Too many relays?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

boji

Well-known member
Joined
Jan 6, 2010
Messages
2,375
Location
Maryland, USA
I kinda worked myself into a mechanical rats nest with relay logic.

What at first seemed like the cleanest way to route signal, has now become a question if passing balanced signal through 8 relays degrades signal.  What's your opinion on this?

Of course it increases the chance for mech. failure. But does the signal suffer after too many relays?
I'm using Panasonic TQ12's.
 
should be OK with decent quality relays, but seems like a lot...

Back in the early days of digital computers using relay logic, moths would get caught in the relays... so they would have to be "debugged".  8)

JR 
 
Should not be a problem. Probably no more relay switched signal  connections in the signal path than there would be with regular switched ones without the relays. At least the ones in the relays are in a dust free environment.

Cheers

Ian
 
boji said:
Thank you gentlemen. I'll try not to lose sleep over this one too.

In mixer builds, as often as not the things you worry about do not come to pass. But the unexpected ones do.

Cheers

Ian
 
Hope not Ian! But I hear you. 

So while I'm at it... what's the real world failure rate of these things? Has anyone around here had a Panasonic relay go bad? Started out making boards that could be quick replaced, but found I had width constraints as a result.
 
Those are a well regarded series of relays.  Don't have real world numbers,  but looking at the datasheet should give information on the failure rate.

I don't think signal integrity through the relays would be a big issue.  But do be careful to keep the switching noise out of the audio.

Keep the relays if necessary,  but imo it's always good to get rid of extra parts, meaning if you can do with less,  then do so.
 
it's always good to get rid of extra parts, meaning if you can do with less,  then do so.
I may take this advise to heart, and do away with the extra fancy group cards that double as input cards. It would vastly simplify things at the cost of what? Ok so a few more sometimes analog mix channels, but they won't have aux outs so their use is limited.

If I'm to be honest, this is me wanting to avoid another bucket of input cards. I should just suck it up and do the metal work.
 
+ some more :)

For OP's question: all the new-fangled "passive relay volume controllers" run signal through 6-8 relays all day. In my G22, I run signal through 2x9 relays when in MS and hard-linked. Never had a relay failing me (we use Takamigawa (sp)).

Jakob E.
 
Metal work is always the nemesis of electronics.

For my background, it's far less intimidating than interpreting / modifying circuits. It's just darn laborious. I am kinda blown that eyed-only cuts fell within tolerances. 

Not perfect, but I consider it a 'sans cad/cam' win. 

Xc8atqb.jpg

Ia4b2wB.jpg

 
For OP's question: all the new-fangled "passive relay volume controllers" run signal through 6-8 relays all day. In my G22, I run signal through 2x9 relays when in MS and hard-linked. Never had a relay failing me (we use Takamigawa (sp)).

Oh yes, that's reassuring. This concern is officially struck from the list. Thanks!

Edit: I'm still going to take your advice and simplify. This 'small studio mixer'  is going to need a pallet jack when I'm done. :/
 
john12ax7 said:
Metal work is always the nemesis of electronics.
I've told this story before but when stacking together tens of modules width tolerance matters. For production metal in my last big console, we used a forming tool so the width of the individual strips were held to tight tolerances. That said the hand bent prototype metal was only held to shop tolerances.  ::)

The night before the first trade show had me trying to shoehorn 64 single width strips and a master section into the frame for it's introduction. Random width errors would average out, but I wasn't that lucky...  :eek: We got it together (involving an all nighter and a mallet) and drove it to the trade show with another engineer arriving the morning of the last setup day (we both worked to keep each other awake during the drive). 

When we arrived and carried it into a dealer pre-show already in progress they thought it was a planned dramatic entrance (it wasn't).

JR
 
boji said:
Good story.  Get any grey hairs out of the bargain?
Any correlation to the width problem?  ;D
No, in production the metal was the least of my concerns (hard tooling works).

IMO it marked the end for large hardware (for me at least). I was trying to sell a $40k split console for only $20k but the market wanted to buy a $4k inline (Mackie 8-bus) instead.  The rest is how they say history...

big 24T tape machines are likewise past their prime as digital becomes cheaper and all pervasive. Not much call for a 36x24 split these days (or a couple decades ago).  ::)

JR



 
gyraf said:
For OP's question: all the new-fangled "passive relay volume controllers" run signal through 6-8 relays all day. In my G22, I run signal through 2x9 relays when in MS and hard-linked. Never had a relay failing me (we use Takamigawa (sp)).

One thing I wondered about relay-switched audio paths: is it better to have the "normal" path with the relays de-energised or energised? I can see advantages for both

I would imagine that using the energised gives regular wiping action to keep the switch contacts clean - but that might be wishful thinking

Nick Froome
 
I've built some modules with the relays de-energized as the default.  So you have to activate the relays to bypass the unit.  You conserve some power in normal operation, also normal and bypass current draws are closer.

The opposite way is better when there is a loss of power,  as the unit can still then pass signal.

So in part it depends on your priorities. But I'm also curious to hear others thoughts on this.

 
I am not a big fan of relays but have used a few over the years for hardwire bypass when power is removed... so obviously in bypass without power.

JR

PS: I recall back in the 70s-80s some audiophool puke complaining that he could "hear" the sound of a single switch contact in his audio path... Clearly he has never been within a mile of a recording console with multiple switches in every audio path. If you can hear a switch, you just found a bad switch.  ::) 
 

Latest posts

Back
Top