| ??? 03/09/06 13:38 Read: times |
#111698 - Transmission lines Responding to: ???'s previous message |
Mahmood said:
I thought the purpose of ground plane was to stop signals from outside sources to interfere with the board and to stop signals emissions from the board to the outside. If noise gets generated internaly, how on earth can ground plane stop it? There are very many aspects how a solid ground plane works and helps. One major component of emitted interference is common mode noise, which results from the fact, that there is potenial shifts on ground between different points on the board. The use of a solid ground plane can nearly eliminate this ground noise, at least to a level, where you can pass the CE-tests. A solid ground plane short circuits most of the ground noise! Mahmood said:
Maybe as Kai suggested on top, adding RLC to the vcc pin of each IC or oscillator to stop the leak through power supply line. This is only to avoid that supply voltage line will emit interference too. But it's not a substitute for a solid ground plane. Mahmood said:
Also signals on the same board interfere with each other if tracks are close to each other, remedy is to make 2 ground lines surrounding the track that carries high frequency signals or high current pulses to make interference as small as possible To shield a signal trace by 2 surrounding ground tracks will only help for very slow signals. It doesn't work with your fast clock signals. The only shielding that works is to have a solid ground plane directly under the signal trace, where the distance to ground plane equalls the signal trace width. When dealing with so fast signals, forget that you have currents that flow along signal traces. You don't have any longer simple currents, but electromagnetic wave packets filling the space within transmission lines. Fast signals need always their ground return current "path" nearby. Would you ever split a coax cable and pull appart the shield from the inner wire by several centimeters, when you want it to use for your wireless set or satellite communication? But with the printed circuit board you do it... Mahmood said:
I think the external oscillator is better than internal oscillator in terms of frequency stability, starting and signal levels. But the disadvantage is that it generates amplified signal and the track supplying the frequency is relatively long before it gets to the microcontroller so it acts as antenna transmitting rf. while the internal oscillator has a lower amplitude and since the amplifier is inside the micro, the track is shorter making less emmissions. Exactly! But don't forget the ground noise such a clock line produces! Kai |
| Topic | Author | Date |
| AD9835 noise problem | 01/01/70 00:00 | |
| noise | 01/01/70 00:00 | |
| Recommendations | 01/01/70 00:00 | |
| quick fix? | 01/01/70 00:00 | |
| Maybe | 01/01/70 00:00 | |
| Transmission lines | 01/01/70 00:00 | |
| 5MHz? | 01/01/70 00:00 | |
| Spectrum analyser | 01/01/70 00:00 | |
| It's ground noise | 01/01/70 00:00 | |
| ground noise and groundhogs | 01/01/70 00:00 | |
| Harmonics as well | 01/01/70 00:00 | |
| comments | 01/01/70 00:00 | |
| Why separate the analog & digital gnds? | 01/01/70 00:00 | |
| Splitting the ground plane | 01/01/70 00:00 | |
| Evaluation board | 01/01/70 00:00 | |
| thin ground? | 01/01/70 00:00 | |
| that is why you use star | 01/01/70 00:00 | |
| Ground wire in parallel | 01/01/70 00:00 | |
| and a twisted pair? | 01/01/70 00:00 | |
| Would do! | 01/01/70 00:00 | |
| does not exist, Jan if you used twisted | 01/01/70 00:00 | |
| The clock signal? | 01/01/70 00:00 | |
| oh well | 01/01/70 00:00 | |
| Indeed | 01/01/70 00:00 | |
| is there a way... | 01/01/70 00:00 | |
| Hardly | 01/01/70 00:00 | |
| unfortunately not | 01/01/70 00:00 | |
| twisted pair | 01/01/70 00:00 | |
| is that not obvious | 01/01/70 00:00 | |
| No ground traces, but patches | 01/01/70 00:00 | |
| the bane of many | 01/01/70 00:00 | |
| Why 1 or 2 layers | 01/01/70 00:00 | |
| it's nice! | 01/01/70 00:00 | |
| No Secrets | 01/01/70 00:00 | |
| DALO 2M professional | 01/01/70 00:00 | |
| UV ink Problem | 01/01/70 00:00 | |
| Mixed method? | 01/01/70 00:00 | |
| No Need for DALO | 01/01/70 00:00 | |
| CE | 01/01/70 00:00 | |
| Single step mode | 01/01/70 00:00 | |
| If you want... | 01/01/70 00:00 | |
| mmm | 01/01/70 00:00 | |
| Dreaming | 01/01/70 00:00 | |
| LAYOUT | 01/01/70 00:00 | |
| 100nF caps | 01/01/70 00:00 | |
| 100nF ? | 01/01/70 00:00 | |
| ok | 01/01/70 00:00 | |
| Some checks | 01/01/70 00:00 | |
| It's my daily job! | 01/01/70 00:00 | |
| Unfortunately, no | 01/01/70 00:00 | |
| simulate double-layer | 01/01/70 00:00 | |
| Seperate Analog and Digital Supply | 01/01/70 00:00 | |
| RLC-filters don't help? | 01/01/70 00:00 | |
| separate | 01/01/70 00:00 | |
| And... | 01/01/70 00:00 | |
| Seperate Supplies | 01/01/70 00:00 | |
| Evaluation board | 01/01/70 00:00 | |
| No not that one... | 01/01/70 00:00 | |
| Curly | 01/01/70 00:00 | |
| OT: Hair | 01/01/70 00:00 | |
| lets have pictures | 01/01/70 00:00 | |
| No no... | 01/01/70 00:00 | |
| http://www.k6ese.com/Mar_Vista_Shack.htm | 01/01/70 00:00 | |
| but it is relevant | 01/01/70 00:00 | |
| No wonder... | 01/01/70 00:00 | |
| waveforms | 01/01/70 00:00 | |
| Link to DDS page | 01/01/70 00:00 | |
| Add the second stage... | 01/01/70 00:00 | |
| Filter components | 01/01/70 00:00 | |
| clock | 01/01/70 00:00 | |
| What clock | 01/01/70 00:00 | |
| ? | 01/01/70 00:00 | |
| Sampling theorem | 01/01/70 00:00 | |
| But... | 01/01/70 00:00 | |
| Can't find it now | 01/01/70 00:00 | |
| Ground of probe? | 01/01/70 00:00 | |
| even better | 01/01/70 00:00 | |
| Good trick! | 01/01/70 00:00 | |
| same point | 01/01/70 00:00 | |
| Both | 01/01/70 00:00 | |
| I feel strong now! | 01/01/70 00:00 | |
| Funny! | 01/01/70 00:00 | |
| Depends... | 01/01/70 00:00 | |
| Oh Dear | 01/01/70 00:00 | |
| NO analog outputs | 01/01/70 00:00 | |
| since there is no schematics presented | 01/01/70 00:00 | |
| Publishing failed | 01/01/70 00:00 | |
| then give a link to an ORCAD drawing | 01/01/70 00:00 | |
| Schematics | 01/01/70 00:00 | |
| corrections | 01/01/70 00:00 | |
| Big test | 01/01/70 00:00 | |
| Try This Schematic... | 01/01/70 00:00 | |
| one more problem | 01/01/70 00:00 | |
| I miss the second 300R resistor | 01/01/70 00:00 | |
| Terminating Resistor. | 01/01/70 00:00 | |
| Answers | 01/01/70 00:00 | |
| University of Kai | 01/01/70 00:00 | |
| Keep us informed... | 01/01/70 00:00 | |
| of course | 01/01/70 00:00 | |
| Update | 01/01/70 00:00 | |
| Eager to know | 01/01/70 00:00 | |
| Me too | 01/01/70 00:00 | |
| Autoreload | 01/01/70 00:00 | |
| Current source output | 01/01/70 00:00 | |
| Electronic World Feb 06 | 01/01/70 00:00 | |
| Superb! | 01/01/70 00:00 | |
| No surprise | 01/01/70 00:00 | |
| AD9835 Cycles | 01/01/70 00:00 | |
| Solid Ground Plane Update!!!!! | 01/01/70 00:00 | |
| Very nice results! | 01/01/70 00:00 | |
| Low frequency | 01/01/70 00:00 | |
| SLIGHt correction | 01/01/70 00:00 | |
| OPA27? | 01/01/70 00:00 | |
| Cross talk on PCB tracks | 01/01/70 00:00 | |
| Have a look at this plot | 01/01/70 00:00 | |
| Interesting question | 01/01/70 00:00 | |
further interesting question | 01/01/70 00:00 |