[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Discuss-gnuradio] interpolation and output level
From: |
Matt Ettus |
Subject: |
Re: [Discuss-gnuradio] interpolation and output level |
Date: |
Tue, 8 Mar 2005 11:01:08 -0800 |
User-agent: |
Internet Messaging Program (IMP) 4.0-cvs |
Quoting Chuck Swiger <address@hidden>:
> Hi - Just curious whats the mechanism behind this: using a usrp xmit
> interpolation of 64 we get max signal level, and it drops off a lot if
> changed, say to
> 40 or 80. With interp=64, sample rate of 2000000, sig ampl of 6000, center
> at 14000e3 and freq of 72e3 my output is 3v peak (external amps).
> Change it to interp=80, sample rate of 1600000 it drops way down to .6v peak
> (-14db). Try 50, 3200000 and get .5v peak out. Even stranger using interp=60
> gets 2.7v peak but 68 it's down to .6v.
This happens both with the interpolator and decimator. It is due to a
simplification I made to save space in the FPGA. I will be fixing it in the
near future.
Basically, if you interp/decimate by a power of 2 you get full amplitude
signals. If you use a non-power of 2, you will get lower amplitudes. Using a
value just greater than a power of 2 will have the worst effect. A value just
less than a power of 2 should be less noticeable. This is why 68 is so much
lower than 60 in your example above.
Matt