DP83846A: can't access from CPU

Hi Shinichi,

Can you describe how often the fault state is seen in power-up? Is the device fully functional in other power-ups? 

Can you include the difference in the current consumption between a good and faulty start-up, and broken down by supply? Please also include the operational mode you are using the PHY in and any bootstrap settings you have enabled. 

Please note the DP83846 is not recommended for new designs. Can describe the status of this design and reason the DP83846 is selected? We can help to suggest new designs that are robust and cost-effective.

Regards,
Justin 

Hi Justin,

Thank you for your reply and sorry for late reply.

I confirm your comment with the customer.

When I get it, I'll feedback to you.

I appreciate your great help.

Best regards,

Shinichi

Hi Shinichi,

Please help to communicate the DP83846 is not a recommended device due to the release date of the device and support for this device is limited compared to the latest generation of PHYs.

Regards,

Justin 

Hi Justin,

I could obtain the answer from the customer.

Could you please advise from listed below.

---

(1) The frequency of failure at power-on is about once in 10 times.

Other than PHY is operating, there is no response from PHY, so the CPU is waiting.

(2) The current consumption is about 10mA higher in both normal and abnormal cases.

The operation mode of PHY is auto negotiation.

I'm not sure about the boot setting because the OS driver does it.

OS uses TRON.

---

When I could get a new information, I'll send you soon.

I appreciate your great help.

Best regards,

Shinichi

Hi Shinichi,

Does holding the RESET pin low, result in the device going into a low-power mode? Does a hardware reset (holding RESET low for >160us) cause the PHY to re-initialize? 

Can you share the power sequence, ramp times, and delay times between each supply as well? 

I am not sure what you mean current consumption is higher in both normal AND abnormal modes, can you please clarify? Can you also share which supply is seeing the increase in current?

Regards,
Justin 

Hi Justin,

Thank you for your reply.

Regarding current consumption, it increases by 10mA at abnormal condition.

I'll confirm the customer about reset and power sequence and then feedback to you.

I appreciate your great help.

Best regards,

Shinichi 

Hi Shinichi,

Please share the information requested above. A 10mA increase is about the expected current draw of the LEDs so I would not expect that to be a latch-up event.

Regards,
Justin 

Hi Justin,

Thank you for your reply.

I'm waiting for the response from the customer.

When I get it, I'll share with you.

I appreciate your great help.

Best regards,

Shinichi

Hi Justin,

I obtain waveform of power up regarding NG and OK.

It seems to be the same waveforms.

NG waveform

OK waveform

So I requested the customer expand waveforms.

Next, when the latch up occur, reset signal is ignored and LED keeps lighting.

It works when the power is turned off and on.

Customers continue to use the DP83846A on various models, but no problems have occurred.

At this time, the power supply filter was changed from inductor to ferrite beads, and the ceramic capacitor was changed to electrolytic capacitor.

curcuit around VDD.pptx

Could this lead to latch-up?

I'll request the customer change to the inductance and ceramic capacitor fomerly used.

Could you please advise the other reason lead to latch up.

ｰｰｰ

I appreciate your great help and cooperation.

Best regards,

Shinichi

Hi Shinichi,

I don't see a difference between the two power ramps. If the customer uses the the inductor and ceramic capacitor does the issue go away? Can you try slowing the ramp rate of each supply? You can also try reordering the power sequence. 

Unfortunately because this is such an old device, I cannot provide deeper insights since this device is no longer supported. 

Regards,
Justin