Linux/AM3352: About RGMII glitch Issue

yimin bai34

Genius 4755 points

Part Number: AM3352

Tool/software: Linux

Hi SIr

We measured the RX CLK signal in RGMII interface and found the issue

During the CLK rising & falling time , the glitch would show up.

Question:

Whether the glitch causes emac to misjudge data while latch data

Thanks

Yimin

over 7 years ago

0 Biser Gatchev-XID over 7 years ago

TI__Guru**** 393215 points

Hi,

I don't see any glitches in this waveform. There is some ringing, which may be due to PCB trace impedance mismatch, routing length, spacing, etc. Do you have a serial resistor on this line? Check the Ethernet PHY documentation for routing recommendations.

0 yimin bai34 over 7 years ago in reply to Biser Gatchev-XID

Genius 4755 points

Hi SIr

Sorry for that.

Above photo is the is the clock source from MAC chip, the waveform is good.

below photo is RGMII CLK to CPU, we can find some glitches in the waveform.

Does the glitch cause AM3352's EMAC to misjudge data during latch data ?

Yimin

0 -DK- over 7 years ago in reply to yimin bai34

TI__Mastermind 27890 points

Hello Yimin,

While not ideal, I’ve certainly seen worse glitches without any functional failures. Are you seeing functional failures?

This looks like an impedance mismatch in your trace. Likely caused by a parasitic influence of a component on the trace, or a GND reference change+mismatch.

Processors

Processors forum

Linux/AM3352: About RGMII glitch Issue