DP83869HM: DP83869HMRGZR communication issue after replacing DP83867IRRGZ

Rory Wu

Part Number: DP83869HM
Other Parts Discussed in Thread: DP83869

Hello Team

My customer used DP83869HMRGZR to replace DP83867IRRGZ and did a compliant design, please see below schematic, they use some NC resistor to jump some pin connections.

DP83869原理图.pdf

When using DP83867IRRGZ, it can communication normally, but in DP83869HMRGZR, they met packet loss issue when communication with external PC as well as in loopback test.

They short the RJ45 1,3 and 2,6, and set below register for the loopback test:

0x1F = 0x8000

0x31 = 0x10b0

0x00 = 0x2100

0x16 = 0x10

0x10 = 0x5008

0x1F = 0x4000

I asked them to dump the all register:

Fullscreen DP83869HM register dump.txt Download

read 0x0000=0x1140.
read 0x0001=0x796d.
read 0x0002=0x2000.
read 0x0003=0xa0f1.
read 0x0004=0x01e1.
read 0x0005=0xc1e1.
read 0x0006=0x006f.
read 0x0007=0x2001.
read 0x0008=0x4000.
read 0x0009=0x0000.
read 0x000a=0x0000.
read 0x000f=0xf000.
read 0x0010=0x5048.
read 0x0011=0x7c02.
read 0x0012=0x0000.
read 0x0013=0x1c40.
read 0x0014=0x29c7.
read 0x0015=0x0003.
read 0x0016=0x0000.
read 0x0017=0x01e1.
read 0x0018=0x6150.
read 0x0019=0x4444.
read 0x001a=0x0002.
read 0x001e=0x0012.
read 0x001f=0x0000.
read 0x0025=0x0480.
read 0x002c=0x141f.
read 0x002d=0x0000.
read 0x002e=0x0221.
read 0x0031=0x10b0.
read 0x0032=0x00d0.
read 0x0033=0x0000.
read 0x0037=0x0000.
read 0x0039=0x0000.
read 0x003a=0x0000.
read 0x0043=0x07a0.
read 0x004f=0x0200.
read 0x006e=0x1020.
read 0x0086=0x0377.
read 0x0134=0x1000.
read 0x0135=0x0000.
read 0x0170=0x0c0f.
read 0x0180=0x0752.
read 0x0181=0xc850.
read 0x0182=0x5326.
read 0x0183=0xa01e.
read 0x0184=0xe976.
read 0x0185=0x19cf.
read 0x0190=0x0000.
read 0x0191=0x0000.
read 0x0192=0x0000.
read 0x0193=0x0000.
read 0x0194=0x0000.
read 0x0195=0x0000.
read 0x0196=0x0000.
read 0x0197=0x0000.
read 0x0198=0x0000.
read 0x0199=0x0000.
read 0x01a4=0x0000.
read 0x01a5=0x0000.
read 0x01a6=0x0000.
read 0x01df=0x0000.
read 0x01e0=0x417a.
read 0x0c00=0x0140.
read 0x0c01=0x6149.
read 0x0c02=0x2000.
read 0x0c03=0xa0f1.
read 0x0c04=0x0020.
read 0x0c05=0x0000.
read 0x0c06=0x0000.
read 0x0c07=0x2001.
read 0x0c08=0x0000.
read 0x0c18=0x01ff.
read 0x0c19=0x0010.

Can you please kindly help to see any potential root cause we can find? Thanks.

BR,

Rory

over 3 years ago

0 Gerome Cacho over 3 years ago

TI__Mastermind 39795 points

Hi Rory,

Schematic review typically take a 5 business day turn around. I will have an update by CoB 11/5 at the latest.

Sincerely,
Gerome

0 Rory Wu over 3 years ago in reply to Gerome Cacho

TI__Intellectual 1050 points

Cacho,

It is not schematic review, it is an urgent issue, can you please check the root cause of this case?

Rory

0 Gerome Cacho over 3 years ago in reply to Rory Wu

TI__Mastermind 39795 points

Hi Rory,

Part of the issue may lie in the implementation of the DP83869 in the schematic. Reviewing the schematic to ensure correct implementation of this PHY from a previous DP83867 system takes time. This is in addition to reviewing the register data that you had sent. I will try to make sure to get a response back to you as fast as possible, but understand that this takes time.

Sincerely,

Gerome

0 Gerome Cacho over 3 years ago in reply to Gerome Cacho

TI__Mastermind 39795 points

Hi Rory,

Is it possible to ask customer for updated schematic? They seem to use DP83867 symbol while making adjustments to make it DP83869 compatible. This make it confusing to properly navigate. I think having proper symbol makes it easier to see what is going on. If they could clean up the net names that could also help a bunch.

Among first glances, some notes:

- Have customer follow recommended decoupling scheme for power supplies

- Ask customer whether or not they are supplying 1.1V or 1.0V from LDO. DP83869 needs 1.1V.

- Ensure that crystal (and load caps) meets specs of DP83869

- Ensure that magnetics also meets specs of DP83869

- There are nodes that are being placed on pins that shouldn't be driven (ie SIN/SIP/SON/SOP, CLK_OUT)

- RX_CTRL strapping needs to change. This is a 2 level strap for this PHY.

Sincerely,

Gerome

0 Gerome Cacho over 3 years ago in reply to Gerome Cacho

TI__Mastermind 39795 points

This matter has been taken offline.

0 Kallikuppa Sreenivasa over 3 years ago in reply to Gerome Cacho

TI__Guru**** 190079 points

Hello Rory,

Based on the layout review here are some inputs to improve the layout.

Based on the layout analysis, capacitive terminations for the magnetics on the PHY side has to be provided. Additionally please ensure spacing between the magnetics and the EPHY is > 1 inch and also the spacing from the PHY to the MAC is > 1 inch. Do not place sensitive components near to the magnetics.

Regards,

Sreenivasa

0 Kallikuppa Sreenivasa over 3 years ago in reply to Kallikuppa Sreenivasa

TI__Guru**** 190079 points

Hello Rory,

I will close this thread as we continue to discuss over email.

Regards,

Sreenivasa

0 Rory Wu over 3 years ago in reply to Kallikuppa Sreenivasa

TI__Intellectual 1050 points

Thanks Sreenivasa!

Interface

Interface forum

DP83869HM: DP83869HMRGZR communication issue after replacing DP83867IRRGZ