On Mon, Jul 23, 2012 at 2:37 AM, Andrey Sviyazov andrey.sviyazov@fairwaves.ru wrote:
Thomas, thanks for file. But, may be you know, how to convert it to *.cfile for tx_samples_from_file.exe? Here attached zip with example of *.cfile which Alexander made for me in somewhere like MathLab, I think so. Rename zi_ to zip first.
Alexander, may be you know and then you can convert this file or part of it?
You do not need to convert, AFAIK.
Overall measurements are close (either above or below) to 1.5/5 for the 900 band in general. 945 MHz is currently the best case, but that is the frequency I spent the most time on with calibration. Do we need to be consistently below 1.5/5 or just close?
Thanks for the measurements. At the end we should be below that at all GSM bands. For now it's fine to be close.
I also noticed that your frequency is far off - could you calibrate it and make pictures of the best and worst case in the 900 band?
Perhaps this is a silly question, but how do I calibrate the clock? or use an external reference signal?
Actually you can tune LMS frequensy with 1Hz step to compensate known error.
This won't produce correct symbol speed.