add docs on filter design

docs/src/lib/constructors.md

@@ -20,4 +20,5 @@ zpk
 delay
 pade
 ssdata
+ControlSystemsBase.seriesform
 ```

docs/src/man/creating_systems.md

@@ -433,4 +433,19 @@ uF  │       │ │  │       ├──────►         │ yC │  uP
                              └───────────────────────────────────┘
 ```
 
-See code example [complicated_feedback.jl](https://github.com/JuliaControl/RobustAndOptimalControl.jl/blob/master/examples/complicated_feedback.jl).
+See code example [complicated_feedback.jl](https://github.com/JuliaControl/RobustAndOptimalControl.jl/blob/master/examples/complicated_feedback.jl).
+
+## Filter design
+Filters can be designed using [DSP.jl](https://docs.juliadsp.org/stable/filters/). This results in filter objects with types from the DSP package, which can be converted to transfer functions using [`tf`](@ref) from ControlSystemsBase.
+
+```@example FilterDesign
+using DSP, ControlSystemsBase, Plots
+
+fs = 100
+df = digitalfilter(Bandpass(5, 10; fs), Butterworth(2))
+G = tf(df, 1/fs) # Sample time must be provided in the conversion to get the correct frequency scale in the Bode plot
+bodeplot(G, xscale=:identity, yscale=:identity, hz=true)
+```
+
+See also
+- [`ControlSystemsBase.seriesform`](@ref)

lib/ControlSystemsBase/src/dsp.jl

@@ -18,7 +18,7 @@ end
 
 
 """
-    Gs, k = seriesform(G::TransferFunction)
+    Gs, k = seriesform(G::TransferFunction{Discrete})
 
 Convert a transfer function `G` to a vector of second-order transfer functions and a scalar gain `k`, the product of which equals `G`.
 """