??? 04/16/05 15:04 Read: times Msg Score: +1 +1 Good Answer/Helpful |
#91727 - Well....Here's One Idea Responding to: ???'s previous message |
Well Well....there is not really any definitive idea or method to solve problems with chips like this one. The biggest problem with these microcontrollers is the multi-use / multi-function pins. The choice of ground plane layout becomes more a function of how the device is used in the circuit as opposed to the chip itself. The SiLabs chips are far superior in the A/D converter inputs area in that they assign these pins as fixed function inputs....i.e. the 12 bit A/D inputs of say an 'F12x part.
So I tried to sketch up a more detailed idea of how you could attack the GND planes for this chip. Keep in mind that I went completely on the sketch info you provided. I did not look in the chip data sheet nor did I have any idea of how you were to use any of the pin connections that you highlited. I think that the XTAL pins are not particularly troublesome if you can handle them similar to what I show. You may have to select a longer bypass cap than you would otherwise use to permit the three connections shown to pass under it at right angles. So here is the split idea: ![]() Bypass caps are shown as the dark brown rectangles. The ferrite that would bridge the analog GND to the DGND could be the violet colored rectangle. If you use a ferrite in this way do note to use care to select something that does not compromise the DC and low frequency behavior of your analog subsystem. Also note that there will be a need to cut out all digital PWR plane whereever there is analog GND plane. You can then fill in an analog PWR plane in this area or double the analog GND in here or leave it open. Michael Karas |
Topic | Author | Date |
WEOT: Ground layout problems. | 01/01/70 00:00 | |
Multi-Layer Boards | 01/01/70 00:00 | |
For Your Chio Example | 01/01/70 00:00 | |
Thats great. | 01/01/70 00:00 | |
Well....Here's One Idea | 01/01/70 00:00 | |
Thanks Michael,![]() | 01/01/70 00:00 |