TUSB2046B problem negotiation with some memory sticks

Ramin Fardi

Other Parts Discussed in Thread: TUSB2046B, TUSB2036

Hi,

I have a new design where I use TUSB2046B as an embedded hub connected to a single board computer running WIN CE. The TUSB2046B seems to have problems negotiation with some memory sticks but otherwise works fine. All low speed devices that we have tested such as keyboard, mouse etc. work fine too. Most of brand name memory sticks such as Sandisk or Kingstone wok fine as well but most of cheaper sticks won’t come up. If I connect a cheap stick directly (behind the hub) to the WIN CE computer it will work. If I connect an ordinary external hub to my TUSB2046B and connect the same nonworking cheap stick to the external hub it will work too. It is worth mentioning that in the older version of the same design we used TUSB2036 hub and never had this problem, but we needed more ports so I changed. What could be the problem?

Regards

Ramin Fardi

over 11 years ago

0 Joel Jimenez0 over 11 years ago

TI__Mastermind 18655 points

Hello Ramin,

We are reviewing your question, Could you post your schematic for review?

Regards.
Joel

0 Ramin Fardi over 11 years ago in reply to Joel Jimenez0

Intellectual 280 points

Hi,

Thanks for your attention. Here is the schematics. The USB signals are routed from the ZIF connector J4 to another board where the USB connectors are. U11 is a single board ARM9 computer running WIN CE.

Regards2605.Controller Board.pdf

Ramin Fardi

0 Joel Jimenez0 over 11 years ago in reply to Ramin Fardi

TI__Mastermind 18655 points

Hello Ramin,

Some devices have less control on D+/D- than others. An implementation may use small capacitors at the transceiver for purposes of edge rate control.
Could you add a 22pF capacitor between the transceiver pins and Rs as shown below?

Could you get some USB bus trace captures between the TUSB2046B and the device(cheap stick )?

Regards.

0 Ramin Fardi over 11 years ago in reply to Joel Jimenez0

Intellectual 280 points

Hi,

I've tried with capacitor without any result. Here are some screen captures. 2.gif is the second packet. 3.gif is a random packet in the middle. 4.gif is the last packet before it stops. Please let me know if you need any other measurement.

First packet

Second packet

Random packet in the middle

Last packet

Regards

Ramin

0 Joel Jimenez0 over 11 years ago in reply to Ramin Fardi

TI__Mastermind 18655 points

Hello,

It is very difficult to speculate on where the initialization is breaking down without looking at the whole bus trace.

Do you have access to a protocol analyzer? If it be so, could you send some pictures of the passing case and the failure?

If not , Are you able to ship one of your platform for testing?

Regards.

0 Ramin Fardi over 11 years ago in reply to Joel Jimenez0

Intellectual 280 points

Hi,

I understand the difficulty. But Unfortunately we do not have a bus analyzer. And thanks for the offer but we decided that the most probable outcome of any further trouble shooting is that we end up with the conclusion that some sticks are poorly designed (yet USB compliance) and this particular hub for some reason refuses to let those stick get through. In order to save time we are already on track to change to another hub which we tested positively with those poorly designed sticks.

Thanks for your efforts

Ramin

0 Joel Jimenez0 over 11 years ago in reply to Ramin Fardi

TI__Mastermind 18655 points

Ramin,

I double-checked your schematic and I noticed that the terminal TSTPLL/ 48MCLK is floating. This pin must be tied to ground for normal 6-MHz operation.

Regards.

0 Ramin Fardi over 11 years ago in reply to Joel Jimenez0

Intellectual 280 points

I tried that some time ago. Actually all the traces are taken with that pin grounded.

Thanks

Ramin

Interface

Interface forum

TUSB2046B problem negotiation with some memory sticks