board with ADSP-21369 and W5300; transformer Wuerth 749013010 (meets exactly your spec)
W5300 in internal PHY mode (Test-mode = 0000); OP-mode = 000 or 111;
OPEN socket0 in UDP mode (S0_SSR response ok); try to broadcast data
Result: internally, command SEND (or SEND_MAC) is finalized with SEND_OK interrupt; TX_LED is active; but:
external pins TXOP, TXON (and RXIP, RXIN) do not carry any voltage (< 10 mV, but there is NO short-cut to GND)
LINK_LED is not active (and there is really no link)
Switching OP-mode from 000 to 111 increases duration interval between command SEND and Interrupt SEND_OK by factor 13 (in OP-mode 000, duration is 5µs/8 Byte user data, corresponding to about 1Byte/100ns);
but on the outside TXO and RXI pins - no voltage, no link, all the same.
Do you have an idea, what prevents the voltage activation (should be in the order of 1.6V dc) of these PHY pins?
Hi, Wolfagang.
Would you please attach the board schematic and transformer’s datasheet in this post?
(and other material that will help to solve the problem anything)
I can reply after reviewing your materials I guess.
attached are transformer datasheet, Orcad Board design, and the W5300 part as image.
Please, consider, we are not dealing with slightly bad bahaviour, but:
the dc bias at the RX and TX pins is fully missing, all the time, also during switching on, it never appears.
And consequently, also any data modulation never can appear at these pins.
While internally, the software behaves as it would send data.
Some comments to board design: 1.8V is usually taken from an extra supply on the board. But I also switched it to the 1V8O of W5300, no changes.
There is some different with our W5300reference.
I think Pin2, 7 Of Transformer should connect to 3.3VA.
Please try it and compare with w5300 internal phy reference again.(as below link.) wiznet.co.kr/sub_modules/en/ … 12&cType=2
thank you, I will try this amendment in the schematic.
Sorry, but I followed only to “Fig. 19 Internal PHY & LED Signals” in your datasheet, where this additional power supply is not visible.
I’ll come back to you with the experimental results, as soon as I’ve checked your proposition.
Already now, thank you all for your remarkable reliable communication.
Dear Peter,
I added these 2 power supply knots on my board, and it works now.
Thank you, great help! (Although you could add them already to the W5300 data sheet)
Wolfgang.