DP83869HM: Used as Media Converter from wired PHY to SFP - Works only with EVM but not with other external Media Converters.

Chester Kronke

Part Number: DP83869HM
Other Parts Discussed in Thread: MSP-EXP430F5529LP, DP83869EVM, DP83869, USB-2-MDIO

We have an in-house custom designed board that utilizes the TI DP83869HM as a media converter interface. The intent of this design is to interface an existing 10/100Base-T wired Ethernet interface (of a daughter card module) to an on-board SFP socket. The SFP socket then allows the use of SFP modules (wired, MMF, or SMF) to be installed in each board depending on link distance required for that unit’s field application (i.e., could vary from less than 100 ft per link to several miles). Initially we proved the design using wired SFPs which is the simplest case (i.e., does not require another external SFP to translate back to wired interface at opposite end of external link). Once we switched to fiber optic SFPs, we have had issues.

Our custom board design used the DP83869EVM as design reference for the DP83869HM portion of our custom board. It includes all the same configuration strap jumpers that are provided by the EVM board which are all implemented in the same way as shown in the EVM schematics. In our design, the DP83869HM is used as an unmanaged converter so we did not include the MSP430 launch pad or any other MCU on board. However, the MDIO and MDC pins go to an SMI connector same as on EVM board. This allows us to connect an external launch pad. I have purchased the TI MSP-EXP430F5529LP for this purpose.

I also have two DP83869EVMs that I can use as independent external media converters.

I have a simple strategy where I prove out any fiber optic SFP modules prior to testing them with our custom design. To do this I use the following basic test setup:

PC1 <====> [UWS1] <====> M/C1:SFP1 <::::x::::> SFP2 :M/C2 <====> [UWS2] <====> PC2

where:

PC1 & PC2 are PCs with 10/100/1000 NICs configured with separate fixed local IP addresses.

[UWS1] and [UWS2] are optional independent 10/100BASE-T Unmanaged Wired Switches.

M/C1 and M/C2 are external individual media converters (with single SFP socket and single RJ45 jack).

SFP1 and SFP2 are two fiber optic SFPs being tested (usually identical models but can also be mixed).

<====> represents a single wired Ethernet jumper.

<::::x::::> represents a dual fiber jumper with appropriate cross-over of fiber between the end connectors.

The above chain is everything in the basic SFP proving test setup. There are no other network connections to the PCs nor to the optional switches.

The two unmanaged switches allow me to force the link speeds on either side down to 100 Mbps. Without the switches, the PC NIC to M/C link speed will be 1000 Mbps. This is verified by link speed indicators on the M/Cs.

To pre-test a pair of FO SFPs, I set them up in the above configuration and verify I can communicate from PC1 to PC2 (simple ping tests both ways). I have 3 different types of packaged commercial M/Cs (in addition to the two EVMs) that I can use for the M/Cs. Once I have a particular pair of SFPs working in this proof setup, I then can remove SFP2 from M/C2 and plug it into our custom board's SFP socket to test if the SFP works in our design. In this way, I am making sure that everything from PC1 to SFP2 is working before testing with our design. The test setup with our custom board is then as follows:

PC1 <====> [UWS1] <====> M/C1:SFP1 <::::x::::> SFP2 : Custom Board : Daughter Card

The PHY interface on the Daughter Card is also configured with an independent fixed IP address. In this setup, if I can ping from PC1 through to the daughter card, then the test passes.

The problem is that I can get SFPs that will work between two EVMs as M/Cs and in this case I can then get the same SFPs to work with EVM as M/C1 and with SFP2 plugged into our custom board. Some of those same SFPs will also work using the commercial packaged M/Cs on both sides in the basic proving test, but I can never get a commercial M/C in M/C1 position to work with EVM as M/C2 or with our custom board.

I would have expected that if an SFP works with the commercial M/Cs on both sides and the same SFPs also work with EVMs as M/Cs on both sides, that I should then be able to get that same pair of SFPs to work with commercial M/C on the left and EVM M/C on the right. But I have not been able to get this to work. Can someone please shed some light on what I might be missing here that is preventing the commercial M/Cs from working with the EVM which in turn means they won’t work with our custom board? Or alternatively, anything that I can try on the EVMs to get it to work with a commercial M/C? If we can solve that, then we should be able to setup our custom board with the same configuration as EVM and get it to work with the commercial M/C too.

Thanks in advance for any help or guidance you can provide.

over 3 years ago

0 Cecilia Reyes over 3 years ago

TI__Genius 9920 points

Hi Chester,

Could you elaborate on what you mean when you say that it is not working? Are you able to get a link between devices? Or is this is all a ping test?

Thanks,

Cecilia

0 Chester Kronke over 3 years ago in reply to Cecilia Reyes

Prodigy 230 points

Cecilia,

Thanks for your reply. I use the ping test to verify end-to-end connectivity. If ping works, I then know that the intended data transport for our application will work. If ping fails, the application will fail as well.

Sorry but I'm not certain what "get a link between devices" means exactly. Do you mean the link status LED indicators? Unfortunately, the EVM is not providing a link status LED indicator. The commercial media convertors do have a link status LED which indicates a connection is established once I connect the fiber between the SFPs being tested. If I'm misunderstanding this, please let me know.

Best Regards,

Chester

0 Cecilia Reyes over 3 years ago in reply to Chester Kronke

TI__Genius 9920 points

Hi Chester,

Sorry for the confusion, usually when we ensure there is a link between devices we are reading the registers on the DP83869 on the BMSR reg 0x1 to confirm link up or link down.

Do you have the datasheets for the SFPs? One item that we have run into recently with being unable to successfully ping is how the SFP is connected to the SD pin on the PHY. In our next datasheet refresh we highlight that pin 24 is actually an active low signal meaning that the pin is expected to be low when a fiber signal is present. Perhaps if you are not getting a successful ping this could be something that you can verify.

Thanks,

Cecilia

0 Chester Kronke over 3 years ago in reply to Cecilia Reyes

Prodigy 230 points

Celilia,

Thanks again for your reply. I just want to clarify that I am able to get some SFP pairs to work in a the following arrangements:

Using two TI DP83869EVMs as both M/C1 and M/C2.
Using two commercial packaged media converters as both M/C1 and M/C2.
Using one TI DP83869EVM as M/C1 and plugging SFP2 into our custom designed board (who's design is based on EVM).

With the same pair of SFPs that work in the above cases, I cannot get the following to work (yet):

Using one commercial media converter as M/C1 and one TI DP83869EVM as M/C2.
Using one commercial media converter as M/C1 and plugging SFP2 into our custom designed board.

Wouldn't the pin 24 issue exist in all cases above? I mean if the SFP works in the 3 cases shown above, doesn't this mean the SD pin is not an issue?

Anyway, I will try to check both the link status by reading the registers and I will also try to examine the state of Pin 24 SD signal and get back to you.

Thanks again!

Best Regards,

Chester

0 Cecilia Reyes over 3 years ago in reply to Chester Kronke

TI__Genius 9920 points

Hi Chester,

Thanks for the clarity on the cases.

Please let me know when you are able to provide the register reads and if you are able to see a link.

Non working scenarios:

1. Using one commercial media converter as M/C1 and one TI DP83869EVM as M/C2.

2. Using one commercial media converter as M/C1 and plugging SFP2 into our custom designed board.

In scenario 1, what is the speed configured for both M/C1 and the EVM? One other thing we can try and test for link is to see whether you can configure the OP mode of the EVM from media converter mode to 1000Base-FX to RGMII. Can you see if you are able to get a link from that?

Thanks,

Cecilia

0 Cecilia Reyes over 3 years ago in reply to Cecilia Reyes

TI__Genius 9920 points

Hi Chester,

Can you please also provide what fiber transceivers you are using and the datasheets? Are they the same transceivers for the commercial media converter that you are using for the EVMs as well?

Thanks,

Cecilia

0 Chester Kronke over 3 years ago in reply to Cecilia Reyes

Prodigy 230 points

Cecelia,

For clarity I'll itemize each of the test arrangement cases I described previously:

Case A: Using two TI DP83869EVMs as both M/C1 and M/C2.

Case B: Using one TI DP83869EVM as M/C1 and plugging SFP2 into our custom board.

Case C: Using two commercial packaged media converters as both M/C1 and M/C2.

Case D: Using one commercial media converter as M/C1 and one TI DP83869EVM as M/C2.

Case E: Using one commercial media converter as M/C1 and plugging SFP2 into our custom board.

Our custom board is basically the same as an EVM media converter except that we have dedicated the wired side of the DP83869HM PHY interface to connect directly to our video module daughter card instead of having an UTP RJ45 jack that connects to a test PC node.

In all test cases, I test using a pair of identical model SFPs, i.e., the same model SFP on each side of the center fiber or wired link between the media converters. Also, in all cases, the UTP wired connections to the test PC nodes goes through the unmanaged 10/100BASE-T switch to force the wired link speed down to 100Mbps. This is the same speed that the video daughter card also uses for its wired Ethernet connection.

All of the following SFP models have worked in some of the above arrangement cases:

1. IPOLEX - Model: ASF85-24-X2 - 1000BASE-SX 850nm MMF Dual-LC - www.ipolex.com/goods-67-ASF85-24-X2-D.html

2. INTELLINET - Model: 545006 - 1000BASE-SX 850nm MMF Dual-LC - intellinetsolutions.com/intellinet-en-gigabit-fiber-sfp-optical-transceiver-module-545006.html

3. INTELLINET - Model: 506724 - 1000BASE-LX 1310nm SMF Dual-LC - intellinetnetwork.eu/intellinet-en-gigabit-fiber-sfp-optical-transceiver-module-506724.html

4. UBIQUITI - Model: UF-MM-1G - 1250 Mbps, 850nm MMF Dual-LC - www.ui.com/.../

5. UBIQUITI - Model: UF-MM-10G - 10 Gbps, 850nm MMF Dual-LC - Same link as above model.

6. UBIQUITI – Model: UF-SM-10G - 10 Gbps, 1310n SMF Dual-LC - Same link as above model.

7. BEL - Model: SFP-1GBT-05 - 10/100/1000BASE-T, UTP Wired SFP w/RJ45 (SGMII) - www.belfuse.com/.../ds-bps-sfp-1gbt-05-series.pdf

SFP models 1 through 6 work in arrangement cases A, B, & C. SFP model 7 works in cases A & B only. None of these SFP models have worked in arrangement Cases D or E. Data sheets or general technical information for each model is available at the URL links provided above.

Yesterday, I had many issues getting the MSP driver and USB2MDIO GUI installed and working. I was finally successful installing the driver and GUI on a Windows 10 Tablet PC. For SFP #2, when I read the contents of BMSR Register (0001), I get 796D. I get the same value from the EVM as M/C2 regardless of whether M/C1 is an EVM (i.e., a working arrangement) or a commercial packaged media converter (a non-working arrangement).

I was going to check status of SD line as you suggested, but I don’t find an SD signal to check. Pin 24 on DP83869HM is JTAG TDI in the EVM and does not connect to SFP socket.

Bes Regards,

Chester

0 Chester Kronke over 3 years ago in reply to Cecilia Reyes

Prodigy 230 points

Cecelia,

For clarity I’m itemizing each of the test arrangement cases I described previously:

Case A: Using two TI DP83869EVMs as both M/C1 and M/C2.

Case B: Using one TI DP83869EVM as M/C1 and plugging SFP2 into our custom board.

Case C: Using two commercial packaged media converters as both M/C1 and M/C2.

Case D: Using one commercial media converter as M/C1 and one TI DP83869EVM as M/C2.

Case E: Using one commercial media converter as M/C1 and plugging SFP2 into our custom board.

Our custom board is basically the same as an EVM media converter except we dedicate the wired Ethernet side of the DP83869 PHY interface to connect to our video module daughter card instead of an RJ45 for UTP connection to a test PC node.

In all test cases, I test using a pair of identical model SFPs, i.e., the same model SFP on each side for the center fiber or wired link between the media converters.

All of the following SFP models have worked in some of the above arrangement cases:

1. IPOLEX - Model: ASF85-24-X2 - 1000BASE-SX 850nm MMF Dual-LC - www.ipolex.com/goods-67-ASF85-24-X2-D.html

2. INTELLINET - Model: 545006 - 1000BASE-SX 850nm MMF Dual-LC - intellinetsolutions.com/intellinet-en-gigabit-fiber-sfp-optical-transceiver-module-545006.html

3. INTELLINET - Model: 506724 - 1000BASE-LX 1310nm SMF Dual-LC - intellinetnetwork.eu/intellinet-en-gigabit-fiber-sfp-optical-transceiver-module-506724.html

4. UBIQUITI - Model: UF-MM-1G - 1250 Mbps, 850nm MMF Dual-LC - www.ui.com/.../

5. UBIQUITI - Model: UF-MM-10G - 10 Gbps, 850nm MMF Dual-LC - Same link as previous model.

6. UBIQUITI – Model: UF-SM-10G - 10 Gbps, 1310n SMF Dual-LC - Same link as previous model.

7. BEL - Model: SFP-1GBT-05 - 10/100/1000BASE-T, UTP Wired SFP w/RJ45 (SGMII) - www.belfuse.com/.../ds-bps-sfp-1gbt-05-series.pdf

SFP models 1 through 6 work in arrangement cases A, B, & C. SFP model 7 works in cases A & B only. None of these SFP models have worked in arrangement Cases D or E. Data sheets or technical details on each model are available at the URL links listed above.

I was going to check status of SD line as you suggested, but I don’t find an SD signal to check. Pin 24 on DP83869HM is JTAG TDI in the EVM and does not connect to SFP socket.

0 Chester Kronke over 3 years ago in reply to Cecilia Reyes

Prodigy 230 points

Cecelia,

For clarity I’m itemizing each of the test arrangement cases I described previously:

Case A: Using two TI DP83869EVMs as both M/C1 and M/C2.

Case B: Using one TI DP83869EVM as M/C1 and plugging SFP2 into our custom board.

Case C: Using two commercial packaged media converters as both M/C1 and M/C2.

Case D: Using one commercial media converter as M/C1 and one TI DP83869EVM as M/C2.

Case E: Using one commercial media converter as M/C1 and plugging SFP2 into our custom board.

In all test cases, I test using a pair of identical model SFPs, i.e., the same model SFP on each side for the center fiber or wired link between the media converters.

All of the following SFP models have worked in some of the above arrangement cases:

1. IPOLEX - Model: ASF85-24-X2 - 1000BASE-SX 850nm MMF Dual-LC - www.ipolex(dot)com/goods-67-ASF85-24-X2-D.html

2. INTELLINET - Model: 545006 - 1000BASE-SX 850nm MMF Dual-LC - intellinetsolutions(dot)com/intellinet-en-gigabit-fiber-sfp-optical-transceiver-module-545006.html

3. INTELLINET - Model: 506724 - 1000BASE-LX 1310nm SMF Dual-LC - intellinetnetwork(dot)eu/intellinet-en-gigabit-fiber-sfp-optical-transceiver-module-506724.html

4. UBIQUITI - Model: UF-MM-1G - 1250 Mbps, 850nm MMF Dual-LC - www.ui(dot)com/accessories/fiber-modules-cable/

5. UBIQUITI - Model: UF-MM-10G - 10 Gbps, 850nm MMF Dual-LC - Same link as previous model.

6. UBIQUITI – Model: UF-SM-10G - 10 Gbps, 1310n SMF Dual-LC - Same link as previous model.

7. BEL - Model: SFP-1GBT-05 - 10/100/1000BASE-T, UTP Wired SFP w/RJ45 (SGMII) - www.belfuse(dot)com/resources/datasheets/powersolutions/ds-bps-sfp-1gbt-05-series.pdf

I was going to check status of SD line as you suggested, but I don’t find an SD signal to check. Pin 24 on DP83869HM is JTAG TDI in the EVM and does not connect to SFP socket.

0 Cecilia Reyes over 3 years ago in reply to Chester Kronke

TI__Genius 9920 points

Hi Chester,

JTAG_TDI actually has a net to R22 going to the SFP_LOS pin (page 2 of EVM schematic)

Glad to hear you were able to get the USB2MDIO software set up. So it looks like even in a working or non-working scenario, reg 0x1 reads 796D meaning there is a link (Referring to Bit 2 LINK_STS1 of Reg 0x1).

In the test you conducted yesterday, which of these model #s outlined did you use for SFP1 and SFP2?

Thanks,

Cecilia

0 Cecilia Reyes over 3 years ago in reply to Cecilia Reyes

TI__Genius 9920 points

With your GUI up and running properly, could you now also provide registers 0-1F?

0 Chester Kronke over 3 years ago in reply to Cecilia Reyes

Prodigy 230 points

Cecilia,

Sorry for the triple posting of my previous reply. The first two didn't seem to get posted and I figured it was due the the URL links. Thus, I edited to make the URLs not appear as direct links. When I posted that last version, it showed up here right away. I see now (much later) that my first 2 reply attempts were only delayed in getting posted. Unfortunately the interface here wprovides no feedback as to what is going on when it is delayed like that.

Anyway, did I miss anything you needed from me? Is there anything else I need to try or investigate in my test setup that may help?

Best Regards,

Chester

0 Cecilia Reyes over 3 years ago in reply to Chester Kronke

TI__Genius 9920 points

Hi Chester,

No problem. My message may have gotten lost in the 3 replies:

Copied from previous reply-

JTAG_TDI actually has a net to R22 going to the SFP_LOS pin (page 2 of EVM schematic) so you can try to probe the signal there

In the test you conducted yesterday, which of these model #s outlined did you use for SFP1 and SFP2?

With your GUI up and running properly, could you now also provide registers 0-1F? (Working scenario and non-working scenario)

Thanks,

Cecilia

0 Chester Kronke over 3 years ago in reply to Cecilia Reyes

Prodigy 230 points

Cecilia,

It took a lot longer to get the register data than I expected. GUI kept freezing up and having to be restarted but I was eventually able to collect all the data.

By the way, I noticed a few register's who's data varied from first read to an accidental subsequent read. This is why a few register reads are repeated below.

Yesterday, I tried several different pairs of SFPs but I wasn't seeing a difference in the BMSR (01) register data value of 796D. All of the data below was collected using a pair of Intellinet 545006 SFPs which is also one of the ones I tested yesterday.

I checked JTAG_TDI state and found it to be low (0V) on EVM2 in Test Arrangement Case D and on both EVM1 & EVM2 in Case A.

Thanks!

Chester

---------------------------------------------------------------------------

With IMC1:SFP1<:::::>SFP2:EVM2 (Test Arrangement Case D - ping not working)
the following is register data from EVM2...

-----------------------------------------------------------------------

With EVM1:SFP1<:::::>SFP2:EVM2 (Test Arrangement Case A - ping working)
the following is register data from EVM2...

The following is register data from EVM1...

0 Cecilia Reyes over 3 years ago in reply to Chester Kronke

TI__Genius 9920 points

Hi Chester,

Thanks for sharing the registers. Our USB2MDIO user's guide has useful info on how to script the reads for these registers. Was the GUI still causing you issues if you loaded a script to the register reads?

It looks like with based on both scenarios for the copper side you are able to see a link by register 0x1. Now we can review the link between the fiber ends.

1) On the EVM or by register read do you have link loss pass through enabled?

2) Could you read 0xC00 through 0xC19? These are our register controls for the fiber portion of the DP83869. If you are still running into issues where this may take too long to record all the data then we can take a look at just:

- 0xC00

- 0xC01

- 0xC04

- 0xC05

Thanks,

Cecilia

0 Chester Kronke over 3 years ago in reply to Cecilia Reyes

Prodigy 230 points

Cecilia,

I have now tried Test Arrangement Case D both with and without Link Loss jumper installed on the EVM. To be clear the jumper I’m referring to is labeled as LINK_LOSS on EVM board and is just to the left of RXDO test point via. In the DP83869EVM User’s Guide on page 5, this jumper position is labeled OP_MD[0] but this is obviously incorrect. In addition, schematic in EVM UG uses only Jxx designators and several of the jumpers on the board do not include silkscreen Jxx labels or it is not clear which Jxx label applies to which jumper.

For reading the register data from the EVM with USB2MDIO, using script files helped but I can only get one script file execution to work per USB2MDIO connection session. Subsequent attempts at using the same or a different script file always fails and I end up having to abort the GUI, restart it, reconfigure & re-establish connection, reset the PHY & mode, then I can again load & run a script (but again, only once).

I'm attaching files with the read data that I captured. The long filenames should indicate the test case that applies for the data but I've also added comments at the top of each file to clarify each test case.

The commercial EVMs I have are INTELLINET Model 510493 and 10GBIT Model A7S2-33-1GX1GT-SFP/HDW1. Both of these use the Qualcom QCA8334-AL3C chip (4-port Gigabit Ethernet Switch IC). I have two of each of these models so that I can test them as matched M/C pairs in my test arrangement Case C. I also have a single ADNET AN-UM-SMC-AS-J media converter which uses RealTek RTL8213 IC.

Please let me know the M/Cs that have actually been tested and are known to work with the DP83869HM. Please include specific brand and model numbers and I will see about getting a pair of each for our testing.

Thanks!

Best Regards,

Chester

EVM1 Data 0000..001F - MC1&2_EVM - SFP1&2_Intellinet545006 - Working.txt

//BatchRead-0000..001F applied to EVM as MC1 in Test Arrangement Case A:
//
//  M/C1 & M/C2: Both EVM	SFP Pair: Intellinet 545006	Ping: Success
//
BatchRead-0000..001F.txt file is open...
Register 0000 is: 1140
Register 0001 is: 796D
Register 0002 is: 2000
Register 0003 is: A0F1
Register 0004 is: 01E1
Register 0005 is: 4DE1
Register 0006 is: 0067
Register 0007 is: 2001
Register 0008 is: 0000
Register 0009 is: 0300
Register 000A is: 0000
Register 000B is: 0000
Register 0000 is: 1140
Register 000D is: 0000
Register 000E is: 0000
Register 000F is: F000
Register 0000 is: 1140
Register 0011 is: 7C02
Register 0012 is: 0000
Register 0013 is: 1C42
Register 0014 is: 29C7
Register 0015 is: 0000
Register 0016 is: 0000
Register 0017 is: 0040
Register 0018 is: 6150
Register 0019 is: 4444
Register 001A is: 0002
Register 001B is: 0000
Register 001C is: 0000
Register 001D is: 0000
Register 001E is: 0012
Register 001F is: 0000

End of file.

EVM1 Data 0C00..0C1F - MC1&2_EVM - SFP1&2_Intellinet545006 - Working.txt

//BatchRead-0C00..0C1F applied to EVM as MC1 in Test Arrangement Case A:
//
//  M/C1 & M/C2: Both EVM	SFP Pair: Intellinet 545006	Ping: Success
//
BatchRead-0C00..0C1F.txt file is open...
Register 0C00 is: 1140
Register 0C01 is: 796D
Register 0C02 is: 2000
Register 0C03 is: A0F1
Register 0C04 is: 01E1
Register 0C05 is: 4DE1
Register 0C06 is: 0065
Register 0C07 is: 2001
Register 0C08 is: 0000
Register 0C09 is: 0300
Register 0C0A is: 0000
Register 0C0B is: 0000
Register 0C0C is: 0000
Register 0C0D is: 0000
Register 0C0E is: 0000
Register 0C0F is: F000
Register 0C00 is: 1140
Register 0C11 is: 6C02
Register 0C12 is: 0000
Register 0C13 is: 0000
Register 0C14 is: 29C7
Register 0C15 is: 0000
Register 0C16 is: 0000
Register 0C17 is: 0040
Register 0C18 is: 6150
Register 0C19 is: 4444
Register 0C1A is: 0002
Register 0C1B is: 0000
Register 0C1C is: 0000
Register 0C1D is: 0000
Register 0C1E is: 0012
Register 0C1F is: 0000

End of file.

EVM2 Data 0000..001F - MC1&2_EVM - SFP1&2_Intellinet545006 - Working.txt

//BatchRead-0000..001F applied to EVM as MC2 in Test Arrangement Case A:
//
//  M/C1 & M/C2: Both EVM	SFP Pair: Intellinet 545006	Ping: Success
//
BatchRead-0000..001F.txt file is open...BatchRead-0000..001F.txt file is open...
Register 0000 is: 1140
Register 0001 is: 796D
Register 0002 is: 2000
Register 0003 is: A0F1
Register 0004 is: 01E1
Register 0005 is: CDE1
Register 0006 is: 006D
Register 0007 is: 2001
Register 0008 is: 6001
Register 0009 is: 0300
Register 000A is: 0000
Register 000B is: 0000
Register 0000 is: 1140
Register 000D is: 0000
Register 000E is: 0000
Register 000F is: F000
Register 0000 is: 1140
Register 0011 is: 6F02
Register 0012 is: 0000
Register 0013 is: 0000
Register 0014 is: 29C7
Register 0015 is: 0000
Register 0016 is: 0000
Register 0017 is: 0040
Register 0018 is: 6150
Register 0019 is: 4444
Register 001A is: 0002
Register 001B is: 0000
Register 001C is: 0000
Register 001D is: 0000
Register 001E is: 0012
Register 001F is: 0000

End of file.

EVM2 Data 0C00..0C1F - MC1&2_EVM - SFP1&2_Intellinet545006 - Working.txt

//BatchRead-0C00..0C1F applied to EVM as MC2 in Test Arrangement Case A:
//
//  M/C1 & M/C2: Both EVM	SFP Pair: Intellinet 545006	Ping: Success
//
BatchRead-0C00..0C1F.txt file is open...
Register 0C00 is: 1140
Register 0C01 is: 796D
Register 0C02 is: 2000
Register 0C03 is: A0F1
Register 0C04 is: 01E1
Register 0C05 is: CDE1
Register 0C06 is: 006F
Register 0C07 is: 2001
Register 0C08 is: 6001
Register 0C09 is: 0300
Register 0C0A is: 0000
Register 0C0B is: 0000
Register 0C0C is: 0000
Register 0C0D is: 0000
Register 0C0E is: 0000
Register 0C0F is: F000
Register 0C00 is: 1140
Register 0C11 is: 7C02
Register 0C12 is: 0000
Register 0C13 is: 1C42
Register 0C14 is: 29C7
Register 0C15 is: 0000
Register 0C16 is: 0000
Register 0C17 is: 0040
Register 0C18 is: 6150
Register 0C19 is: 4444
Register 0C1A is: 0002
Register 0C1B is: 0000
Register 0C1C is: 0000
Register 0C1D is: 0000
Register 0C1E is: 0012
Register 0C1F is: 0000

End of file.

EVM2 Data 0000..001F - MC1_Intellinet510493-MC2_EVM - SFP1&2_Intellinet545006 - Not working.txt

//BatchRead-0000..001F applied to EVM as MC2 in Test Arrangement Case D:
//
//  M/C1: Intellenet 5104932	SFP Pair: Intellinet 545006	Ping: FAILS
//
BatchRead-0000..001F.txt file is open...BatchRead-0000..001F.txt file is open...
Register 0000 is: 1140
Register 0001 is: 796D
Register 0002 is: 2000
Register 0003 is: A0F1
Register 0004 is: 01E1
Register 0005 is: CDE1
Register 0006 is: 006D
Register 0007 is: 2001
Register 0008 is: 6001
Register 0009 is: 0300
Register 000A is: 0000
Register 000B is: 0000
Register 0000 is: 1140
Register 000D is: 0000
Register 000E is: 0000
Register 000F is: F000
Register 0000 is: 1140
Register 0011 is: 6F02
Register 0012 is: 0000
Register 0013 is: 0000
Register 0014 is: 29C7
Register 0015 is: 0000
Register 0016 is: 0000
Register 0017 is: 0040
Register 0018 is: 6150
Register 0019 is: 4444
Register 001A is: 0002
Register 001B is: 0000
Register 001C is: 0000
Register 001D is: 0000
Register 001E is: 0012
Register 001F is: 0000

End of file.

EVM2 Data 0C00..0C1F - MC1_Intellinet510493-MC2_EVM - SFP1&2_Intellinet545006 - Not working.txt

//BatchRead-0C00..0C1F applied to EVM as MC2 in Test Arrangement Case D:
//
//  M/C1: Intellenet 5104932	SFP Pair: Intellinet 545006	Ping: FAILS
//
BatchRead-0C00..0C1F.txt file is open...
Register 0C00 is: 1140
Register 0C01 is: 796D
Register 0C02 is: 2000
Register 0C03 is: A0F1
Register 0C04 is: 01E1
Register 0C05 is: CDE1
Register 0C06 is: 006D
Register 0C07 is: 2001
Register 0C08 is: 6001
Register 0C09 is: 0300
Register 0C0A is: 0000
Register 0C0B is: 0000
Register 0C0C is: 0000
Register 0C0D is: 0000
Register 0C0E is: 0000
Register 0C0F is: F000
Register 0C00 is: 1140
Register 0C11 is: 6F02
Register 0C12 is: 0000
Register 0C13 is: 0000
Register 0C14 is: 29C7
Register 0C15 is: 0000
Register 0C16 is: 0000
Register 0C17 is: 0040
Register 0C18 is: 6150
Register 0C19 is: 4444
Register 0C1A is: 0002
Register 0C1B is: 0000
Register 0C1C is: 0000
Register 0C1D is: 0000
Register 0C1E is: 0012
Register 0C1F is: 0000

End of file.

0 Cecilia Reyes over 3 years ago in reply to Chester Kronke

TI__Genius 9920 points

Hi Chester,

Thanks for sharing your feedback. I'll take a look what we can do to optimize some of the silk screening on our EVMs for better readability.

After reading both test cases it still looks to be that there is a link on the fiber side. (Reading register 0xC01 bit 2)

As mentioned in our call can you please try testing with 1000Mbps MC mode then read the 0x0C00 registers again and also try the ping again?

There are step by step instructions on how to configure to this Media Converter mode:

http://www.ti.com/lit/an/snla318/snla318.pdf

Section 1.6 should help you set this up.

In parallel I will try and find the SFPs and Media Converters our team has used in the past that have worked for our EVMs.

0 Chester Kronke over 3 years ago in reply to Cecilia Reyes

Prodigy 230 points

Cecilia,

I tried following the instructions under section 1.6 - 1000Mbps Media Converter of the Understanding different modes of operation in DP83869 guide but had issues. First, I setup hardware using my test arrangement case A (i.e., using two EVMs for both M/C1 & M/C2). Again, I want to make certain I get working arrangement between the EVMs as M/Cs before testing with Commercial EVMs. However, once I've set the jumpers on the EVMs as described in section 1.6, when i read register 01DF, I get F000 instead of the expected 0044. Also, when I first read register 01EC it returns 0000. If I write 1FFC to 01EC as prescribed, the GUI responds with "Register 01EC is 1FFC" however subsequent read back returns "Register 01EC is 0000". I get the same result on both EVMs.

The only difference I saw between the two EVMs was that LED2 was illuminated on one and off on the other.

Anyway, I tried the ping test despite the issues above and this test failed as well. I tried this using both 100 Mbs & 1000 Mps speeds on the wired side. Without the EVM-to-EVM case working, I did not bother trying to test Commercial M/C to EVM case.

Any ideas of what I may be doing wrong regarding the Section 1.6 instructions? Or, what to try next?

Best regards,

Chester

0 Cecilia Reyes over 3 years ago in reply to Chester Kronke

TI__Genius 9920 points

Hi Chester,

Did this issue with the reading and writing only occur after you reconfigured the jumpers? Could you confirm some of the other registers like 0x0 - 0x1F were also not showing any issues?

One idea we can try is to put the jumpers back in the original settings and do the configuration through register writes?

• Write 0x0044 to register 0x01DF

• Write 0x1140 to register 0x0000

• Write 0x5048 to register 0x0010

• Write 0x1140 to register 0x0C00

• Write 0x1FFC to register 0x01EC (sets bit [0] to 0)

• Write 0x4000 to register 0x001F

Let me know if this works for you.

Thanks,

Cecilia

0 Chester Kronke over 3 years ago in reply to Cecilia Reyes

Prodigy 230 points

Cecilia Reyes said:

Did this issue with the reading and writing only occur after you reconfigured the jumpers? Could you confirm some of the other registers like 0x0 - 0x1F were also not showing any issues?

I followed the instructions in the order that they were written in the manual, so I configured the jumpers first and then tried the registers reads & writes. I have now returned the EVMs to the previous condition (as best I can tell) and there's really no change in what I read from registers 01DF & 01EC:

I also tried writing to register 01EC before changing the jumpers and when I do, I again still see the same write and post-write read-back results:

// Writing 0x1FFC to register 01EC:
Register 01EC is: 1FFC
// Reading register 01EC...
Register 01EC is: 0000

The above is all still using the pair INTELLINET 545006 SFPs. There was no change in the operability before to after writing to 01EC. These SFPs work together with the EVMs at either 100M or 1G speeds on the wired connections. However, I cannot mix wired speeds on opposite ends in this test arrangement. If I attempt to run one EVM with wired connection speed of 100M and the other EVM with 1G wired connection, the ping tests fail. The speed of wired connection is verified both by LED indicator on EVMs and the NIC status on the PC nodes.

One thing I did discover was the cause of the LED indication discrepancy between the 2 EVMs. It was because my shorting jumper for LED_0 on one EVM wasn't shorting as it should have been. With that corrected, both EVMs show the same LED indications.

I was about to repeat everything but before attempting anything further, I did some checking of the documentation, specifically on register 01EC. Please note that this register is not listed in the Register Maps of the DP83869 data sheet. Also, that section of the data sheet specifically states:

All register offset addresses not listed in Table 19 should be considered as reserved locations and the register contents should not be modified.

However, I also find in the 9.4.7 Media Converter Mode section:

In Managed mode, Media Convertor can still be activated via straps but register configuration option are also used. When Media Convertor Mode is enabled then bit [0] of register 0x01EC should be set to 0.

I would also like to direct your attention to section 9.4.8 Register Configuration for Operational Modes, which is supposed to list all register configurations necessary for each mode. Here it states specifically…

9.4.8.4 1000M Media Convertor Mode

After configuring register 0x01DF, perform the following operations.
• Write 0x1140 to register 0x0000
• Write 0x5048 to register 0x0010
• Write 0x1140 to register 0x0C00 9 <<< What is the “9” here? Another mistake I guess.

This information obviously conflicts with the instructions given in section 1.6 of the Modes of Operation guide and the previous section 9.4.7 of the data sheet.

So, I’m lost. Is register 01EC, the correct register or not? If it is, what is its total description since this register is not listed in the Register Map? Shouldn’t reading back the register after writing to it reflect what was written? Also, register address 0x1DF is “OP_MODE_DECODE Register”. Since this register is supposed to reflect the current mode of operation, I’m confused by the fact that its value is “F000” despite changing the operation mode with jumper selections and also because reading it does not agree with what is specified in section 1.6 of the modes guide.

I guess I will try the different registers writes listed above from instructions in data sheet section 9.4.8.4. . I will let you know the results. Also, depending on results, I will try the more detailed register configuration you provide above. If possible, please provide better guidance for anything dealing with registers 01DF & 01EC, since what I’m seeing isn’t matching instructions and/or there is conflicting information given.

By the way, I also received two Perle Model S-1110-SFP Media Converters that are known to work with the DP83869 and fiber SFP. I have once again had mixed results with these new M/Cs but still am not the able to get these to work in the arrangement we require for our application. I will post a separate description on my testing & results with the Perle devices.

Thanks!

0 Chester Kronke over 3 years ago in reply to Chester Kronke

Prodigy 230 points

The following tests were performed using a pair of INTELLINET 545006 1000BASE-SX 850nm MMF SFPs. Always common for the configuration of the EVMs are the following jumpers: Installed shorting jumpers: J2:1-2, J4:1-2, J7:1-2, & J8; Open jumpers: J2:2-3, J4:2-3, J7:2-3, & J6

Results for testing with Perle S-1110-SFP M/Cs as commercial packaged media converters:

Test 1 – MIXED RESULTS: Wired speeds must match on both sides to PASS – Conditions:

Test Arrangement Case A
EVMs Jumpers: OPMODE_0(J19)=open, OPMODE_1(J11)=shorted, OPMODE_2(J10)=shorted

A. Wired connection speed 100M (both sides) - Ping works both ways.
B. Wired connection speed 1G (both sides) - Ping works both ways.
C. Wired connection speed 1G on one side & 100M on opposite side – Ping fails (timed out) both ways.

Test 2 – PASSED: All wired speeds tested worked including mixed speed – Conditions:

Test Arrangement Case C
Perle as both M/C1 & M/C2 with switch settings all UP.

Test 3 – MIXED RESULTS – Conditions:

Test Arrangement Case D
Perle M/C1 switch settings (all UP except SW2 down).
NOTE: Perle M/C1 w/all switches UP always fails w/ping response of “Destination host unreachable”.
EVM as M/C2 jumpers: OPMODE_0=open, OPMODE_1=shorted, OPMODE_2=shorted

A. Wired connection speed 100M (both sides) - Ping fails (timed out) both ways.
B. Wired connection speed 1G (both sides) - Ping works both ways.
C. Wired speed 1G at M/C1 & 100M at M/C2 - Ping fails (timed out) both ways.

Our application requires passing Test 3A or 3C, both of which are failing.

EVM register reads were captured for all cases above where an EVM is used. The captured data is provided in the files attached to this post. Files are named according to the test number with the case letter and M/C number that applies for each file.

2020-05-20 - Test 1A - MC1 Registers.txt

BatchRead-Main & Fiber Status Registers.txt file is open...
Register 0000 is: 1140

Register 0001 is: 796D

Register 0002 is: 2000

Register 0003 is: A0F1

Register 0004 is: 01E1

Register 0005 is: CDE1

Register 0006 is: 006F

Register 0007 is: 2001

Register 0008 is: 6001

Register 0009 is: 0300

Register 000A is: 0000

Register 000D is: 0000

Register 000E is: 0000

Register 000F is: F000

Register 0010 is: 5048

Register 0011 is: 7C02

Register 0012 is: 0000

Register 0013 is: 1C42

Register 0014 is: 29C7

Register 0015 is: 0000

Register 0016 is: 0000

Register 0017 is: 0040

Register 0018 is: 6150

Register 0019 is: 4444

Register 001A is: 0002

Register 001E is: 0012

Register 001F is: 0000

Register 0C00 is: 1140

Register 0C01 is: 796D

Register 0C02 is: 2000

Register 0C03 is: A0F1

Register 0C04 is: 01E1

Register 0C05 is: CDE1

Register 0C06 is: 006D

Register 0C07 is: 2001

Register 0C08 is: 6001

Register 0C18 is: 6150

Register 0C19 is: 4444

End of file.

2020-05-20 - Test 1A - MC2 Registers.txt

BatchRead-Main & Fiber Status Registers.txt file is open...
Register 0000 is: 1140

Register 0001 is: 7969

Register 0002 is: 2000

Register 0003 is: A0F1

Register 0004 is: 01E1

Register 0005 is: 4DE1

Register 0006 is: 0067

Register 0007 is: 2001

Register 0008 is: 0000

Register 0009 is: 0300

Register 000A is: 0000

Register 000D is: 0000

Register 000E is: 0000

Register 000F is: F000

Register 0010 is: 5048

Register 0011 is: 7F02

Register 0012 is: 0000

Register 0013 is: 5D42

Register 0014 is: 29C7

Register 0015 is: 0000

Register 0016 is: 0000

Register 0017 is: 0040

Register 0018 is: 6150

Register 0019 is: 4444

Register 001A is: 0002

Register 001E is: 0012

Register 001F is: 0000

Register 0C00 is: 1140

Register 0C01 is: 796D

Register 0C02 is: 2000

Register 0C03 is: A0F1

Register 0C04 is: 01E1

Register 0C05 is: 4DE1

Register 0C06 is: 0065

Register 0C07 is: 2001

Register 0C08 is: 0000

Register 0C18 is: 6150

Register 0C19 is: 4444

End of file.

2020-05-20 - Test 1B - MC1 Registers.txt

BatchRead-Main & Fiber Status Registers.txt file is open...
Register 0000 is: 1140

Register 0001 is: 7969

Register 0002 is: 2000

Register 0003 is: A0F1

Register 0004 is: 01E1

Register 0005 is: CDE1

Register 0006 is: 006F

Register 0007 is: 2001

Register 0008 is: 6001

Register 0009 is: 0300

Register 000A is: 0000

Register 000D is: 0000

Register 000E is: 0000

Register 000F is: F000

Register 0010 is: 5048

Register 0011 is: 7F02

Register 0012 is: 0000

Register 0013 is: 5D42

Register 0014 is: 29C7

Register 0015 is: 0000

Register 0016 is: 0000

Register 0017 is: 0040

Register 0018 is: 6150

Register 0019 is: 4444

Register 001A is: 0002

Register 001E is: 0012

Register 001F is: 0000

Register 0C00 is: 1140

Register 0C01 is: 796D

Register 0C02 is: 2000

Register 0C03 is: A0F1

Register 0C04 is: 01E1

Register 0C05 is: CDE1

Register 0C06 is: 006D

Register 0C07 is: 2001

Register 0C08 is: 6001

Register 0C18 is: 6150

Register 0C19 is: 4444

End of file.

2020-05-20 - Test 1B - MC2 Registers.txt

BatchRead-Main & Fiber Status Registers.txt file is open...
Register 0000 is: 1140

Register 0001 is: 796D

Register 0002 is: 2000

Register 0003 is: A0F1

Register 0004 is: 01E1

Register 0005 is: CDE1

Register 0006 is: 006F

Register 0007 is: 2001

Register 0008 is: 6001

Register 0009 is: 0300

Register 000A is: 3800

Register 000D is: 0000

Register 000E is: 0000

Register 000F is: F000

Register 0010 is: 5048

Register 0011 is: BC02

Register 0012 is: 0000

Register 0013 is: 1C42

Register 0014 is: 29C7

Register 0015 is: 0000

Register 0016 is: 0000

Register 0017 is: 0040

Register 0018 is: 6150

Register 0019 is: 4444

Register 001A is: 0002

Register 001E is: 0012

Register 001F is: 0000

Register 0C00 is: 1140

Register 0C01 is: 796D

Register 0C02 is: 2000

Register 0C03 is: A0F1

Register 0C04 is: 01E1

Register 0C05 is: CDE1

Register 0C06 is: 006D

Register 0C07 is: 2001

Register 0C08 is: 6001

Register 0C18 is: 6150

Register 0C19 is: 4444

End of file.

2020-05-20 - Test 1C - MC1 (100M) Registers.txt

BatchRead-Main & Fiber Status Registers.txt file is open...
Register 0000 is: 1140

Register 0001 is: 796D

Register 0002 is: 2000

Register 0003 is: A0F1

Register 0004 is: 01E1

Register 0005 is: CDE1

Register 0006 is: 006F

Register 0007 is: 2001

Register 0008 is: 6001

Register 0009 is: 0300

Register 000A is: 3800

Register 000D is: 0000

Register 000E is: 0000

Register 000F is: F000

Register 0010 is: 5048

Register 0011 is: BC02

Register 0012 is: 0000

Register 0013 is: 1C42

Register 0014 is: 29C7

Register 0015 is: 0000

Register 0016 is: 0000

Register 0017 is: 0040

Register 0018 is: 6150

Register 0019 is: 4444

Register 001A is: 0002

Register 001E is: 0012

Register 001F is: 0000

Register 0C00 is: 1140

Register 0C01 is: 796D

Register 0C02 is: 2000

Register 0C03 is: A0F1

Register 0C04 is: 01E1

Register 0C05 is: CDE1

Register 0C06 is: 006D

Register 0C07 is: 2001

Register 0C08 is: 6001

Register 0C18 is: 6150

Register 0C19 is: 4444

End of file.

2020-05-20 - Test 1C - MC2 (1G) Registers.txt

BatchRead-Main & Fiber Status Registers.txt file is open...
Register 0000 is: 1140

Register 0001 is: 796D

Register 0002 is: 2000

Register 0003 is: A0F1

Register 0004 is: 01E1

Register 0005 is: CDE1

Register 0006 is: 006F

Register 0007 is: 2001

Register 0008 is: 6001

Register 0009 is: 0300

Register 000A is: 7800

Register 000D is: 0000

Register 000E is: 0000

Register 000F is: F000

Register 0010 is: 5048

Register 0011 is: BF02

Register 0012 is: 0000

Register 0013 is: 1C42

Register 0014 is: 29C7

Register 0015 is: 0000

Register 0016 is: 0000

Register 0017 is: 0040

Register 0018 is: 6150

Register 0019 is: 4444

Register 001A is: 0002

Register 001E is: 0012

Register 001F is: 0000

Register 0C00 is: 1140

Register 0C01 is: 796D

Register 0C02 is: 2000

Register 0C03 is: A0F1

Register 0C04 is: 01E1

Register 0C05 is: CDE1

Register 0C06 is: 006D

Register 0C07 is: 2001

Register 0C08 is: 6001

Register 0C18 is: 6150

Register 0C19 is: 4444

End of file.

2020-05-20 - Test 3A - MC2 (100M) Registers.txt

BatchRead-Main & Fiber Status Registers.txt file is open...
Register 0000 is: 1140

Register 0001 is: 796D

Register 0002 is: 2000

Register 0003 is: A0F1

Register 0004 is: 01E1

Register 0005 is: 4DE1

Register 0006 is: 0067

Register 0007 is: 2001

Register 0008 is: 0000

Register 0009 is: 0300

Register 000A is: 0000

Register 000D is: 0000

Register 000E is: 0000

Register 000F is: F000

Register 0010 is: 5048

Register 0011 is: 7C02

Register 0012 is: 0000

Register 0013 is: 1C42

Register 0014 is: 29C7

Register 0015 is: 0000

Register 0016 is: 0000

Register 0017 is: 0040

Register 0018 is: 6150

Register 0019 is: 4444

Register 001A is: 0002

Register 001E is: 0012

Register 001F is: 0000

Register 0C00 is: 1140

Register 0C01 is: 796D

Register 0C02 is: 2000

Register 0C03 is: A0F1

Register 0C04 is: 01E1

Register 0C05 is: 4DE1

Register 0C06 is: 0065

Register 0C07 is: 2001

Register 0C08 is: 0000

Register 0C18 is: 6150

Register 0C19 is: 4444

End of file.

2020-05-20 - Test 3B - MC2 (1G) Registers.txt

BatchRead-Main & Fiber Status Registers.txt file is open...
Register 0000 is: 1140

Register 0001 is: 796D

Register 0002 is: 2000

Register 0003 is: A0F1

Register 0004 is: 01E1

Register 0005 is: CDE1

Register 0006 is: 006D

Register 0007 is: 2001

Register 0008 is: 6001

Register 0009 is: 0300

Register 000A is: 38FF

Register 000D is: 0000

Register 000E is: 0000

Register 000F is: F000

Register 0010 is: 5048

Register 0011 is: BF02

Register 0012 is: 0000

Register 0013 is: 1D42

Register 0014 is: 29C7

Register 0015 is: 0000

Register 0016 is: 0000

Register 0017 is: 0040

Register 0018 is: 6150

Register 0019 is: 4444

Register 001A is: 0002

Register 001E is: 0012

Register 001F is: 0000

Register 0C00 is: 1140

Register 0C01 is: 796D

Register 0C02 is: 2000

Register 0C03 is: A0F1

Register 0C04 is: 01E1

Register 0C05 is: CDE1

Register 0C06 is: 006D

Register 0C07 is: 2001

Register 0C08 is: 6001

Register 0C18 is: 6150

Register 0C19 is: 4444

End of file.

2020-05-20 - Test 3C - MC2 (100M) Registers.txt

BatchRead-Main & Fiber Status Registers.txt file is open...
Register 0000 is: 1140

Register 0001 is: 7969

Register 0002 is: 2000

Register 0003 is: A0F1

Register 0004 is: 01E1

Register 0005 is: 4DE1

Register 0006 is: 0067

Register 0007 is: 2001

Register 0008 is: 0000

Register 0009 is: 0300

Register 000A is: 0000

Register 000D is: 0000

Register 000E is: 0000

Register 000F is: F000

Register 0010 is: 5048

Register 0011 is: 7F02

Register 0012 is: 0000

Register 0013 is: 1C42

Register 0014 is: 29C7

Register 0015 is: 0000

Register 0016 is: 0000

Register 0017 is: 0040

Register 0018 is: 6150

Register 0019 is: 4444

Register 001A is: 0002

Register 001E is: 0012

Register 001F is: 0000

Register 0C00 is: 1140

Register 0C01 is: 796D

Register 0C02 is: 2000

Register 0C03 is: A0F1

Register 0C04 is: 01E1

Register 0C05 is: 4DE1

Register 0C06 is: 0065

Register 0C07 is: 2001

Register 0C08 is: 0000

Register 0C18 is: 6150

Register 0C19 is: 4444

End of file.

0 Cecilia Reyes over 3 years ago in reply to Chester Kronke

TI__Genius 9920 points

Hi Chester - thanks for sharing these tests. Is this the first time you were able to see a ping between a M/C and the DP83869EVM with 1G? (Test 3B)

I am currently working with our validation team to get their feedback and see if they can verify the tests. One comment they did have when I communicated to them your test procedure is that they have SW2 and SW3 down on Perle. Can you please see if these settings change?

Otherwise, I am getting their feedback and will give you more information tomororw.

Thank you,

Cecilia

0 Cecilia Reyes over 3 years ago in reply to Cecilia Reyes

TI__Genius 9920 points

Hi Chester,

We noticed in the Perle installation guide that: "These products convert a 10/100/1000Base-T cable connection (copper) to a 1000Base-X connection (fiber)"

Per your message it does look like you were able to ping between Perle MC1 and EVM MC2 at 1G. When you tested SW2 up did you try it with both speeds on the EVM? And another test from my previous comment is in regards to toggling SW3 as well.

Thanks,

Cecilia

0 Chester Kronke over 3 years ago in reply to Cecilia Reyes

Prodigy 230 points

Cecelia,

For previous testing I did, I only changed SW2 on Perle M/C for Test 3 - Arrangement Case D (i.e., Perle as M/C1 and EVM as M/C2). I have always had all other switches (including SW3) UP.

Yes, using the Perle M/C is the first time I have had any success with fiber SFPs in Test Arrangement Case D. Unfortunately, it only worked with 1G wired speed on DP83869 side. For our application, we require that the DP83869 side (i.e., EVM as M/C2) work at wired speeds of 10/100M since the target daughter card only supports 10/100M.

Regarding...

Yes, in Test 3 with all Perle switches up, I could not get ping test to work. I tried at both 100M & 1G wired speeds on both sides (i.e., 100M <=> 100M, 1G <=> 1G, 1G <=> 1M, and 1M <=> 1G ALL failed). In my tests, I power off the Perle M/C1 when switching the wired connection speeds. This is because I observed that at power-up I initially see all LEDs on Perle come on except LKF but approximately 18 seconds later 100/1000 & FDC LEDs both turn off and LKC starts blinking slowly.

In repeating the Test 3 wired speed tests this morning (to verify the LED states for each), I accidentally discovered that ping works in certain wired speeds briefly, basically while 100/1000, FDC, & LKC are ON. Investigating further I see the brief window of pings working for 100M <=>1G and 1G <=> 1G cases (i.e., works when EVM wired speed is 1G) but fails for 100M <=> 100M and 1G ó 100M (i.e., fails when wired speed for EVM is 100M). After the 18 seconds where LEDs change, no case works. Again, this is with Perle switches ALL UP.

With this accidental discovery, I repeated Test 3 with Perle SW2 down. I found that as long as EVM wired speed is 1G the pings work and when EVM wired speed in 100M, the pings fail. This is the same as the brief working period for All switches UP except now the working/non-working state and LED ON/OFF states are maintained indefinitely. To me this seems to indicate the problem is on the DP83869 / EVM side.

I repeated the tests with both SW2 and SW3 DOWN on Perle and all other switches UP, and results are same as the SW2 DOWN and all others UP cases, i.e, ping works only when EVM wired speed is 1G.

QUESTION: Is the DP83869, in any mode of operation, capable of translating fiber connection speeds >100Mbps down to 100Mbps on wired interface? If not, can we force fiber speeds to be 100M?

I’m also still awaiting on answers to my questions regarding registers 0x1DF & 0x1EC and the discrepancies in the documentation and instructions.

In the meantime, am turning my attention to testing only 100M wired speeds on both sides of the connection. We need wired connection of DP83869 to work at 10/100BASE-T/TX. I am currently going back through all the fiber SFPs I have to verify which ones work in Test Arrangement Case A only with wired speeds on both sides at 100M. Then I will repeat testing all SFPs that work in Case A, in Test Arrangement Case C (again only at 100M wired speeds). And finally, I’ll repeat testing the working SFPs for Cases A & C in Test Arrangement Case D. I’ll post the results here after I’m done.

Thanks,
Chester

0 Chester Kronke over 3 years ago in reply to Chester Kronke

Prodigy 230 points

Cecelia,

Restricting my testing to only wired speeds of 100Mbps on both sides and only using Perle S-1110-SFP as the commercial M/Cs, I have tested 7 different SFP models that all work in Test Arrangement Cases A and C but fail in Case D. The 7 SFP models are of 3 different types:

Group A: Four 1000Base-SX MM dual-fiber SFPs.
Group B: Two 1000Base-LX SM dual-fiber SFPs.
Group C: One 1000Base-BX SM single-fiber BiDi SFP.

The SFPs in Groups A & B are interchangeable within each group. By this I mean any Group A SFP will work with any other Group A SFP at the opposite end. Likewise, either Group B SFP will work with any Group B SFP on the opposite end.

I also tested 4 other SFP models that have not yet worked with the EVMs. I’m listing these as Group F for now. See the attached PDF for all the specific model fiber SFPs I have that were tested.

Thanks,
Chester

6786.2020-05-22 SFPs Tested with MCs at 100Mbps Wired Speeds.pdf

0 Chester Kronke over 3 years ago in reply to Cecilia Reyes

Prodigy 230 points

Cecelia,

I want to emphasize the importance of our need for corrected information on the register reads and writes that are described & noted as required in the Application Note “Understanding different modes of operation in DP83869” document. Have you been able to find out anything more on this issue?

Since the information provided does not match what I’m seeing, I cannot follow the instructions to test some of the modes described. For register 0x1DF reads I always get 0xF000 and never see the values that the app note says I should see. Also, regardless of what I write to register 0x1EC, I only ever get 0 when I read back the register. Thus, I’m precluded from setting the DP83869HM into the 100 Mbps Media Converter mode of operation. I believe this is the mode we actually need to get the 100Base-FX SFPs to work.

Thanks,
Chester

0 Cecilia Reyes over 3 years ago in reply to Chester Kronke

TI__Genius 9920 points

Hi Chester,

I am currently investigating whether or not there are discrepancies with the app note and seeing if there is a reason as to why you are unable to read and write to register 0x1DF and 0x1EC. I don't believe we had issues with our media converter using this app note in the past.

Just to confirm, you are using the extended registers and still have to read/write registers 0x0-0x1F properly? Are both EVMs reading improperly? If you configure to another mode through straps are you still unable to properly read register 0x1DF?

And to your question:

QUESTION: Is the DP83869, in any mode of operation, capable of translating fiber connection speeds >100Mbps down to 100Mbps on wired interface? If not, can we force fiber speeds to be 100M?

The DP83869 must be forced to the same speed as the wired interface. This is also done through the op mode straps or register writes.

0 Chester Kronke over 3 years ago in reply to Cecilia Reyes

Prodigy 230 points

Cecilia,

Regarding your questions...

Re: Just to confirm, you are using the extended registers and still have to read/write registers 0x0-0x1F properly?

I'm taking this as two questions:

1. "Am I using extended registers?" - I take this to mean "Am I setting the Extended Register option to "Yes" in GUI?" The answer is no. I did not realize I was supposed to be using extended register option. When I initially had issues with USB-2-MDIO, I had tried multiple times with this option set both ways. And when I finally did start getting results, it was with Extended Register option set to No. Thus, I just kept using it that way. All the register read data I have posted on this thread prior to this point has been with Extended Register option set to No. Also, since I was getting data & GUI did not show errors, I assumed this was the proper way to do it. I also reviewed the information on page 40 of data sheet regarding Extended Registers. At that time I missed that Extended Register R/Ws are required for any registers other than 0..31. I mistakenly thought the Extended Register R/Ws only dealt with reading and writing registers with indirect and/or post address increment methods. Also, the Register Maps section of the data sheet makes no distinction of Extended vs. Standard registers.

2. Not sure what you mean by the last part "... and still have to read/write 0x0-0x1F properly?" Are you saying I have to use Standard register R/Ws (i.e., via normal direct-MDIO access) or that I should use Extended mode (i.e., via MDIO Manageable Device indirect method)?

Re: Are both EVMs reading improperly? If you configure to another mode through straps are you still unable to properly read register 0x1DF?

I believe I had already provided these answers in my previous posts but yes, I was getting the same results with both EVMs and changing mode straps did not affect results from reading 0x1DF.

Retrying using Extended Register mode, I get better results yet still not completely agreeing with what is described in the Understanding Modes of Operation in DP83869 Application Report guide. The 0x1EC writes and read backs look good but the 0x1DF register reads always have result of 000_ instead of the 004_ results prescribed (e.g., I get 0004 instead of 0044, 0005 instead of 0045, and 0006 instead of 0046).

For 1.6 1000 Mbps Media Converter mode (page 8):

Register 01DF is: 0004     // EXTENDED READ - Note that this disagrees with 0x0044 that guide says to expect.
Register 01EC is: 1FFD    // EXTENDED READ - Read value before I attempt to write.
Register 01EC is: 1FFC    // EXTENDED WRITE - Result displayed for the write operation.
Register 01EC is: 1FFC    // EXTENDED READ - Read back correct value after the write operation.
Register 01DF is: 0004     // EXTENDED READ - Still get 0x0004 instead of 0x0044 expected value.

For 1.7 100 Mbps Media Converter mode (page 9):

Register 01DF is: 0005     // EXTENDED READ - Note that this disagrees with 0x0045 that guide says to expect.
Register 01EC is: 1FFD    // EXTENDED READ - Read value before I attempt to write.
Register 01EC is: 1FFC    // EXTENDED WRITE - Result displayed for the write operation.
Register 01EC is: 1FFC    // EXTENDED READ - Read back correct value after the write operation.
Register 01DF is: 0005     // EXTENDED READ - Still get 0x0005 instead of 0x0045 expected value.

For 1.8 SGMII-to-Copper (10Base-Te/100Base-TX/1000Base-T) mode (page 10):

Register 01DF is: 0006 // EXTENDED READ - Note that this disagrees with 0x0046 that guide says to expect.
Register 01EC is: 1FFD // EXTENDED READ - Read value only, this mode does not require a write to 0x1EC.

Since I’m am now getting the correct 1EC settings data, I retested the 100BASE-FX & 100BASE-LX SFPs I have. I was able to only get the AVAGO HFBR-57E5APZ SFPs to work with EVMs (Test Arrangement Case A) for both the 100 Mbps Media Converter mode and the SGMII-to-Copper mode. I note that in the cases that worked, both LED_0 and LED_1 are ON and LED_2 blinks when there is data traffic. Unfortunately, neither of those two modes on EVM will work with any of the 4 different commercial SFPs that I have (i.e., Test Arrangement Case D and with wired speeds at 100M on both sides).

Regarding the following previous comment you made on the Perle S-1110-SFP on May 22…

We noticed in the Perle installation guide that: "These products convert a 10/100/1000Base-T cable connection (copper) to a 1000Base-X connection (fiber)"

On the Perle website, the S-1110-SFP is listed as a “Media and Rate Converter 10/100/1000Base-T to 100/1000Base-X Conversion” and specifically states these two features:

10/100/1000Base-T to 1000Base-X or 100Base-X Fiber SFP Media Converters
Connect 10/100 devices to Fast Ethernet or Gigabit backbone

Still, I’m wondering if we need to exchange or replace our Perle S-1110-SFP Media Converters with something like:
SR-100-SFP, S-110P-SFP, or S110P-SFP-XT.

0 Cecilia Reyes over 3 years ago in reply to Chester Kronke

TI__Genius 9920 points

Hi Chester,

Glad to see you were able to get the proper register reads in and it seems like you are actually in the correct modes.

Based on the order of the app note, the reason the document describes you being able to read 0x44 on reg 0x1DF is because the tests were done in a sequential order on modes. If you see section 1.4 and 1.5 you can see that there was a shift in testing the SGMII to RGMII bridge which is what drives bit 6 of 0x1DF high. In that section the document stated to strap RX_CTRL high therefore keeping it in SGMII to RGMII bridge mode even in the next section, 1000M MC mode. That is why we expected the 0x44 in this case.

However, bit 6 can be ignored in your scenario since we only care about the op mode selection which looks to be correctly strapped when switching between 100M and 1000M.

I think it would be best to contact the Perle team to know exactly how you can verify what mode you are selecting since both the EVM and Perle need to be configured at the same speed. Please let us know as it would be helpful for us to understand how exactly to confirm on the fiber side.

Thanks,

Cecilia

0 Cecilia Reyes over 3 years ago in reply to Cecilia Reyes

TI__Genius 9920 points

Hi Chester,

We heard from our field team you were able to get a successful ping for 100M. Could you please share the status and changes you made to achieve the communication?

Thanks,

Cecilia

0 Chester Kronke over 3 years ago in reply to Cecilia Reyes

Prodigy 230 points

Cecilia,

Yes, we have finally been able to get both the EVM and our custom board design to work with Perle Media Converters (M/Cs) at the required 100Mpbs wired speeds. For the full solution, it seems it was a combination of getting appropriate configuration settings for both the Perle as M/C1 and for DP83869HM on the M/C2 end, plus for our custom board, fixing one signal connection error.

Keep in mind that I have always been discussing here my efforts on getting/proving a working solution for our application on our custom board by using the EVM as the DP83869HM side of the target case. Thus, the signal connection error on our custom design was not a contributing factor for the previous tests & register data that has been discussed thus far.

As I noted in my previous message, I was at first only able to get AVAGO HFBR-57E5APZ SFPs to work with EVMs (Test Arrangement Case A) for both the 100 Mbps Media Converter mode and the SGMII-to-Copper mode but could not get those same SFPs to work with any of my commercial M/Cs (Test Arrangement Case D) for our required 100mbps speed on wired side of DP83869HM. However, Monday June 1st, I received the following from Dennis Oyama, a Perle Technical Sales Engineer, regarding the Perle S-1110-SFP M/C:

This media converter does support both 1000base-X and 100base-X modules … the Install Guide is in error not to mention this.

By default the 10/100/1000base-T copper port will negotiate with the attached device to match speed/duplex.

So if the copper device has a 10/100base-TX copper interface, they should both negotiate 100Mbps/full duplex.

… from the email trail and forum discussion, it appears that that does not work ? (<==<< I confirmed.)

[Device + 10/100TX] --- copper – [10/100/1000TX + S-1110 + 100base-FX] --- mmf --- [100base-FX + Device]

I do not see a problem with the above setup.

Note: you can force the S-1110-SFP to a fixed rate of 100base-TX on the copper by setting the following DIP switches:

1 - down (auto neg off

4 - up (full duplex)

5 - up (100 Mbps)

And so, I changed the Perle M/C config switch 1 to the down position, and then the Avago SFP worked in Case D at the required 100Mbps speed, which was the first time I had a commercial M/C working with DP83869 at the 100 Mbps wired speed we need.

However, switching to our end goal Test Arrangement Case E (i.e., replacing M/C2, USW2, & PC2 with our target custom board), the link still failed. At this point I knew there was no good reason it shouldn’t work, so after much effort (verifying jumpers, verifying all hardware still working, etc.), I finally discovered a design error. The required connection of JTAG_TDI from DP83869 to SFP_LOS pin 5 of SFP socket was missing. I added this signal connection to our custom board and the target case finally worked. I also retested other 100Base-xx Fast Ethernet SFPs and the Cisco GLC-FE-100FX, also a 100Base-FX device same as the AVAGO, now also works. I have yet to get 100Base-SX or -LX SFPs I have to work.

In the end, the following are the configuration settings that are currently working for the 100Base-FX SFPs:

Perle S-1110-SFP as M/C1:

Configuration switches 1, 2, & 3 down, and all others up. All internal jumpers in their factory default state.

DP83860HM on EVM or our custom board:

                        Installed jumpers for OPMODE_0, OPMODE_2, & MIRROR_EN.
                        Removed jumpers for OPMODE_1 and LINK_LOSS.
                        Set 3 pin jumpers to 1 (HIGH pull-up side) for ANEGG_DIS, ANEDSEL_0, and ANEGSEL_1.
                        All registers in their default state (i.e., I did not need to change any register settings).

Note also that with the above configuration I get good LED status indications for M/C and DP83869, namely:

Perle S-1110-SFP M/C: 100/1000 LED Yellow (100 Mbps) and all other LEDs green. LKF & LKC blink with traffic.

DP83869HM: LED_0, _1, and _2 all on and LED_2 blinks with traffic.

Yesterday (6/8), I also received and tested the Perle S-110P-SFP 100Base-X Fast Ethernet Media Converter. This M/C also worked with the 100Base-FX SFPs I have but does not yet work with the -SX or -LX devices. This is with all configuration options of the Perle S-110P-SFP in the factory default configuration (all external switches UP and all internal options unchanged).

I am still hoping we can get other SFPs to work, especially the 100Base-SX and -LX I already have. I was surprised these did not work. Any advice?

0 Cecilia Reyes over 3 years ago in reply to Chester Kronke

TI__Genius 9920 points

Hi Chester - I am glad to hear that both EVM and custom design have been able to work. I will need to reach out to our validation team to see if they have any comments of your recent test conditions.

Thanks,

Cecilia

Interface

Interface forum

DP83869HM: Used as Media Converter from wired PHY to SFP - Works only with EVM but not with other external Media Converters.