TLK10232: RX differential Vp-p is over SPEC

Kai Hsiao

Part Number: TLK10232

Hi Sir

We test some of SPF+ module and measure the module to HSRX 's signal ,but the Vp-p result is 1200mV~1300mV and over module's SPEC (300~850mV)

May I know have any register can fixed this issue? Just like termination resistor?

Thanks

over 7 years ago

0 Yaser Ibrahim over 7 years ago

TI__Genius 10250 points

Hello,

If I understand your question right, you are saying the High-Speed side output Vpp from TLK10232 is too high. If this is the case, the field HS_SWING[3:0] of register HS_SERDES_CONTROL_2 can be used to control this. Please refer to page 43 and page 44 in the datasheet. The Low-Speed side also has similar control.

Regards,
Yaser

0 Eric Lo72350 over 7 years ago in reply to Yaser Ibrahim

TI__Intellectual 915 points

Hi Yaser

You miss understanding customer's question.

The SFP+ module output spec is only 850mV maximum but they saw 1200~1500mV on TLK10232 input of their PCB as below block diagram.

customer want to know why? does TLK10232 has register for internal termination impedance setting?

0 Yaser Ibrahim over 7 years ago in reply to Eric Lo72350

TI__Genius 10250 points

Hi Kai,

Thanks for clarifying. There is no register to set the termination in TLK10232. The termination is 100 ohm +/-15%. Does this happen with every paltform (combination of SFP+ + TLK10232, or just some of them? Have they checked the transmitter (SFP+) to make sure its output is actually < 850 mV?

Regards,
Yaser

0 Kai Hsiao over 7 years ago in reply to Yaser Ibrahim

Guru 12675 points

Hi Yaser

1. Remove TLK10232 to measure HSRX Diff PAD of TLK10232 ,the Vpp is 1.5V

2. Remove TLK10232 and mount a 100ohm resistor between HSRX Diff PAD , the Vpp is 800mV

Have anything I need to do ?

0 Yaser Ibrahim over 7 years ago in reply to Kai Hsiao

TI__Genius 10250 points

Hi Kai,

Thanks for providing the details. Does this happen with every TLK10232 device?

Regards,
Yaser

0 Kai Hsiao over 7 years ago in reply to Yaser Ibrahim

Guru 12675 points

Hi Yaser

So far measure 3pcs of TLK10232 ,These 3pcs TLK10232 have the same issue
Could you kindly help to check this issue on your LAB?
Due to it have already impact our product mass production

Thanks

0 Kai Hsiao over 7 years ago in reply to Kai Hsiao

Guru 12675 points

Hi Yaser

Looking forward your reply

Thanks

0 Yaser Ibrahim over 7 years ago in reply to Kai Hsiao

TI__Genius 10250 points

Hi Kai,

Sorry, we have been really busy and didn't have a chance to check this out in the lab yet. We will try to get to this as soon as possible.

Regards,
Yaser

0 Malik Barton57 over 7 years ago in reply to Kai Hsiao

TI__Guru 64326 points

Kai,

What is the trace impedance of the traces leading into the TLK10232 receiver?

0 Malik Barton57 over 7 years ago in reply to Malik Barton57

TI__Guru 64326 points

Kai,

Is any more support needed for this issue? If so please reply with any relevant details so that I can further assist you. For now I will be marking this thread as "TI Thinks Resolved".

0 Kai Hsiao over 7 years ago in reply to Malik Barton57

Guru 12675 points

Hi Malik

Trace impedance : Differential 100ohm

Trace length Diff+: 1179.26 mil

Trace length Diff- : 1179.25 mil

Maybe need your fully support to check this issue

0 Malik Barton57 over 7 years ago in reply to Kai Hsiao

TI__Guru 64326 points

Kai,

Could you please post your schematic for this application? Are there any unused pins that are not terminated correctly?

0 Roger Chang7 over 7 years ago in reply to Malik Barton57

Prodigy 60 points

Hi Malik,

Sorry for late reply for the last 3 days are the public holidays in Taiwan.

The attachment is the related part of the schematic.

Please check it.

Regards,

Roger C.

ATEN Inc.

TLK10232_ckt_for_TI.pdf

0 Roger Chang7 over 7 years ago in reply to Malik Barton57

Prodigy 60 points

Hi Malik,

Any idea?

Regards,

Roger C.

ATEN Inc.

0 Malik Barton57 over 7 years ago in reply to Roger Chang7

TI__Guru 64326 points

Roger C.,

I am currently reviewing your schematic and will try to measure the termination on my side. through a TLK10232 EVM. I will update as soon as I have further details.

0 Kai Hsiao over 7 years ago in reply to Malik Barton57

Guru 12675 points

Hi Malik

May I know the currently status ?

Thanks

0 Malik Barton57 over 7 years ago in reply to Kai Hsiao

TI__Guru 64326 points

Kai,

I have measured the termination in our lab and was only able to get nominal values around 100 ohm. Could you try to measure the termination on your end with optical module?

0 Kai Hsiao over 7 years ago in reply to Malik Barton57

Guru 12675 points

Hi Malik

Could you kindly help to provide the picture with your test method
It will be easy help us to verify

Thanks

0 Roger Chang7 over 7 years ago in reply to Malik Barton57

Prodigy 60 points

Hi Malik,

Could you let us know your detail method and steps to measure the termination in your lab?

We will try to duplicate it in our lib.

Right now, we are trying to use the network analyzer to measure the trace impedance between SFP+ and TLK10232 (D1_XFI_RXIN_P / D1_XFI_RXIN_N)

to verify it with the impedance setting of our PCB layout rule.

Following is the related schematic and PCB layout rule, FYI.

0 Malik Barton57 over 7 years ago in reply to Roger Chang7

TI__Guru 64326 points

Kai,

The test was very simple. I measured the DC resistance as seen across the RX diff pair using a multi-meter.

Power up TLK10232EVM.
Ensure receiver termination is enabled through HS_ENRX bit field.
Measure the RX termination with multi-meter via the on board SMA connectors.

The method you are using right now is good to verify the trace impedance but be sure to try an measure the DC resistance at the RX diff pairs of TLK10232 when it is powered on and the HS_ENRX bit is high.

0 Malik Barton57 over 7 years ago in reply to Malik Barton57

TI__Guru 64326 points

Kai,

Is any more support needed for this issue? If so please reply with any relevant details so that I can further assist you. For now I will be marking this thread as "TI Thinks Resolved".

0 Kai Hsiao over 7 years ago in reply to Malik Barton57

Guru 12675 points

Hi Malik

Please kindly help to check my question ,May I know your idea?

1. Yes!! We follow your test step ,we can measure the 100ohm on HSRX pin ,but in this moment we use differential probe to measure the HSRX pin's waveform

the eye amplitude is 1300mVpp and over SFP+ spec (SPEC is 300~800mVpp)

2. When we remove TLK10232 IC on our board and add a 100ohm on HSRX pin ,the amplitude is 796mVpp (meet SPF+ SPEC) ,

0 Malik Barton57 over 7 years ago in reply to Kai Hsiao

TI__Guru 64326 points

Kai,

Are you using a high impedance probe? What are the termination settings on the scope that you are using? The first image seems to be affected by the loading caused by the probes and potential improper termination on the scope. What model of scope is being used in this test?

0 Malik Barton57 over 7 years ago in reply to Malik Barton57

TI__Guru 64326 points

Kai,

Is any more support needed for this issue? If so please reply with any relevant details so that I can further assist you. For now I will be marking this thread as "TI Thinks Resolved".

Interface

Interface forum

TLK10232: RX differential Vp-p is over SPEC