Using Electronic switching to route signals
As a follow up to our article on multiplexing (and demultiplexing) or MUX/DEMUX and running my own experimentation with a CD4051 it comes as no surprise that this chip works well but there are limitations.
As I am no expert in this field I must confess that the biggest obstacle IMO when switching or routing signals is not the degrading of signal quality necessarily but the dangers of phantom power. As most microphone preamplifiers use an XLR plug or TRS combo, the TRS does not pass the +48V but the XLR does. Now if someone unwittingly connected his expensive tape deck to XLR outputs and then into an XLR mic preamp there is a chance that they would end up with a toasted front end to the deck. Yes, we all like XLR plugs but this can be a problem. How to eliminate this problem?
Phantom power – plugging into the unknown
Phantom power is usually used to power condenser microphones through a balanced signal input. Pins 2 and 3 of the XLR plug both receive +48V which causes 0V potential difference across the 2 and 3 connectors.
If one needs to take sockets to the front of a console and these are connected to a microphone preamplifier with phantom power there is a very strong likelihood that the user at some stage is going to connect a tape deck or preamp output to this socket. The output may very well be unbalanced which will cause +48V at very low current to be imposed across pins 1 and 3. This can damage the output circuitry of the source. At the very least this may cause a very loud pop through the speakers.
Blowing a MUX chip
If the signals are routed via a CMOS 4067 or 4051 chip the likelihood of causing damage is very real. In this case it becomes very apparent that manufacturers of patchbays use only TRS plugs and sockets, with no phantom power. But mistakes happen and us DIYers will use a combination of XLR and TRS to achieve a desired outcome. In most cases it is advisable to use combo connectors or separate phantom and non-phantom powers connectors. These must be visibly marked.
In the schematic above two reed relays can be used to bypass the coupling capacitors. This circuit is not advised however, separate your powered and non powered jacks.
As we have seen, mixing and matching XLR and TRS connectors can be a troublesome approach if using MUX/DEMUX chips. What is the recommendation?
In almost every case complaints arise out of either the Ron resistance creating distortion (my tests prove very little) even with the cheaper TI chips. Enthusiasts recommend reed relays. Shop around, they can be obtained very cheaply – Mantech sells them for about R15.00 each. Check under code 35M1042, 12V at 500 Ohm coil resistance.
Power relays are not recommended. The switch contacts are designed for a minimum current flow and audio signals are often in the microAmperes.
Wafer switches are tried and trusted. The pitfall is often the limitation on poles and amount of contacts.
In the circuitry covered up to now we are using a switch in series with the signal path but what happens if we apply a muting configuration by grounding the signal? Most of the muting circuits I have come across pull down the audio signal to near ground and open circuit to next stage input – to combat noise this is by far a better approach.
Australian Rod Elliott as always comes up with a solution to most of our audio woes on his pages and I was really intrigued by his BJT circuit. Of course all of us have read that muting a signal with BJTs is a big nada but here is a very good example of how to get an Ron resistance of near 3 Ohms with a BC549. And 0.02% distortion.
Muting Transistors – Muting Transistor Attenuator Circuits and the 2SC2878 – Jim Keith (thought provoking past, a must read)