Karteikarten
Sigsys munta
Karten in diesem Stapel (15)
- 1Causal filterdepend on past and current but NOT future
- 2Decibel to attenuation- x dB => 10^-x/20
- 3Attenuation to DB20log(x)
- 4what has to be true for a filter to be causal?h(t) = 0 for t<0
- 5what happens if h[k] has infinite range with a FIR filter? (i.e non causal)we must have sharp transitions in h(w)
- 6periodic in one domain..discrete in the other
- 7symmetric in one domainreal valued in the other
- 8pros and cons with tapsA lot of taps -> more delay in time domain
but also better roll-of - 9rect vs triangular vs hammingrect: +fasest rolloff, -ripples
triang: +less ripples, -slower rolloff
hamming: +no ripples, -slower rolloff - 10what do we map between in bilinear transform?s = jw and e^jwTs (using real frequency)
- 11what is mapping between H(z) and H(w) when using real freq w?z^jwTs
- 12Transer function and frequency response of FIRtransfer: H(z) =SUM bn*z^-n
H(w)=SUM bn*e^-jwn (norm)
H(w)=SUM bn*e^-jwn*Ts (real ) - 13Important difference between FIR and IIRFIR does not depend on past values of y
- 14How do we derive the taps for FIR?a continous function can be expressed with fourier series, in the frequency domain we can approximate the coeffecients of the fourier-series to be the taps in the Z domain. The same taps we'll have in the time domain
- 15pole-zero mapping1. find poles of H(s)
2. use z=e^sTs (where s are the poles or zeros)
Informationen
- Beschreibung
- Sigsys Karteikarten am Sigsys munta, erstellt von Emma Tysk am 11/03/2018.
- Erstellt von
- Emma Tysk
- Aufrufe
- 3
- Bewertung