DP83867IRPAP-EVM: DP83867IRPAP-EVM Register Access through USB-2-MDIO

Hi Gerome

Thank you for the answer.

You specifically mentioned  MSP430F5529LP.  May I know the reason why specifically  MSP430F5529LP ? and why not  MSP-EXP430G2ET ? 

Would you point me a TI Document that describe the overall connections between MSP430* launchpad and EVM board ?

Do I need any drivers to be installed for this EVM board ? 

Thanks

Krish

Hi Gerome

You mentioned about newer EVMs have builtin IC MSP430F5529.

But, In my case, i have DP83867IRPAP-EVM and not sure it is a newer EVM or not.

So, i still have to connect MSP430F5529 with my board DP83867IRPAP-EVM externally ?

Pls confirm MSP430F5529 launch PAD is compatible with with board DP83867IRPAP-EVM.and connection need to be done externally.

Pls confirm the above before i purchase them.

Thank You.

Krish

Hello Gerome

1.  Connected pins 4.1 (MDC) and 4.2 (MDIO) to the respective parts in the EVM

2. I did flash the Launch PAD with the text file that you sent in the 1st post in here. => Successfully done.

3. Tried writing/Reading the Registers using USB-2-MDIO

4. But, I am getting the "Device not Connected" Message when I read or write register. 

5. How do we fix the "Device not connected" Issue ? 

Hi Krish,

Please ensure that the PHY is on and connected to the launchpad before powering on the launchpad via USB.

In addition, I noticed that the VDDIO of the EVM is at 2.5V. With the launchpad being 3.3V, there is a conflict of VDDIO domains. To bridge this gap, can you please disconnect the 3.3V header in the middle of the launchpad, and then attach a power supply to the 3.3V pin at the bottom of the launchpad with this supply set to 2.5V. This will make the launchpad 2.5V VDDIO compatible.

Sincerely,

Gerome

Hi Gerome

    Yes, I did power up EVM - PHY First before the Launch PAD.  But, still getting "Device not connected".

Are there any jumpers need to be removed in the Launch PAD (MSP430F5529) as per the following TI guide ?

Hello,

Edited my response above for 2.5V operation.

Sincerely,

Gerome

Hi Gerome

    Need explanation on your last answer.

>>In addition, I noticed that the VDDIO of the EVM is at 2.5V. With the launchpad being 3.3V, there is a conflict of VDDIO domains. To bridge this gap, can >>you please disconnect the 3.3V header in the middle of the launchpad, and then attach a power supply to the 3.3V pin at the bottom of the launchpad > >> with this supply set to 2.5V. This will make the launchpad 2.5V VDDIO compatible.

1. Launchpad is already powered by 5v power supply at the top port through USB.  Do you want me provide another additional 3.3v power supply to the right bottom pins ?   Two different power supplies ( 5v and 3.3v) at the same time to the launchpad ?  

2. As per the following diagram, the 3.3v power supply is derived (internal to launchpad) from 5v power supply and the 3.3v is supplied to the Target Device (MSP430F) automatically by the launchpad.  Then,  why do we have to re-supply 3.3v at the PIN level to this launchpad while 5v is already supplied through USB.

Page 10 from MSP430F5529 LaunchPad Development Kit (MSP-EXP430F5529LP) User's Guide (Rev. D) (ti.com)

Thanks

Krish

Hi Krish,

While it may seem unorthodox, this is the correct interpretation. In your diagram, you can see the jumper in red is the 3.3V getting supplied by the LDO on top of the launchpad. We would like to remove this jumper to cut this connection off. Instead, we would like to power 2.5V from an external supply to the 3.3V header connected to the target MCU. This can be done via the 3.3V supply pin at the bottom right side of the board. Again, as the 3.3V LDO is cut off, this voltage will only be supplied from the external power supply. This is done to manipulate the VDDIO domain of the MCU to be 2.5V instead of 3.3V, thus matching the VDDIO of the PHY.

Sincerely,

Gerome

HI Gerome

I thought I need 3.3v power supply to be applied to 3.3v supply pin at the bottom right side of the board.

But, as per your last answer, I need 2.5v power supply to be applied to the 3.3v supply pin ? Pls confirm that I need to buy 2.5v power supply ? 

Can you pls suggest the correct part number of this power supply that specially fit into the PINs at the bottom of the board. 

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

Also, I want to know that Is this unorthodox method is due to the MSP430F5529 launchpad ?   

Can you pls suggest me another launchpad where I do not have to go through this unorthodox method. 

Hi Krish,

This is true, the 2.5V supply needs to be connected to the 3.3V pin to supply 2.5V VDDIO to the MCU. 

Any typical variable lab power supply should be sufficient.

Sincerely,

Gerome

Hi Gerome

     Thank you for clarification.

I am still wondering why this unorthodox way of connecting external power supply WHILE we already provide 5v through USB to the Launchpad.

Pls kindly explain the following questions. 

1. Is this external power supply of 2.5v is needed due to this specific EVM such as DP83867IRPAP where VDDIO is 2.5v ?  

2. Does the EVM such as DP83867ERGZ - R also have VDDIO of 2.5v ?  Even if I bought DP83867ERGZ EVM , I would have needed this special 2.5 v power supply ? 

3. Let me know what combination of EVMs (?) and Launch PAD (?) would work at the same VDDIO with out the need of this unorthodox way of connecting external power supply ? 

4.  The 2.5v Power supply is hard to get and could not find one at amazon. Does the 3v or 3.3v external Power Supply would work at LaunchPAD ?  

Thanks

Krish

Hi Krish,

Yes, this is due to matching the VDDIO of the DP83867IRPAP EVM.

Sincerely

Gerome

____

Update 1/25/24: Hi Krish, I would like to correct my previous statement where VDDIO of the RGZ EVM is 3.3V. It is actually 2.5V. However, I have personally tested the launchpad as is with the EVM and the connections can work as is, but may possibly need to go through the iteration of power cycling the board and launchpad multiple times for the FW to register the PHY to allow access.

You can also utilize a secondary option; DP83TC812EVM-MC. This is an all-in-one option which can be supplied by a single USB supply and have register access via that same cable. The only drawback is that the GPIO is a NC, so to probe you would need to directly probe on the PHY pin.

I apologize for the confusion and frustration this entire process has caused. 

Sincerely,

Gerome

Hi Gerome

   I just saw your Corrected Version of the answer. 

1. You said "VDDIO of the RGZ EVM is 3.3V. It is actually 2.5V".   Does it mean that I still need to apply 2.5 v externally to the LaunchPAD MSP430F5529 even for DP83867ERGZ EVM ?? 

2. You said  " However, I have personally tested the launchpad as is with the EVM and the connections can work as is, but may possibly need to go through the iteration of power cycling the board and launchpad multiple times for the FW to register the PHY to allow access. " 

Can you pls point me the TI Spec for the above statement of how to do the the Power Cycling. 

3.  Is this Power Cycling Process is ONLY applicable for RGZ EVM & MSP430F5529 ?   This process is not needed for DP83867IRPAP. ? 

4. I am looking for EVM board that have the feature of => SFD IEEE 1588 timestamp.  I have only 2 options. => DP83867ERGZ  Or   DP83867IRPAP.  I have the LaunchPAD MSP430F5529.

These EVM and Launch PAD are suggested by TI Technical Support Team. Now, I am EVM and LaunchPAD suggested by TI Technical team are not working together or not compatible and making the entire process more confusion and frustration.

Pls suggest me what is the correct Procedure and what is the correct combination of EVM and Launchpad. 

Please answer the above questions as well.

Hi Gerome, Waiting for your answer for my last post. 

Most importantly, For the Combination of (DP83867ERGZ EVM and MSP430F5529) => Do we still need to apply external 2.5v power supply to MSP430F5529  Or, it will be work fine with just 5v power supply through USB. ? 

Hi Krish,

Using 5V supply via USB should be sufficient for this, but there may need to be some iterations in power cycling to ensure the MSP is able to properly recognize the PHY. Previously, you had expressed difficulty in obtaining a 2.5V supply so I am trying to give options to work around this.

If you are looking for an all-in-one solution to evaluate the PHY's SFD on a GPIO while also powering and accessing registers via USB, I would recommend checking out DP83TC812EVM-MC as this does have this capabilities, albeit with the GPIO pin as a NC so you would need to probe directly.

There are also other EVMs in our DP83867IR product page which also have a MCU with our PHY for proper register access.

I apologize for the frustration this has caused. I have noted this for next revision of these EVMs.

Sincerely,
Gerome

Hi Gerome

        Thank you for your answer.

You suggested  DP83TC812EVM-MC as all in one solution.  But, I reviewed the Spec for this EVM.  But, it does not explicitly say that it supports start-of-frame detect IEEE 1588 timestamp.

Can you please point me the TI Specification for DP83TC812EVM-MC where it says it supports the following.

• SFD IEEE 1588 timestamp

Hi Gerome

   As per your suggestion, I bought DP83TC812EVM-MC recently.

I tested write/read for the PHY registers on the board and it works fine. That is great. 

Specifically, I want to detect SFD on DP83TC812EVM-MC and should be able to check visually either through PINs/GPIO.

The following TI Specification shown below ( Section 4, Page 3) has the procedure to detect the SFD specifically for EVMs such as DP83867IRPAP,  DP83867ERGZ-S,  DP83867ERGZ-R.   

How to Configure DP83867 Start of Frame Detect (ti.com)

I tried exact the same procedure on this new board DP83TC812EVM-MC  and it does Not seem to work for DP83TC812EVM-MC. 

1. What registers ( address, data) that need to be programmed in order to detect the SFD on this DP83TC812EVM-MC.  ?   

2. Once Step (1) is completed, What PINs or GPIO that I can check SFD detection ? 

3. Pls let me know the complete procedure to detect SFD specific to DP83TC812EVM-MC.

Look forward to hear from you as early as possible as my project depending on this SFD detection. 

Thanks

Krish

Hi Krish,

This is odd as our team had recently tested this on the ERGZ EVM with no issue. The register configurations were per the app note, but the same settings should apply regardless of board.

The procedure also requires data to be sent on the MDI so it would pass through the PHY, with the probe on GPIO_0 pin of DP83867.

Can you please provide your procedure that you are conducting? What PHY ID? Picture of setup where the probe is connected to the pin.

Sincerely,

Gerome

HI Gerome

     The register programming that you sent was  for the ERGZ EVM.

But, in my last post, I asked the register programming for new board DP83TC812EVM-MC ( Not  ERGZ EVM).

Pls let me know the correct register programming procedure and Prob setup to detect SFD on DP83TC812EVM-MC.

I have only ONE DP83TC812EVM-MC.  The packet generator can be connected to the EVM through Ethernet cable.

Exactly which PINs on DP83TC812EVM-MC need to be probed to detect the SFD ? 

Is the following test setup is enough to detect the SFD ? 

Packet generator => DP83TC812EVM-MC => Probes ?  

Hi Krish,

This would still work as the PHY is the same; DP83867. You will need to probe on the DP83867 pins directly per the app note.

Sincerely,

Gerome

HI Gerome

Just for your Information :  Your previous answer is not correct and does not applicable to DP83TC812EVM-MC.  

Correct Procedure for DP83TC812EVM-MC.  

1. Writing 0x1080 to Register 0x04A0 

2. Writing 0x0406 to Register 0x0452  <=  New Information.

3. Software restart : Writing 0x4000 to Register 0x001F

Hi Krish,

Which PHY are you writing to on this board?

Sincerely,

Gerome

Hi Gerome

The SFD register for the DP83867 is different than  the one in DP83TC812. 

Section 8.6.2.44 IO_MUX_CFG_1 Register 0x0452 is the register for configuring SFD for DP83TC812.

Hi Krish,

I am curious as to why you are looking at the DP83TC812 when all of the prior debug was geared towards the DP83867. The board has both PHYs and thus you can ignore the 100Base-T1 portion of the board for all intents and purposes of your debug.

Sincerely,

Gerome