| ??? 10/26/04 02:07 Read: times |
#79813 - RE: RF interference of 11.0592MHz Responding to: ???'s previous message |
Kaiwalya said:
If crystal of frequency 27.175MHz is placed in receiver & if any unshielded probe or wire is touched to the ground pin of controller running at 11.0592 MHz (present on same board), receiver acts like actual signals from corresponding transmitter are received & triggers the event of kart passing the finish line. Hallo Kaiwalya, as Steve and Erik already stated, you have a relevant grounding problem! Only the use of PCB containing a solid ground plane, or better a 4-layer board containing this, and, if this is not enough, the use of a Faraday cage will help. Think of the Faraday enclosures that are used with TV tuners, then you will see what I mean: 27MHz reciever and microcontroller circuit are sitting in separate sections (Faraday enclosures), with their solid ground planes soldered directly to Faraday enclosure, all over its length! What you observe is a susceptibility against ground noise, which results from the lack of proper shielding and grounding. Means: 'Ground A', observed at microcontrollers ground pin and 'ground B', signal ground of 27MHz reciever DIFFER, i.e. ground noise is superimposed between point 'A' and point 'B'. The wire you connect to 'ground A' acts like an antenna, 'transmitting' the ground noise directly to antenna input of 27MHz reciever. And because sensitivity of a normal reciever is extremely high, this injected noise appears as big noise signal and an unexpected performance occurs. Very probably some intermodulation takes place, which makes the ground noise at point 'A', containing oscillator frequency and its harmonics, interfere with some mixer frequency of superhet reciever. If the reciever input is relevantly overdriven, which is very probably the case in your application, then the very unlinear characteristic of every involved reciever stage can cause a total unpredictable behaviour... The only remedy is to use proper grounding and shielding technique as stated above and by this forcing all ground potentials of your board to be nearly identical. I guess, that you will not be successful, unless you use this Faraday cage concept. A further way to reduce this common mode noise is to introduce common mode filters where ever you cross the separated sections. By the way, do you use any coding of signal? You could make the reciever only to detect a race car, when a certain code is transmitted, instead of simply detecting the presence of carrier. Kai |
