Pelican System Selector Pro Review

 

The System Selector Pro switch by Pelican is shown above (case removed).  Inside, the product has three boards.  The front panel board is hidden in the image.  It contains the toggle buttons and the front AV connectors (identical to rear connectors).  The back panel board has the rear AV, optical, & Ethernet connectors.  The main board contains all the active circuitry.

Construction:

    The first thing we noticed was that all the boards are single layer and all components are through hole.  While this makes the boards very cheap to manufacture, they are not well suited to high frequency signals.  The intrinsic impedance of the signal traces is totally uncontrolled, generally >100K.  The lack of suitable ground routing also makes the design very susceptible to EM interference.

    The AV connectors are of average quality.  RCA connectors are tin coated and the optical inputs do not have shutters.

Design:

    The active circuitry inside the product is very simple.  This device passively switches all inputs except optical through 9 sets of relays.  These relays are all chained together to the output.  All signals are identical circuits, so evaluating one channel is sufficient.

    The Ethernet inputs are also switched just like the other signals.  This device cannot be used as a router, only one input is selected at a time.  Switching 10Mbps should be ok, but we did not test.

    The TOSLINK audio is switched through a TI SN74HC4066 silicon switch.  This should be fine for the digital audio signal.  We did not test this.

    The relays used for all the video channels and the Ethernet channels is the Songle SRC relay.  It is not rated for bandwidth by the manufacturer.  The contact resistance is specified as 100 mOhm.

Testing:

    The first thing we tested was the input to output real resistance of the switch.
        Ron : 2.2 Ohms        (A little high, but not terrible)

    Next we tested the bandwidth of the switch.  We used a 720p HDTV multiburst pattern so the test is a perfect representation of actual use.

Top signal is source, bottom is Pelican output.  2 ft cables are used on both input and output.

The signal is 6.2 dB down at 30 Mhz.  At 15 Mhz, it is <1 dB.  Between 15 and 30 Mhz the ringing (reflections) distort the waveform significantly. 

Bandwidth is sufficient for 480p, but not for HDTV.

Next, we investigated the effects of the poor impedance matching inside the box.

Top signal is input to Pelican (source), bottom signal is output of Pelican.   2 ft cables are used on both input and output.

You can see that even with short cables, there is significant ringing at edges of signals.  Since this switch is passive, there is no termination inside the box.  The ringing from the output will travel back through the box to the input.  That is why it is also visible at the input to the switch.

This level of ringing will be visible in the video image, but only at large signal transitions.

Top signal is input to Pelican (source), bottom signal is output of Pelican.   2 ft cables are used on input and 50 ft cables on the output.  Second image is a close up of the edges.

You can see that with a long cable, the time constant of the ringing is much longer.  This can produce a lot of image distortion.

Compare this with the output of the CSW02 active switch using 100 ft cables (below).  The edges are clean with zero ringing.