|
本帖最后由 超級狗 于 2015-9-25 13:35 编辑
* M, @7 S. _: b4 j1 L) z
3 s# i4 D0 u! |) _6 y5 ^( P7 X晶振電路緩起振(Long Start-Up Time)問題,樓上兩位的意見都不錯。這篇肥死卡好(Freescale)的技術文檔中,有較詳細的文字說明,之前貼過一次。- D @( M4 v ]* c7 ^% [6 ^
5 h4 F$ T/ Y; y8 z
7 Insufficient Loop Gain! z# c R; Z/ k/ p9 R
Just as overdriving the crystal can be a problem, insufficient drive can also be the source of many failures. For adequate loop gain, load capacitors must be sized correctly, and RB and RS values must be appropriate in the Pierce configuration. Load capacitance affects loop gain since the feedback voltage is obtained in both configurations from the voltage divider formed by C1 and the crystal, so it is very important to account for stray capacitance when
( Z! l- X- } J' H% w' i! Ycalculating the value of C1 and C2. In the Pierce configuration, adding load capacitance will reduce loop gain in some cases. Resistors RB and RS also have an effect on loop gain in the Pierce configuration. Making RB bigger increases the loop gain, while reducing RB decreases it. In general, lower-frequency crystals require higher values for RB because their impedance is normally higher than that of high-frequency crystals. A typical value for RB is usually given in the microcontrollers' datasheet. Resistor RS has the opposite effect on loop gain, since loop gain is reduced when the resistor is increased and is increased when the resistor is reduced. To make sure that loop gain is sufficient, a potentiometer can be placed in series with the crystal. If loop gain is barely sufficient, increasing the resistance will soon prevent the oscillator from starting. The recommended procedure is to start with the potentiometer set at 0 Ω, then slowly increase the resistance. After each increment, power to the board should be removed and then restored. Eventually, the circuit will fail to start. When this happens, the total resistance (crystal + potentiometer) should be substantially larger than the worst-case resistance specified for the crystal. For example, for a good margin of safety, it would be nice to have the circuit oscillate with at least twice the maximum specified crystal resistance. The size of the potentiometer required for the test will increase as the crystal requency decreases (i.e., 1 kΩ for a 10 MHz crystal and 10 kΩ for a 32 kHz crystal). It is recommended that the previous test also be carried out at the highest temperature and lowest VDD at which the circuit is expected to operate, since loop gain in many oscillators tends to decrease as temperature increases and as VDD decreases.
' I \0 S% G* E* _2 ] e
* T- N" r& m0 R: x* E6 `6 r- |8 J8 J" E% w7 E7 q( h$ D1 ]
8 Long Start-Up Time# ?' j5 h r0 N5 J% T5 {
A long start-up time is usually a more common problem with low-frequency crystals, since they tend to start much more slowly than high-frequency crystals. One of the causes of a long start-up time is weak loop gain, which was addressed in Section 7, “Insufficient Loop Gain”. Oscillator start-up times will also be affected by the rise time of the power supply. When the power supply has a sharp rise time, the crystal will experience an energy impulse that will usually make it start faster than it would using a power supply with a slow rise time.
* T0 y* T' N$ Y9 W# R' l; b' V1 ? |
|