Relays as input switch?

How you wire the relays will be determied by how you intend to tell them to switch:
--Using a front panel switch to route DC to the coils--piece of cake. (Yes it's nice to use a front panel switch <i>and</i> relays, because you can use small relays with nice, gold-plated contacts and a cheap switch. The DC doesn't effect the signal.)
--Using a logic chip to take charge of the process--this will depend on the chip you choose.

DC doesn't effect the signal

That is true. However, magnetic field from the coil might degrade the sound. Many preamp constructions pass audio input and output signals through an activated relay and the magnetic field of its coils. The better way for input switching is to do it in the opposite manner by conducting the signal through an inactive relay while all non-conductive relays are activated. Thus, the magnetic field cannot adversely affect the sound quality. It is also a good idea to switch not only the input signals but also corresponding grounds.
As long as you're using DC on the coil, I'm not sure I see how the coil field will effect the signal. If you get into a steady state magnetic field causing audible differences, then we need to start accounting for the Earth's magnetic field, the magnets in phono cartridges, magnets in CD drive motors, etc. etc. etc.
AC, of course, is a different matter.
I'd leave all grounds hooked up at all times. All you need is one chassis at a different ground potential (not at all uncommon) and you're likely to get a substantial <i>pop!</i> as you switch inputs. Not to mention the unfortunate results should you get a bit of dust or corrosion on the ground contact so that it doesn't "make," leaving the hot the only thing that connects...

In my experience relays DO do damage, or at least alter the audio output.
To prove this, add a wire (wires) to short across the relay signal contacts, and then alternately connect or isolate the relay coil connection.
IME you get two (slightly) different sounds.
Also try this with amplifier output protection relay.

Conclusion - the coil current induced magnetic field has an audible effect.

BTW - Earth's magnetic field is much lower level than that in a relay.

Regards, Eric.
Affect of DC on signal paths nearby.

Although purists may object, I think there is merit in using the sonic change (should there be one) resulting from DC on the relay coil to help tailor the sound of your amplifier.
I remember owning a Sony CD player where the wires leading to the analog output jacks were bundled with the power supply wires for the analog ouput stage. On first seeing this arrangement I thought this was just a way of making the internal wiring look cleaner. After re-routing the wiring; keeping the output and power lines as far apart as possible I noticed that the sound became fuzzy and overly warm. Returning the wiring restored the original sound. It seems Sony knew what they were doing and were tweaking the sound of this CDP using the nearby supply lines.
It is true that this method (if it's true and I'm not just imagining things) is only distorting an already distorted signal, and the best thing would be to clean the original signal. But as long as it sounds better I have nothing against it.
Trying both active and inert relays for the selected source to see which one you like sonics-wise may be a good idea.
My biggest (though purely theoretical) objection to a relay was the physical path. Depending on the relay design, you are going through several unkown materials. The relay contacts, the wires from the relay contacts to the relay pins. And the pins themselves.

Maybe some one out there has "auditioned" different make and models of relays to see which ones sound the best......

The only reason I can think of using a relay is to integrate with some sort of TTL type logic and/or remote control application.

If your design still requires that you walk up to your pre-amp and throw some switch, then my advice is to just pass your signal through your preferred brand of audio switch. 2 position DPDT switches work nicely, you can switch both the ground and the hot at the same time. Do not underestimate the value of this.

If you still wish to use relays, there is an arcane type of logic called "relay logic." There are many tricks to it. One is that you use extra contacts of the relay to energise itself. (The relay's coil is in series with one of it's own contacts and the relay coil power supply) This priniciple can be used to "lock out" the other relays.

An added trick is it add capacitors to the coil circuit so that when you power down the 1st coil, the 1st coil stays enregised until the 2nd coil discharges the capacitor. If you do it right, the 2nd source's relay will be energised before the 1st relay's coil discharges. This gives you a "make before break" switch. Neat.

Old electronics books from the library an intrnet search would be useful.

Back to the scheduled programme....

Thanks for all your replies people. What I´m actually looking for is some way of switching the relays via separate push-buttons for each input. What I need to know is how to make a circuit that will switch on just one input at a time and then keep this relay on until I press another input select button (or turn the thing off, obviously :) I found a project over at headwize that pretty much describes how to do the rest of the circuit, but it uses a rotary switch, which I, for various reasons, would like to avoid.

Can anyone help with this??
different ground potential

I'd leave all grounds hooked up at all times. All you need is one chassis at a different ground potential (not at all uncommon) and you're likely to get a substantial pop! as you switch inputs. Not to mention the unfortunate results should you get a bit of dust or corrosion on the ground contact so that it doesn't "make," leaving the hot the only thing that connects...


Recently I bought an upsampler from Monarchy and added to the chain between DVD player and MSB Link Dac. Since it is hooked up with the rest of the system I get a very loud pop and click from my home brew switch box (which is a double pole DACT 5 IN 1 OUT selection switch) when switch into and out of the upsampler's path. After reading your previous post here I reckon I must have this upsampler at a different ground potential. Is there a way to saturate the different grounds?

Please help.

You must be new here :)

Gain devices add a considerably higher degree of coloration and/or distortion than any relay, which is essentially a straight wire. The concern over magnetic influence is splitting hairs in comparison. Naturally, this is what makes people audiophiles -- they care about these details.

Ben Duncan used these a few years ago in one of his preamp articles in Hi-Fi news, might be worth considering.

<a href="">Pickering Series 103 pdf</a>

<font size = "+0" color="#400000" face="times">A version with an internal coaxial electrostatic screen is available which is ideal for applications where capacitively coupled noise from switch to coil connections is undesirable. The coaxial device has a characteristic impedance of 50 ohms and is also suitable for RF applications, HF performance is similar to the Series 102M.

The range consists of two basic types, the first achieves ultra low capacitance levels of typically 0.1pf from each switch connection to the coil and typically 0.08pf across the open switch contacts, by virtue of an internal coaxial electrostatic screen or guard connection.

Where it is not possible to drive a guard, the second type has inherently low capacitance figures of typically 0.4pf from each switch connection to the coil and typically 0.1pf across the open switch contacts.</font>



2002-01-25 9:28 pm
Relay Input Selection Circuit

I designed a preamp input selector circuit, where a single pushbutton switch is used to toggle between 6 inputs. I also included a switched tape output, and a mute feature. Shielded Aromat gold-contact relays are used.

So far, the circuit works well breadboarded - boards are next. This will be the front-end of a Joe Curcio Daniel II preamp/xover.

Once I figure out how to post a .pst file, I will be happy to share it.