This is the place for any magazine-related discussions that don't fit in any of the column discussion boards below.
4 posts • Page 1 of 1
I have an application where we want the user to select an oscillation frequency with jumpers or dip switches. I want 8 frequencies, spaced 300Khz apart starting at 7.3728Mhz and ending at 9.4728Mhz. I prototyped with the LTC1799 from Linear Technologies. It uses a single resistor to create the frequency. I used .1% resistors. We have 2 board that must talk to each other. My prototypes worked good so we went to a board mfg. The mfg of the boards say they get some that don't communicate. They think it's because the chip has a <1.5% error and this is enough to create an unmatched pair. This is not RS232 communication. Are there any other variable-frequency oscillators that have a better tolerance than 1.5%?
Looking at each end of your specs--on one hand you need better than 1.5% accurracy ( how much better?). On the other hand your start/stop frequencys are speced at 10 hz precision which for the fundamentals involved would equate to roughly 1PPM. Do you know what your allowable error is? If it is tight, you may have to go with a simple phase lock loop system that will step it in 300Khz increments. R-C oscillators probably wont give you much more than 0.5% stability.
Users browsing this forum: No registered users and 30 guests