DS90UB953-Q1: About SI analysis method of FPD-LINK3

Tomoaki Yoshida91

Genius 5255 points

Part Number: DS90UB953-Q1
Other Parts Discussed in Thread: DS90UB960-Q1, DS90UB954-Q1,

Hi team,

Our customer would like to do SI analysis of A/W, wire harness and connectors route between 953 and 954.

They are trying to verify using the ideal SPICE model like the following, is it a general method?

To verify with this method, we need to know skew occurring at TX of 953 and skew allowed at RX of 954.

Can I know these skew?

Best regards,

Tomoaki Yoshida

over 7 years ago

0 Davor Glisic over 7 years ago

TI__Expert 8165 points

Hello-

The channel requirements given in Table 7 of the DS90UB960-Q1 datasheet (www.ti.com/.../ds90ub960-q1.pdf) should be the focus of the SI study. The requirements are applicable to DS90UB953-Q1 and DS90UB954-Q1 devices as well.

The intra-pair skew for differential signaling is not specified, however, it is assumed to be minimal (< 5 mils).

Regards,
Davor

0 Tomoaki Yoshida91 over 7 years ago in reply to Davor Glisic

Genius 5255 points

Hi Davor-san,

Thank you for your support.
I understand that the channel request can also be adapted to 953-954.

As I mentioned before, our customers are trying to verify the design of FPD-LINK3 communications by simulation, but I think there is no need for simulation.
Normally, is it necessary to verify by such simulation?

Best regards,
Tomoaki Yoshida

0 Davor Glisic over 7 years ago in reply to Tomoaki Yoshida91

TI__Expert 8165 points

Hello Yoshida-san,

Typically, following good PCB layout practices should be sufficient to meet the channel requirements. However, if s-parameter models are already available, simulations could help ensure 1st pass success of the channel design.

Regards,
Davor

0 Tomoaki Yoshida91 over 7 years ago in reply to Davor Glisic

Genius 5255 points

Hi Davor-san,

Thank you for your support.
When verifying with the model as shown in the figure above, I think that it is good to input the sine wave and check the attenuation at the termination.
Is there no mistake in this way?

Best regards,
Tomoaki Yoshida

0 Davor Glisic over 7 years ago in reply to Tomoaki Yoshida91

TI__Expert 8165 points

Hello Yoshida-san,

The insertion loss (attenuation) data should already be included in the s-parameter models. Simply use a Touchstone (S-parameter) viewer to plot the S12 (insertion loss) and check the insertion loss at the frequencies of interest.

Regards,
Davor

Interface

Interface forum

DS90UB953-Q1: About SI analysis method of FPD-LINK3