Do we really need relays for phantom/pad/polarity switching?

[Svart]

I remember reading a thread somewhere that discussed the legitimacy of using relays for switching of phantom/pad/polarity/etc.  Of course I did a search but nothing of any value showed up, but I don't even remember if it was here that I read that stuff, so lets discuss this.

Do we *really* need to do this?  I don't understand why products tend to switch these things through relays and not through switches.  Is it simply that the relay can be remotely controlled and you don't need to run the audio everywhere to get to the switches?

Any opinions?

And of course this is almost a rhetorical question. 

We don't *really* need to do this fancy stuff, but why are some people convinced that we do?

 

[JohnRoberts]

The only possible justification for relays is that they can be located proximate to preamp circuitry and at some distance from the panel switches. OTOH you now have relays in the middle of microVolt level noise floors, so pick your poison...

JR

 

[audiox]

A long time ago my colleague made a switch/relay comparison in line level switching. The result was that a signal relay (some standard Alcatel low level type) produced approximately 10 dB less switching noise than a switch (C&K toggle with gold plating). I guess that is because the relay switching speed is faster (contact bounce duration is shorter).

Relays have also longer lifetime than switches.

 

[shabtek]

I would guess a quality relay would cost less than a comparable mechanical switch (if available)--relay data sheets spec better low level switching ohms, # cycles etc. than mechanical switches (usually spec'd in amps)... that, and proximity/less wires as mentioned-- for a mic pad relay is good. phanthom not neccesary and polarity do it on the output @ line level either way.
not sure switching noise is the most valid criteria. twill always be greater if done at input of a high gain stage (phase reverse @ input of mic pre)

 

[emrr]

Longer relay life versus ability to source replacement many years down the road.  I'll vote for sourcing a switch over a relay.   The only argument I see in favor of relays, in this context, relates to advantageous physical placement.  

People like to complicate things.  #1 item around here is regulated DC tube filaments.   It's almost as bad as $300 power cables and great sounding wooden knobs.   It's not so much that people have heard a problem they wish to correct, they simply believe it will be better.  Never mind the service problems that come with a more complex system.  

 

[Svart]

good discussion.  personally I don't see a technical need for relays, it seems more psychological.

So, would you pay more for a device with a relay or would you rather save the money?

 

[tv]

It's less "wire" in the signal path (low-level signal wire dosn't need to run to the frontpanel - or traces if integrated on board). Plus, it looks good in brochure.

 

[Svart]

So if it says "relay" it's OK?

What about "sealed relay"? "Automotive"? "Bifurcated contacts"? "low signal"? "Shielded"?

Do people care about these attributes that I would consider more important that simply being a relay?

 

[bruno2000]

I using socketed, nitrogen filled relays for audio switching in all of my new builds.  The switches that control the relays can be of the less expensive variety, and I don’t have to worry about the contacts getting dirty since they are switching 12 VDC.  If the relay goes bad, I just remove it from the socket, and plug in a new one.  The cost of a decent relay plus a modest switch is about the same as a really good switch with gold contacts, and as has been said, it enables much better board layout.
Just my $0.02
Bruno2000

 

[Svart]

hmm I can't tell if you are joking about the relay stuff, but I understand the principles around using them, I want more of the psychological aspect of why people choose to use them over something like a normal switch.  

Do fancy designs really make a better product or do we simply think it does?

In going along with the "nitrogen filled" relay thought, what exactly does that buy you?  You aren't switching power so there is no need to extinguish arcing or flashover.  Doesn't metal on metal mean contact mean the same thing if it's in a switch or in a relay?  Relay contacts aren't exactly the same as a good switch most of the time.  Does a "good" switch really make that much difference besides wearing out less over time?

What about the socket to pin contact resistance?  Wouldn't the small amount of resistance equal that of using a relay with cheap contact materials(although even the cheapest low signal relays use gold..)?

I suppose it's just as easy to step into audiophool territory with this stuff as it is with anything else.

 

[emrr]

The theoreticals around how much time one might buy with 'better' mechanical parts are really hard to gauge.  I have plenty of things I built 10 years ago, using $2 Mountain switches, that work just fine.  Not to curse myself, but I've not had to replace a switch yet.  When will I?  What's an acceptable life span?  I have plenty of 60 year old gear with signal switches that work just fine.  Switches of that age are easy to replace even today.  A dead relay in an early 80's piece might fill me with dread. 

I do approve of small signal relay use when it obviously solves a layout problem.  But many times a layout can be done to account for switch placement on a panel.  So, which is it? 

 

[bruno2000]

Nitrogen filled as opposed to “open air” relays (or switches) are much less susceptible to forming semi-conducting films on the contacts of the relay (or switch).  Basic chemistry.
Bruno2000

 

[JohnRoberts]

A long time ago my colleague made a switch/relay comparison in line level switching. The result was that a signal relay (some standard Alcatel low level type) produced approximately 10 dB less switching noise than a switch (C&K toggle with gold plating). I guess that is because the relay switching speed is faster (contact bounce duration is shorter).

Relays have also longer lifetime than switches.

10dB less switching noise?? ?  Huh.. If any switch contact is switched between two different potentials there will be a step response (click, thump, etc). If there is no potential difference, what does the switch bounce sound like? If anything relays are more likely to introduce noise into sensitive paths than mechanical switches.

There is similar merit to putting long shaft extenders on push-push switches so they can be located closer to where they need to be...

In the grand scheme of audio performance this is counting angels on the pin head. Not as significant as many other things IMO.

JR

 

[owel]

I've used plain DPDT switches before but nowadays I use relays.

I use relays so I don't have to route signal wires all the way to the front panel, and then back again. I can do the switching (via relays) where I need it to be. Only the "control lines" need to be wired to the front panel. Less noise/interference susceptibility.

2nd, I can use a simple SPDT switch. 

3rd, or use a DPDT switch, and use the other pole to turn on/off an indicator lamp... to signify if pad/polarity reverse is engaged or not.

4th, I can control the relays via a microprocessor... something you can't do with mechanical switches.

2 relays only cost an additional $5 total - and these are gold overlay, bifurcated, small signal. 1x10^8 lifetime operations relay. 

In the end, it's just a matter of personal preference.

 

[audiox]

10dB less switching noise?? ?   Huh..

I am not saying that it had any practical meaning. It was about meeting some German broadcast standards. Mechanical switch didn't meet the specifications but relay did (and we were happy). The noise level was measured using a psophometer.

 

[Svart]

Nitrogen filled as opposed to “open air” relays (or switches) are much less susceptible to forming semi-conducting films on the contacts of the relay (or switch).  Basic chemistry.

Sure I understand, I did pretty good in chemistry in college even though I rarely use it these days.. 

But I do know relays since I dabbled in designing power switching using relays and contactors after school.  The best contacts are not pure gold, but an alloy mix of gold and other platinum group metals regardless of the atmosphere that they exist in.  Most common relays are designed to never tarnish under rated load conditions and they have specifications that rate their lifetimes over contact actuations.  Generally these are rated for uOhms to mOhms for initial tests and ALSO for 200k actuations.  I've seen general use(sealed but non inert filled) relays be 100mOhms out of the box and then 200mOhms after half a million actuations. 

Does that really count as a relay wearing out?

 

[emrr]

John brings up a funny related tangent.  I do wish there was still market position for shaft extenders and flexible shafts.  I could use a lot of them for keeping pots closer to desired signal location, but there isn't much on the market at all.  And it's not as cheap as it should be.  I'm unaware of any available flexible shaft couplings. 

 

[clintrubber]

I suppose it's just as easy to step into audiophool territory with this stuff as it is with anything else.

Most likely indeed... what's your application ?

Some units for yourself for which you want to make the best choice ?

Or some units to be built for others ?

If it's the latter, there'll sure be the pretty customer-dependant issue of 'whoah, there's clicking inside, serious piece of kit' vs
'hmm, stupid unit, couldn't they make it a bit cheaper and skipping those fancy relays' ?

Bye,

 Peter

 

[clintrubber]

John brings up a funny related tangent.  I do wish there was still market position for shaft extenders and flexible shafts.  I could use a lot of them for keeping pots closer to desired signal location, but there isn't much on the market at all.  And it's not as cheap as it should be.  I'm unaware of any available flexible shaft couplings. 

I recall those flexible bands in my old Marantz amplifier for the source- & record-select switches, nice stuff...
incidentally, the speaker-relays need to be replaced in that amp (OK, that's no small signal stuff),
but those on-PCB open-air source-selection switches are still doing fine...

 

[clintrubber]

Talking about switches, what are my chances to keep old switchbanks like this (attached)
decently switching line-level signals in the long run ? Don't want to have to spray them once a week.

I'm throwing a little something together using these, but the plan was to directly go for those switchbanks controlling relays.

Regards,

  Peter