??? 04/01/05 12:50 Read: times |
#90799 - 1 Hz Resolution Responding to: ???'s previous message |
This will be a difficult task to achieve to the exact 1Hz resolution. Synthesis of frequencies to exact steps such as you have requested help with often requires the use of a number of carefully selected crystal frequencies and the use of an multiplying / divider PLL circuit.
To show the difficulty let us assume that you have a 8051 style microcontroller of a modern design that has a 16-bit timer like a TIMER2 that can operate in auto-reload mode using the mode that can drive an output pin with a square wave frequency. Such mode is often called the "High Speed Output" mode. In this mode the timer reload register is set to a value that determines the half cycle time of the output waveform. At each timer overflow event the output pin state is toggled. I have made a spreadsheet that permits analysis of the timer frequency generation capabilities. If the 8051 derivative allows the timer to operate at the Fosc rate (i.e. timer clock rate = SYSCLK) then the spreadsheet can show divisors needed to produce certain frequencies. You may access and download the spreadsheet here: Timer Divider Spreadsheet The following chart shows the first 40 or so steps of the spreadsheet with the desired frequency starting at 15.000 KHz and stepping up in 1 Hz steps. As you can see the integer divisions that are acheivable using typical microcontroller timers are unable to achieve the result that you desire. ![]() Michael Karas |
Topic | Author | Date |
frequency resolution | 01/01/70 00:00 | |
1 Hz Resolution | 01/01/70 00:00 | |
AD 9850 ? | 01/01/70 00:00 | |
Yes AD9850. | 01/01/70 00:00 | |
All digital | 01/01/70 00:00 | |
Thanks Steve. | 01/01/70 00:00 | |
Formatting | 01/01/70 00:00 | |
Formatting | 01/01/70 00:00 | |
derivative | 01/01/70 00:00 | |
Try SiLabs![]() | 01/01/70 00:00 |