SK-AM62: Test eth0 using phytool issues the signal which will be shown on oscilloscope, but test eth1 would not.

Hi Kylix,

I need more time to discuss with my team, please allow another day to get back to you with a response.

Regards,

Alvaro

Dear Alvaro,

Thanks for you reply.

With the correct command, we can get waveform for 100 Base Standard MDI and 10 Base Link Pulse.

But when we want to test 10 Base Standard, there is only link pulse after entering the command.

Here are our command lines for each test mode, could you give us some advices?

# eth1-10 Base Link Pulse:
phytool write eth1/1/0x1F 0x8000
phytool write eth1/1/0x00 0x0100
phytool write eth1/1/0x10 0x5008

# eth1-10 Base Standard:
phytool write eth1/1/0x1F 0x8000
phytool write eth1/1/0x00 0x0100
phytool write eth1/1/0x10 0x5008
phytool write eth1/1/0x16 0x0020

Waveform for eth1-10 Base Link Pulse

Hi Austin,

What tests are you trying to run? What DP83867 functionality are we trying to test? 

Regards,

Alvaro

Dear Alvaro

We are verying the Signal Integrity of DP83867 .

We want to run the test below

We expect to get the output as below

Hi Austin,

I tried running the same test as you in lab and ran into the same problem. Please allow me another day to debug the 10 Base Standard Script. I will reply by end of day tomorrow, August 9th.

Regards,

Alvaro

Hi Austin,

I thought I had found a solution earlier this morning, but I noticed an error in my solution.

I apologize for the delay but I need more time to resolve this one. 

What I noticed is that the 100 Base Standard (MDI) script found in this document, the one I believe you to be referencing, it forces MDI, selects a Test Mode, then forces that Test Mode out the MDI side (ethernet cable).

The 10 Base Standard Script does not do this, and instead turns the phy into a Loopback mode. I tried enabling PRBS, copying the 100 Base Standard script except forcing 10 Mbps, and even choosing different loopbacks with no luck. Below is script that managed to send signals out of the MDI side, but it was by setting the Test mode to MLT3 Idles, which I don't think applies to 10BaseTe. 

001F 8000 // hard reset
0000 0100 // Forced 10M
0010 5408 // Forced MDI Mode, Force Link                                                                                                                                                                                0009 BB00 // Test Mode 5
0025 0480 // output test mode to all channels

Regards,

Alvaro 

Hi, Alvaro 

I can't get expeccted waveform with your recommandation.

If I enter the Test Mode 5, the PHY seems work under 100M

If I skip the ccommand line for Test Mode 5, there is only link pulse.

Hi Austin,

It has come to my attention that the test fixture I have in lab is not 10Base-Te compliant. I wonder if you have the same problem.

I will continue to try to find a way around this, but can you check this for the time being? 10Base-T & 10Base-Te are slightly different.

Regards,

Alvaro 

Dear Alvaro

Thanks for your support.

The test is finished.

Will TI update test procedurce document to prevent this issue?