SN75LVCP601 SATA Gen2 Gen3 issue

Prescott Siao

Intellectual 2220 points

Other Parts Discussed in Thread: SN75LVCP601

Hi Team,

We met the issue when customer used SN75LVCP601 on SATA gen2 and gen3 testing

First Customer used DE=0 dB(VOD=550 mV) and EQ=0 dB(SATA gen2 fail, SATA gen3 pass).

We tried to change the EQ setting to 7 and 14 dB, but the eye diagram looks like the same.

The VOD isn't enough, so we change the DE setting to -2 (VOD=830 mV) and -4 dB (VOD=630 mV)

Then gen2 pass but gen3 fail.

Below scope captures are gen2 and gen3 (DE=-2dB, EQ=14dB)

We also changed the EQ=0 and 7 dB, got the same results (gen2 pass, gen3 fail).

So we used the DE setting to -4dB(VOD=630 mV) to get smaller VOD.

But gen2 pass, gen3 fail.

So is there any possible to change the DE and VOD function separately?

Best regards,

Prescott

over 11 years ago

0 Elias Villegas M. over 11 years ago

TI__Guru* 76130 points

Hello,

Where are you taking those measurements? At the LVCP601 output or at the receiver input?

What are the trace lengths before and after the re-driver?

Regards.

0 Prescott Siao over 11 years ago in reply to Elias Villegas M.

TI__Intellectual 2220 points

Hi Elisa,

We tested at SATA conn as below.

and

L1: 1672.69mil

L2: 2484.43mil

L3: 1107.72mil

L4: 1868.34mil

We funded out the output eye of redriver look like below figure (middle eye drop and overlap to mask) if PCH SATA GEN2 provider higher signal swing to redriver.

We don't see this effect if PCH SATA GEN2 provider lower signal swing to redriver.

What is the reason to cause the effect if input signal is higher (middle eye drop and overlap to mask)?

Regards,

Prescott

0 Elias Villegas M. over 11 years ago in reply to Prescott Siao

TI__Guru* 76130 points

Hello,

Please try the first configuration with DEW = Long (VCC)? Changing DE to NC should also give the best eye(Keeping DEW set to VCC).

Both EQ settings should be either 0dB or 7dB.

Regards

Interface

Interface forum

SN75LVCP601 SATA Gen2 Gen3 issue