm that the TUSB9260 can use a 25MHz XTAL if only USB2.0 is used? I would consider switching to the TUSB9261 but there is no stock for any distributors and no samples available from TI, and I'm on a tight schedule. Thanks,

-t

The jitter issues were found on the SSC, it only aims to the Super Speed section, since the TUSB9260 is intended for use with SuperSpeed devices that is why input frequencies other than 40MHz were reserved.

This specific jitter condition will not affect if you use 25MHz input frequency and only USB2.0 interface. In theory you should be fine.

However, You must take in mind the Disclaimer below this web page. I cannot make a commitment of something that is not allowed by datasheet.

Regards.

Ok, understood.

One other question: In the reference schematic,VBUS is provided to the TUSB9260 input (pin 33) through a 90.9k||10k resistive divider. In my application, VBUS is not directly applied to the TUSB9260. If I am using a 3.3V signal instead of 5V, should the new network be 60k||10k ? I can't find a description of the pin 33 input voltage parameters in the datasheet. Thanks,

-t

Hello,

Terminal USB_VBUS is used to detect the presence of USB cable voltage (VBUS = 5V)

So, it doesn't matter if you are powering your board with 3.3V, whenever you plug in a USB cable, you are going to have 5V there. 

Regards.

In my application, the TUSB9260 is downstream of another chip (similar to a host processor). Cable VBUS is applied to the other chip, not to the TUSB9260 directly. Furthermore, my system requires that the TUSB9260 is accessible to the other chip even when not plugged into a cable (e.g.,the host processor wants to access the SATA disk when operating on battery power.) Thanks,

-t

In that case, your design should work fine.

Just make sure your battery supplies the required current and voltage per USB and SATA specification.

Regards.

Just to clarify, do you mean that the voltage divider applied to pin33 should be changed to 60k||10k for 3.3V operation? I don't think that keeping the 90.9k||10k will work when only 3.3V is applied. Thanks,

Yes, 3.3V with voltage divider between 60K and 10K

Regards.

@TI-Support:

Could you post the current version of the firmware (v0.99) with NO POLARITY SWAP?

Hello Sebastian,

Please send me an email to elias.villegas@ti.com

Regards.

Hello Sebastian,

Your email client is blocking me to send the firmware,

Please provide another recipient.

Regards.

Hello,

thanks for your quick response. You can send it to schreiber.seb(at)googlemail.com

Regards

It is still rejecting it.

Delivery has failed to these recipients or groups:

schreiber.seb@googlemail.com
Your message wasn't delivered because the recipient's e-mail provider rejected it.

The following organization rejected your message: gmail-smtp-in.l.google.com.