quaternion
Joined: 12 Nov 2006
Posts: 209
Helped: 11
Location: Cairo , Egypt
|
 16 May 2007 13:16 colpitts crystal oscillator |
|
|
|
|
I am working at discrete level.
I have tried many schematics either in simulation or in practical to build a Colpitts crystal oscillator at 10MHz but all fail. I used ORCAD for simulation but never get oscillation ,when I replace the crystal by a coil I get oscillation but I wants to work with crystal to assure frequency stability.
What is the best configuration (common collector,common base or common emitter) at this frequency(is it common base) also taking into consideration that oscillator coupling could be by a reactive element (say a capacitive coupling to an amplifier).
Also I have 9V bias and I try to work with 2N3904 transistor.
Please help.
|
|
xxargs
Joined: 28 Jul 2002
Posts: 231
Helped: 44
|
16 May 2007 14:52 crystal colpitts configuration |
|
|
|
|
| quaternion wrote: |
I am working at discrete level.
I have tried many schematics either in simulation or in practical to build a Colpitts crystal oscillator at 10MHz but all fail. I used ORCAD for simulation but never get oscillation ,when I replace the crystal by a coil I get oscillation but I wants to work with crystal to assure frequency stability.
What is the best configuration (common collector,common base or common emitter) at this frequency(is it common base) also taking into consideration that oscillator coupling could be by a reactive element (say a capacitive coupling to an amplifier).
Also I have 9V bias and I try to work with 2N3904 transistor.
Please help. |
You have wait long time enough in your simulation???
oscillation with crystal take looooong time to starting and recive full amplitude depend of high Q-value inside crystal.
If crystals Q is around 100000 (usual value), is same as free running oscillating (shorted) crystal need 100000 periods to damping 27.29 dB (1/e^2PI) - simular situation for start/rice amplitude in crystal in oscillator coupling.
You must simulate with high resolution (> 20 - 100 sampel per period) and millions of periodes before oscillator make steady state level of output, ie. 0.5 - 1 second timeline in simulations before oscillator recive full amplitude from time starting very weak but steady initial oscillation.
You can force this if starting process is not important with charging crystals serial capasitance to around 100000 Volt and after in simulation see if amplitude still rice or going down and adjusting cap voltage again to you make steady amplitude level
Simulator have not normaly thermal noise and in starting process can weak oscillating jumping around simular state machine (depend of simulators numeric precision and time step etc.) and never find way to fire up to big oscillating. This is never happen in real world circurits with thermal noise and inteference from other equipment. So, try different intial starting charging value inside crystals L and serial C if oscillator never starting to big oscillating in simulation.
|
|
