Part Number:EK-TM4C1294XL
Hello forum,
Checking two PCB's via datasheet table (27.23) MCU 3v3 ripple >100mV (EVM) against custom PCB, e.g. MOSC 120Mhz @32*C attempting to mitigate increasing DC ripple. Custom PCB powered via liner 24vdc <16mV PP ripple, EVM powered via USB0 (control).
Note when either peripheral is powered yet not connected MCU current draw EVM=102mA, custom PCB=75mA same peripherals configured and powered. Perhaps slight difference results KCPD versus NCPDT PCB layout, e.g, separate DGND/AGND versus EVM single GND layer sandwiched between trace layers.
The trouble both PCB's begins EMAC0 or USB0 plugged to Ethernet switch or USB0 OTG port and MUC 3v3 draw; EVM >150mA, custom PCB 92mA so MCU jumps >17mA on each PCB.
Custom PCB we triple EMAC0 (C18/C22) currently @1uf up to 4.7uf. Fails to reduce Ethernet 3v3 ripple or odd current draw upon Pulse XFMR primary on either PCB. The odd part being increase MCU current just for plugging isolated EMAC0 cable into Pulse XFMR jack. The USB port simply being plugged to computer USB causes little to any added DC ripple. That is until USB0 endpoint client is launched then >200mV ripple of switching spikes appear in the 3v3 already >100mV AC ripple of both EMAC0 and USB0. The Perfect Storm is born, even George Cluny can't save the sinking Andrea Gail.
The 5V buck regulator custom PCB adjusts to current demand MCU upon 3v3 LDO regulator >102mV ripple. Added ferrite beads or even larger bypass caps LDO (TPS73533) fail to reduce >DC ripple of both peripherals.
Has any remedy been discovered or answer as to why EMAC0 current increases just plugging cable, closing loop? Likewise why USB0 bulk device endpoint client simply being launched produce >200mV switching spikes upon 3v3 LDO of either PCB? USB0 OTG port switching spikes (>200mV) seem to be data bus related noise yet are unacceptable ripple levels on either PCB.