DP83TD510E: TDR on long cables or high attenuation

Johan Jensen

Part Number: DP83TD510E

Hi!

I'm exploring the TDR function on the DP83TD510E. I've made measurements up to 625m, but beyond that I cannot get a result. In snla364c pg.6 there is a note:

Note A High Attenuation cable might need a different TDR script. A Register 030C reading of 0000 indicates that the cable is a high attenuation cable.

What do I need to change to make the script work with longer cables or higher attenuation? I'm using the default script from snla364c revision C. Knowing a little more about the TDR-registers would also be helpful (0x300-0x30A and 0x30E-0x310)

The cable is a single twisted pair with a ~100Ohm characteristic impedance, 0.22mm^2/24AWG area.

At 625m I get the following results (open ended cable):

DP83TD510E_TDR_Status.txt file is open...
Register 001F is: 4000
Register 001E is: 0002
Register 030C is: 0EBC
Register 0305 is: 0038
Register 0306 is: 0000
Register 0307 is: 1500
Register 0308 is: 0000
Register 0309 is: 0000
Register 030A is: 0001
End of file.

Which tells me:

		dec
TDR_CFG	TDR_Done	1
	TDR_Fail	0
TDR_Fault_Status	Peak_Detect	1
	Peak_Sign	1
	Peak_Location	700
FAULT_STAT1	peaks_loc_1
	peaks_loc_0	56
FAULT_STAT3	peaks_amp_0	21
	peaks_loc_4	1

At 815m I get the following results (open ended cable):

DP83TD510E_TDR_Status.txt file is open...
Register 001F is: 4000
Register 001E is: 0002
Register 030C is: 0000
Register 0305 is: 0000
Register 0306 is: 0000
Register 0307 is: 0000
Register 0308 is: 0000
Register 0309 is: 0000
Register 030A is: 0000
End of file.

Which tells me the TDR-process has run successfully, but no return signal detected?

Regards,

Johan Kleiberg Jensen.

over 3 years ago

0 Gerome Cacho over 3 years ago

TI__Mastermind 39795 points

Hello,

Thank you for your query. I am out of office and will look at this thread upon my return tomorrow, 9/7/22.

Sincerely,

Gerome

0 Johan Jensen over 3 years ago

Prodigy 10 points

As a note, I reran the TDR-process with a closed (shorted) end, but the results are the same or worse wrt "long" cables.

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

TI__Mastermind 39795 points

Hi Johan,

What type of ethernet cable are you using? CAT5? 6? 7?

Sincerely,

Gerome

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

Prodigy 10 points

It is a custom hybrid cable with one twistede pair used for SPE. Resistance is 92.4 Ohms/km, impedance is ~100 Ohms (Theoretical Impedance 102 Ohm, Delay 0.129 ns/in, Inductance 13.2 nH/in, Capacitance 1.27 pF/in)

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

TI__Mastermind 39795 points

Hi Johan,

We have done our TDR testing based on standard cable CAT6. If you are not using standard cable, then we cannot guarantee result for TDR.

Sincerely,

Gerome

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

Prodigy 10 points

Hi Gerome,

Thank you for your reply. The twisted pair in our cable meets electrical characteristics of a CAT6/CAT5e cable (DC-loop-resistance=0.1848Ω/m, impedance=102Ω, delay=5.08nS/m, inductance=519.8uH/m, capacitance=50pF/m, wire-diameter=0.53mm, cross-talk doesn't apply - only single pair, etc..).

I'm using the 3.2 TDR Test Procedure from Application Note DP83TD510E Cable Diagnostics Toolkit, and there is a note at the bottom page:

Johan Jensen said:
Note A High Attenuation cable might need a different TDR script. A Register 030C reading of 0000 indicates that the cable is a high attenuation cable.

What changes can be made to the TDR script in high attenuation cables?

Regards,

Johan.

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

TI__Mastermind 39795 points

Hi Johan,

I am reaching out to the team regarding this and will get back to you tomorrow.

Sincerely,

Gerome

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

TI__Mastermind 39795 points

Hi Johan,

I am attaching the following scripts; one for calibration using CAT6 cable that team has done, and other for procedure to read/run TDR test. Hopefully, this is able to solve your query.

Fullscreen CAT6_TDR_script (working).txt Download

begin
000D 001F
000E 001F
000D 401F
000E 8000

000D 0007
000E 0200
000D 4007
000E 0000

000D 0001
000E 0834
000D 4001
000E 4000

000D 001F
000E 0301
000D 401F
000E 2407

000D 001F
000E 030D
000D 401F
000E 4825

000D 001F
000E 0303
000D 401F
000E 0432

000D 001F
000E 030E
000D 401F
000E 0412

000D 001F
000E 0304
000D 401F
000E 000C

000D 001F
000E 030f
000D 401F
000E 0008

000D 001F
000E 0302
000D 401F
000E 3e80

000D 001F
000E 0310
000D 401F
000E 0036

000D 001F
000E 001F
000D 401F
000E 4000

//write 001E 8000 and read until it is 0002
//then read the 030C to see where is open or short. (0c00 open 0800 short) (last two digit are the distance)
end

Fullscreen TDR_reading_script (working).txt Download

begin
000D 001F
000E 001F
000D 401F
000E 4000

000D 001F
000E 001E
000D 401F
000E 8000

000D 001F
000E 001E //error checking
000D 401F
000E 

000D 001F
000E 030C //reading TDR result
000D 401F
000E 
end

Sincerely,

Gerome

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

Prodigy 10 points

Hi Gerome, thank you for the script.

I've rerun the TDR-Process with the following results (cable is open ended):

	Script	Default	New	Default	New
Register	Cable Length	600	600	1200	1200
030C	Peak_Detect	1	1	0	1
030C	Peak_sign	1	1	0	0
030C	Peak_Location	663	663	0	675
0305	peaks_loc_0	h35	h35	0	h36
0307	peaks_amp_0	18	10	0	1
030A	peaks_sign_0	1	1	0	0

It seems the new script will detect a weak reflection in the connection where two 600m cables are joined, but otherwise not detect the open end at 1200m.

Is there any chance you can shed some light on registers 030D, 030E, 030F, and 0310?

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

TI__Mastermind 39795 points

Hi Johan,

These are proprietary registers that cannot be shared.

Sincerely,

Gerome

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

Prodigy 10 points

I understand. Do you have any more information about registers listed in the datasheet?:

TDR_CFG1 Register (Address = 0x300)
TDR_CFG2 Register (Address = 0x301)
FAULT_CFG1 Register (Address = 0x303)
FAULT_CFG2 Register (Address = 0x304)

In particular the parameters:

cfg_forward_shadow_1
cfg_forward_shadow_2
cfg_post_silence_time
cfg_pre_silence_time
cfg_end_tap_index_1
cfg_start_tap_index_1
cfg_tdr_flt_loc_offset_1
cfg_tdr_flt_init_1
cfg_tdr_flt_slope_1

It is not clear to me what these parameters do based on the description.

Regards,

Johan.

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

TI__Mastermind 39795 points

Hi Johan,

These are DSP registers within the PHY. There is no more information that I can provide on these registers.

Sincerely,

Gerome

0 Johan Jensen over 2 years ago in reply to Gerome Cacho

Prodigy 10 points

Hi Gerome,

Ok, thanks for looking into that.

What is the (maximum) expected range for the TDR-function?

Regards,

Johan.

0 Gerome Cacho over 2 years ago in reply to Johan Jensen

TI__Mastermind 39795 points

Hi Johan,

I am working with the team on this query. I expect to get back to you tomorrow.

Sincerely,

Gerome

0 Gerome Cacho over 2 years ago in reply to Gerome Cacho

TI__Mastermind 39795 points

Hi Johan,

We have tested CAT5E cables with TDR up to 800m.

On your 1.2km cable, are you able to get clean data transfer?

Sincerely,

Gerome

0 Johan Jensen over 2 years ago in reply to Gerome Cacho

Prodigy 10 points

Hi Gerome,

No, I cannot get a connection at 1.2km.

Regards,

Johan.

+1 Gerome Cacho over 2 years ago in reply to Johan Jensen

TI__Mastermind 39795 points

Hi Johan,

Thank you for your reply. The only recommendation we have on our side to ensure clean communication would be to switch over to standard cabling as that is what we have tested with.

Sincerely,

Gerome

0 Johan Jensen over 2 years ago in reply to Gerome Cacho

Prodigy 10 points

Hi Gerome,

Ok, thanks for looking into this. I'll chalk this down to overall attenuation.

Regards,

Johan.

Interface

Interface forum

DP83TD510E: TDR on long cables or high attenuation