DS90UH929-Q1: How to measure the waveform of FPD-Link III output

Nao

Part Number: DS90UH929-Q1
Other Parts Discussed in Thread: ALP

E2E members,

I would like to measure the waveform of FPD-Link III output that like datasheet "9.2.3.1 Application Performance Plots".

1. Can I get it on DS90UH929-Q1EVM?
2. Could you tell me how to measure waveform of FPD-Link III output on EMV?
- measurement points
- how to set register for measurement

Regards,
Nao

over 5 years ago

0 Nao over 5 years ago

Genius 4905 points

Hi,

Do you have any feedback?

Regards,
Nao

0 Darryl Jeong over 5 years ago in reply to Nao

TI__Expert 4210 points

Hi Nao-san,
You really don't need the UH EVM to do FPD-LinkII output.
You must be an HDCP adopter to purchase the UH device/EVM.
Here is the link to order the DS90UB929 EVM: www.ti.com/.../DS90UB929-Q1EVM
I your work on your #2 question

0 Nao over 5 years ago in reply to Darryl Jeong

Genius 4905 points

Hi, Darryl

Thank you for your reply.

We have the DS90UB929-EVM & DS90UH929-EVM. Des side is DS90UH940-EVM.
Our customer would like to measure the jitter of FPD-Link output from DS90Ux929.
(like a datasheet "9.2.3.1 Application Performance Plots".)

Please tell me some advice.

Regards,
Nao

0 Nao over 5 years ago in reply to Nao

Genius 4905 points

Hi, Darryl

Do you have any update?

Regards,
Nao

0 Darryl Jeong over 5 years ago in reply to Nao

TI__Expert 4210 points

Hi Nao-san,
What kind of oscilloscope do you have? What is the bandwidth of the oscilloscope?
Do you have 50-ohm coax cable with male SMA ends?

Regards,
Darryl

0 Nao over 5 years ago in reply to Darryl Jeong

Genius 4905 points

Hi, Darryl

Thank you for your reply.

#1 We have
Oscillo Scope　　 Tektorinix DPO71604C (Up to 16GHz）
Differential Probe 　Tektorinix　P7313SMA (Up to 13GHz)

#2 Yes.
Could you tell me how to measure waveform of FPD-Link III output on EMV to use SMA cable?
- measurement points
- how to set register for measurement

Regards,
Nao

0 Nao over 5 years ago in reply to Nao

Genius 4905 points

Hi, Darryl

Our customer is planning to measure the waveform of FPD-Link III output at 11/19.
So, Please tell us your feedback soon.

I would like to measure the waveform of FPD-Link III output that like datasheet "9.2.3.1 Application Performance Plots".
Could you tell me how to measure waveform of FPD-Link III output on EMV?
- measurement points
- how to set register for measurement

Regards,
Nao

0 Nao over 5 years ago in reply to Nao

Genius 4905 points

Hi, Darryl

Do you have any update?

Regards,
Nao

0 Nao over 5 years ago in reply to Nao

Genius 4905 points

Hi, Darryl

Can I get your feedback today?
I need your help to measure the FPD-Link output.

Regaerds,
Nao

0 Vijaya Ceekala over 5 years ago in reply to Nao

TI__Intellectual 2930 points

Hi, Nao

Here is the way to attach an SMA adaptor board and measure Vod

0 Nao over 5 years ago in reply to Vijaya Ceekala

Genius 4905 points

Hi, Vijaya

Thank you for your reply.

Can I get it on DS90UH929-Q1EVM?

I would like to measure it in following setting.
But I guess DS90UH929-EVM can not output FPD-Link signal without HDMI input.
DS90UH929-EVM accept HDMI input after link is established.
In this case, FPD-Link signal can not measure to use the Oscillo Scope,
because SMA cables connect between DS90UH929 and Des-EVM(DS90UH940-EVM).

Could you tell me how to measure the waveform of FPD-Link III output on following setting?
- how to set register for measurement.

Regards,

Nao

0 Darryl Jeong over 5 years ago in reply to Vijaya Ceekala

TI__Expert 4210 points

Hi Nao,
The cleanest measurement is with the HSD-to-SMA adapter but
if you do not have the HSD-to-SMA adapter to connect to P1 of the DS90Ux929 EVM, there is an option to use the on-board SMAs J9 and J11.
To use J9 and J11 SMA option, you must first populate R211 and R212 with a 0402 sized zero ohm resistor or bridge the pads with a solder bridge.
R211 and R212 are intentional not stuffed because the traces to SMAs J9 and J11 will be a stub when using P1.
For more details see the DS90Ux929-Q1EVM User's Guide: www.ti.com/.../snlu172.pdf.
Question, on the TEK scope you have, do you have a software package installed called DPOJET? If the eye opening meets the Ew then the software, DPOJET may not be necessary.

0 Nao over 5 years ago in reply to Darryl Jeong

Genius 4905 points

Hi, Daryl

Thank you for your feedback.
Yes. We have DPOJET software. We populated R211 and R212.

Do we need to set some register in DS90UH929 for measurement FPD-Link output with HDMI input?

or

Do we need to set the internal pattern generator of DS90UH929, if we can not connect HDMI input?

Regards,
Nao

0 Nao over 5 years ago in reply to Nao

Genius 4905 points

Hi, Daryl

Do you have any update?

Regards,
Nao

0 Darryl Jeong over 5 years ago in reply to Nao

TI__Expert 4210 points

Hi Nao,
If you do not apply an HDMI input then our internal AON (Always ON oscillator) will output at ~33MHz x 35 (1.155Gbps); no register setting is needed.
You can also set the internal pattern generator of the DS90UH929 so you can get different output frequencies

0 Nao over 5 years ago in reply to Darryl Jeong

Genius 4905 points

Hi, Darryl

Thank you for your reply.
Could you tell me how to set register for measurement, if I apply an HDMI input?
Our customer want to measure the FPD-Link output in state of HDMI input.

Regards,
Nao

0 Nao over 5 years ago in reply to Nao

Genius 4905 points

Hi, Darryl

Do you have any update?

Regards,
Nao

0 Nao over 5 years ago in reply to Nao

Genius 4905 points

Hi, Darryl

Our customer measured FPD-Link output jitter with Tektronix engineer.
But output jitter was higher than 0.3UI. The result is out of datasheet spec(< 0.3UI).

Their measurement condition is here.
HDMI Input 74.25MHｚ → FPD LinkⅢ：2.6Gbps

#1. Can you measure the output jitter in this conditon on your DS90UH929-EVM?
If yes. Please share your measurement data.

#2. If not. Please could you tell how to measure the output jitter that like datasheet
"9.2.3.1 Application Performance Plots" on DS90UH929?

Regards,
Nao

0 Nao over 5 years ago in reply to Nao

Genius 4905 points

Hi, Darryl

There are no updates from Nov 21.
Our customer are waiting for the reply, hence we need your comment about measurement of jitter.

Regards,
Nao

0 Darryl Jeong over 5 years ago in reply to Nao

TI__Expert 4210 points

Hi Nao,

Are you using the DPOJET software?
Which measurement are you using?
I suggest Tj@BER and DDJ
Which clock recovery method are you using?
I suggest Constant Clock - Mean
Which Edge are you selecting?
I suggest Data
Which RjDj configuration did you select?
I suggest Pattern Type: Arbritrary
Window Length: 5UI
Population: 100
Jitter Target BER: BER=1E-10
Which Filter settings did you select?
I suggest High Pass: 3rd Order
Filter frequency = PCLK/20, as stated in the datasheet.
Do a single measurement use the formula Tj@BER - DDJ

0 Darryl Jeong over 5 years ago in reply to Darryl Jeong

TI__Expert 4210 points

Hi Nao,
Did you try measuring the FPD3 eye opening with the DPOJET software?

0 Nao over 5 years ago in reply to Darryl Jeong

Genius 4905 points

Hi, Darryl

Our customer has the following measurement equipment.
Oscilloscope 　　 Tektronix DPO71604C (Up to 16GHz）
Differential Probe 　Tektronix　P7313SMA (Up to 13GHz)

And then, They measured FPD3 eye opening with Tektronix Engineer.
So, I guess they are using DPOJET.

#1 Please let me know, if we need more setting at DPOJET for measuring FPD3 eye opening.
#2 Can you get the data that jitter is lower than 0.3UI by DS90UH929-EVM?

Regards,
Nao

0 Darryl Jeong over 5 years ago in reply to Nao

TI__Expert 4210 points

Hi Nao,

I don't believe they were measuring the eye with DPOJET because they didn't have the settings I sent you to correctly measure the eye.

You should not need the differential probe as you can connect the FPD3 directly into the scope 50 ohm terminated input.

In reply to your questions:

1) settings provided already, no additional should be needed

2) Can you rephrase the question? I'm not exactly sure I understand the question.

0 Nao over 5 years ago in reply to Darryl Jeong

Genius 4905 points

Hi, Darryl

Thank you for your quick reply.

> 2) Can you rephrase the question? I'm not exactly sure I understand the question.

OK. I send you the Original question.

I would like to measure the waveform of FPD-Link III output that like datasheet "9.2.3.1 Application Performance Plots".
jitter is < 0.3UI.

#1. Can we measure it by DS90UH929-EVM?

If yes.
#2. Please tell me the measurement method.
- measurement equipment
- Please tell me how to connect.
- Please tell me the setting, if need it.

If no.
#3. Please tell me the measurement environment.

We do not want to know the bad point of our environment.
We just want to know how TI was measured it.
Our customer prepare the necessary measurement equipment and tools by themselves.

Regards,
Nao

0 Nao over 5 years ago in reply to Nao

Genius 4905 points

Hello Darryl

Do you have any update?
Please let me know how TI was measured it that datasheet "9.2.3.1 Application Performance Plots".

Regards,
Nao

0 Darryl Jeong over 5 years ago in reply to Nao

TI__Expert 4210 points

Hi Nao,
The measurement was done with a different scope than you have. It was done with and Agilent scope but you should be able to get similar results with the TEK scope. Please use the DPOJET and settings I provided earlier.

Regards,

0 Nao over 5 years ago in reply to Darryl Jeong

Genius 4905 points

Hi, Darryl

Thank you for your reply.

> Please use the DPOJET and settings I provided earlier.
You mean that following settings, right?
1. The measurement is Tj@BER and DDJ.
2. The clock recovery method is Constant Clock - Mean.
3. The Edge is Data.
4. The RjDj configuration are
- Pattern Type: Arbritrary
- Window Length: 5UI
- Population: 100
- Jitter Target BER: BER=1E-10
5. The filter settings is High Pass: 3rd Order.

Are there another settings?

Regards,
Nao

0 Nao over 5 years ago in reply to Nao

Genius 4905 points

Hi, Darryl

Could you confirm my last update?

Regards,
Nao

0 Nao over 5 years ago in reply to Nao

Genius 4905 points

Hello, Darryl

Do you have any update?

Regards,
Nao

0 Darryl Jeong over 5 years ago in reply to Nao

TI__Expert 4210 points

You are missing from your last question:

....applying the High Pass: 3rd Order
Filter frequency = PCLK/20, as stated in the datasheet.
Do a single measurement use the formula Tj@BER - DDJ

Is it not working?

0 Nao over 5 years ago in reply to Darryl Jeong

Genius 4905 points

Hi, Darryl

Thank you for pointing it out. I sent our customer these settings.
Please wait the response from our customer.
I was able to get to know about scope settings through talking to with you.
Thank you very much for your support.

BTW, I have additional questions about EVM setting.
How do you set the EVM for measurement of eye diagram?
#1 Do you use an internal pattern generator?
If yes.
#2 Could you please tell me the register settings?
If no.
#3 Could you please tell me how to another way?

Regards,
Nao

0 Darryl Jeong over 5 years ago in reply to Nao

TI__Expert 4210 points

Hi Nao,
You should not be using the pattern generator.
You should provide the external HDMI clock frequency of the application.
Can you supply the HDMI clock in?
This is important to find the true eye of the application with the real source of the application.

0 Nao over 5 years ago in reply to Darryl Jeong

Genius 4905 points

Hi, Darryl

Thank you for your update.
Yes. Our customer has HDMI source.

But basically, just connecting the HDMI source to the DS90UH929 EVM
does not recognize the HDMI signal.We need that HPD signal goes high.
However The DS90UH929 can not go the HPD signal high
without linking with the deserializer (We use DS90UH940 now).

How can I output the HDMI source to FPD-Link with only DS90UH929?
Now I am using "Assert HPD" of ALP. Is this the correct way?

Regards,
Nao

0 Nao over 5 years ago in reply to Nao

Genius 4905 points

Hello Darryl

Do you have any update?
Is my understanding correct?

Regards,
Nao

0 Darryl Jeong over 5 years ago in reply to Nao

TI__Expert 4210 points

Hi Nao,
Since you don't have a deserializer you will need to write to these registers:
0x5C[6] - force link ready
0x27[80] - bypass the state machine

0 Nao over 5 years ago in reply to Darryl Jeong

Genius 4905 points

Hi, Darryl

Thank you for your reply.
I will tell our customer your answer.

Regards,
Nao

0 Nao over 5 years ago in reply to Nao

Genius 4905 points

Hi, Darryl

Our customer was able to measure the waveform of FPD-Link III output.
So, we close this E2E. Thank you very much.

Regards,
Nao

0 Darryl Jeong over 5 years ago in reply to Nao

TI__Expert 4210 points

Hi Nao,
You are very welcome.
Thank you very much for you patience too.

Regards,
Darryl

Interface

Interface forum

DS90UH929-Q1: How to measure the waveform of FPD-Link III output