DP83822I: DP83822I with Capacitive Couple at 10Base-Te operation (Is register write is necessary?)

Daniel Haddad

Intellectual 255 points

Part Number: DP83822I
Other Parts Discussed in Thread: DP83TD510E,

Hi Team,

We are developing a SPE(DP83TD510E) to 10Base-TE(DP83822i) media convertor.

For MDI side on DP83822i we are using capacitive coupling (not magnetics).

Is it necessary to write the following register for 10Base-Te operation?

Line Driver Class Selection Register<0x0404> = 0x0024 To program full MLT-3 on both Tx+ and Tx– (Class A)

Or is this only required when operating PHY with Capacitive Coupling at 100Mbs?

I have referenced the following forum link for the above register write;

DP83822I: Configuration without Magnetics and RJ45 - Interface forum - Interface - TI E2E support forums

Our design will sit on a backplane and PCB trace distance from PHY-PHY is ~100mm.

Kind Regards,

Daniel

over 3 years ago

0 Kallikuppa Sreenivasa over 3 years ago

TI__Guru**** 189539 points

Hello Daniel,

Thank you for the query.

For the 10M default value should work.

We do not have any performance data to share.

Did you have a chance to test ad what are you observations ?

Regards,

Sreenivasa

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

Intellectual 255 points

Hi Sreenivasa,

Thanks for your prompt response!

We have not had a chance to test although we will allow for both Transformer & Capacitor options for testing once we receive prototypes.

Appreciate the reassurance around the default register value.

Kind regards,

Daniel

+1 Kallikuppa Sreenivasa over 3 years ago in reply to Daniel Haddad

TI__Guru**** 189539 points

Hello Daniel,

Thank you for the message. Do reach out to us to review the schematics when your are ready.

If you do not have additional questions, please click the resolved button to close.

regards,

Sreenivasa

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

Intellectual 255 points

Hi Sreenivasa,

That would be great i have a finished schematic i can share with you over private message, i have sent you friend request.

Thanks,

Daniel

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

TI__Guru**** 189539 points

Hello Daniel,

I did a quick review for the DP83822I and 510E based media converter and provided my comments.

Regards,

Sreenivasa

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

TI__Guru**** 189539 points

Hello Daniel,

Any thoughts if the DP83822i with capacitive isolation connects to another board with DP83822I ?

Regards,

Sreenivasa

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

TI__Guru**** 189539 points

Hello Daniel,

Please expect a delay in response due to the year end holidays.

Regards,

Sreenivasa

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

Intellectual 255 points

Hello Sreenivasa,

From your comment on MLT-3 (default) value should work for 10M we had planned once we complete the PCB design we will be able to test as we will have both options fitted.

Do you suggest we test before hand?

Regards,

Daniel

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

TI__Guru**** 189539 points

Hello Daniel

Thank you for the reply.

I would suggest doing a quick test if you have the setup to test before finishing the board.

Regards,

Sreenivasa

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

Intellectual 255 points

Hello Sreenivasa

Okay sounds good we should test once we return on 17th Jan.

Is it okay to test with magjack connector in-circuit and take PHY TX/RX pins to another PHY (via capacitors) as we only have one pair of 510E EVM with DP83822i and magjack fitted?

Regards

Daniel

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

TI__Guru**** 189539 points

Hello Daniel,

Thank you for the reply.

Although this is not recommended i would request you to give a try. Please solder the caps near to the connector.

Regards,

Sreenivasa

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

Intellectual 255 points

Hi Sreenivasa,

Have just completed testing with 100NF and all is working with 10M operation, no need to write register for full MLT on 10M link.

Our network bandwidth testing via Jperf also provided good results as follows;

Duration	Total Data	Send Send Rate	Jitter (Total)	Lost/Total Datagrams
0.0-10.0 sec	11.4 MBytes	9.57 Mbits/sec	4.013 ms	0/ 8143 (0%)

Thanks again for your help with this, much appreciated.

Regards,

Daniel

+1 Kallikuppa Sreenivasa over 3 years ago in reply to Daniel Haddad

TI__Guru**** 189539 points

Hello Daniel,

Good morning and good to hear the testing progress.

Regards,

Sreenivasa

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

TI__Guru**** 189539 points

Hello Daniel,

Quick question - Did you have 1 X 100nF 0r 2 X 100nF when you were communicating between two devices ?

Regards,

Sreenivaa

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

Intellectual 255 points

Hi Sreenivaa,

We have only 1 x 100NF per TX/RX +/- lines (4 Total) between PHY A & PHY B.

Regards,

Daniel

+1 Kallikuppa Sreenivasa over 3 years ago in reply to Daniel Haddad

TI__Guru**** 189539 points

Hello Daniel,

Noted and thank you.

Regards,

Sreenivasa