The idea behind the switch matrix is to have one input that can be routed to one of three outputs which is selected with two digital control lines.

The circuit was at first designed to transform a SP3T-switch with a NC in the middle to a "regular" SP3T-switch. This might sound kind of useless, but this was cheaper for me then buying new switches since i allready had the parts for this. Also, this gives me the possibility to interface the switches digitaly.



The circuit is based around two CMOS 4053 analog multiplexers, which are setup so that there are three routers avaliable. Each router has one analog input, three analog outputs and two digital control inputs. One of the routers have a special alternative analog input which will be discussed later. Externally, each routers control lines are interpreted as the binary code generetable via a SP3T (NC):

• 00 - output A
• 01 - output B
• 10 - output C

Internally, hovewer, each router consists of 2 2-channel multiplexers and for the forthcoming logic to work, the binary input code has to be transformed into the following sequence.

• x0 - output A
• 01 - output B
• 11 - output C

The first multiplexer is directly connected to the signal source, control line 1 is used to route the signal either directly to output A (Control1=0), or to the input of the next multiplexer (Control1=1). The second multiplexer redirects its output either to output B (Control2=0) or output C (Control2=1). The diode connection in the schematic makes the control-line combination 10 impossible.



One of the routers has a special feature; a FET switch that makes sure there is allways a signal present at that output. The original intention of this was to have an ADSR that could be routed to three different functions, including a VCA. However when the ADSR was not routed to the VCA, it would be convenient to have a GATE signal routed to it instead, so that there was allways a signal controlling the VCA.



Since this circuit is essentially composed of FETs, it draws little current, a few µA. On the downside, it has limited voltage swing, the limit for the 4053 is 15v VDD-VEE. For switching CV signals i have yet to see any problems with it, though it's not suitable for high-end audio routing. This circuit is small, simple, and replaced my need for buying new switches so i like it!