| ??? 10/07/05 16:32 Read: times |
#102103 - It's a combination of both Responding to: ???'s previous message |
Jan said:
Ehm, I faintly remember somebody saying here recently that the rise/fall times count more than the actual clock frequency, when it comes to noise... It's a combination of both, frequency of fundamental and rise time: Having a square wave the harmonics follow a slope of 20db/decade. Above a frequency of 1/pi/risetime the harmonics follow a slope of again 20db/decade, giving additonally a slope of 40dB/decade. Noise is now formed by the whole band made of fundamental and harmonics. So, if you make a plot of wide band noise with a fundamental of 7MHz and compare it with that of a 22MHz fundamental with same rise time, then you will notice a certain lower noise with the 7MHz fundamental. On the other hand, if you really want to drastically decrease wide band noise you should also increase rise time of signals. This can be done by inserting suited low pass filters in the lines running to ADCs/DACs, which round the edges a bit and by this decrease the noise. Kai |
| Topic | Author | Date |
| Other crystals? | 01/01/70 00:00 | |
| my experience | 01/01/70 00:00 | |
| dallas | 01/01/70 00:00 | |
| RTFM | 01/01/70 00:00 | |
| the speedy '450 | 01/01/70 00:00 | |
| why 8MHz? | 01/01/70 00:00 | |
| crystals | 01/01/70 00:00 | |
| '420 | 01/01/70 00:00 | |
| errata? | 01/01/70 00:00 | |
| 22.1184MHz | 01/01/70 00:00 | |
| since when? | 01/01/70 00:00 | |
| the tiny asterisk... | 01/01/70 00:00 | |
| Baud rates | 01/01/70 00:00 | |
| crystals, but problem fixed.... sorta :) | 01/01/70 00:00 | |
| 10472 baud | 01/01/70 00:00 | |
| but Atmel | 01/01/70 00:00 | |
| 8 bits autobaud | 01/01/70 00:00 | |
| explanation | 01/01/70 00:00 | |
| dallas | 01/01/70 00:00 | |
| 7.3728MHz | 01/01/70 00:00 | |
| Lower clock frquency | 01/01/70 00:00 | |
| backslashes | 01/01/70 00:00 | |
| New Server Problems? | 01/01/70 00:00 | |
| lol | 01/01/70 00:00 | |
| Go on laughing... | 01/01/70 00:00 | |
| rise/fall time? | 01/01/70 00:00 | |
| that is absolutely correct, but what mak | 01/01/70 00:00 | |
| jetplanes | 01/01/70 00:00 | |
| It's a combination of both | 01/01/70 00:00 | |
| thread | 01/01/70 00:00 | |
This is for the XA which has a UART that | 01/01/70 00:00 |