SX48/52 datasheet question
bunni
Posts: 38
I was working on setting up IRC_CAL for sx48/42 and ran in to this, which the numbers are completely different from the sx18/20/28 datasheet, and frankly don't make much sense.
IRCTRIM2:
IRCTRIM0 Internal RC Oscillator Trim. This 3-bit field adjusts the operation of the internal RC oscillator to make it
operate within the target frequency range of typically 4.0 MHz. Parts are shipped from the factory
untrimmed. The device relies on the programming tool to provide trimming.
100b = maximum frequency
111b = typical
011b = minimum frequency
The sx18/20/28 datasheet says:
000b = minimum frequency
111b = maximum frequency
each step about 3%
Why is binary 4 the max, 7 the middle, and 3 the minimum? I would assume this is a datasheet error, if not, could someone kindly explain the steps for how trimming on the sx48/52 works.
-Kris
Post Edited (bunni) : 11/30/2007 2:50:15 AM GMT
IRCTRIM2:
IRCTRIM0 Internal RC Oscillator Trim. This 3-bit field adjusts the operation of the internal RC oscillator to make it
operate within the target frequency range of typically 4.0 MHz. Parts are shipped from the factory
untrimmed. The device relies on the programming tool to provide trimming.
100b = maximum frequency
111b = typical
011b = minimum frequency
The sx18/20/28 datasheet says:
000b = minimum frequency
111b = maximum frequency
each step about 3%
Why is binary 4 the max, 7 the middle, and 3 the minimum? I would assume this is a datasheet error, if not, could someone kindly explain the steps for how trimming on the sx48/52 works.
-Kris
Post Edited (bunni) : 11/30/2007 2:50:15 AM GMT
Comments
I don't know why you want to fiddle directly with the trim bits, but whatever. If you're confused as to what affect the trim bits have on the frequency (based on the differences in the datasheets), try writing a simple program that outputs a known frequency. Measure the frequency with a frequency counter or O-Scope, modify the trim bits, and then see what happens.
If you don't have the equipment to do that, then use two SX chips. On one, use an external crystal oscillator so you know it's running at the right frequency. On that chip, write a small program that counts incoming pulses during a fixed time interval. On the other chip, do what I wrote about in the previous paragraph. By counting the number of incoming pulses during a fixed time (make it long enough to be useful) you can determine if the change in the trim bits made the SX output more or less clocks. You can even figure out the specific time change for each of the trim bit choices.
Thanks,
PeterM
The reason I want to fiddle with them is because I want them working correctly in my assembler. The SXKey software uses these bits with the IRC_CAL setting, and I want it to do the same. I could scope out all of the variances, but I ask because I figured someone else might have the knowledge as to what the different trimming combinations do and what I am missing with the numbering.
-Kris