Best Selector Switch? ..Advice plz!

Which one is the way to go? (Plz vote!)


  • Total voters
    28
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
You may find that one or both start to suffer poorer performance after many years of wear and dirt accumulation.
There is no way to assess this other than personal experience, or the manufacturer giving lifetime data in the spec sheet.

Cleaning with a switch cleaner can extend life.
Treatment with a switch/contact corrosion protection fluid can extend life.
But some of these products can degrade plastics used in construction.
 

ICG

Disabled Account
Joined 2007
You may find that one or both start to suffer poorer performance after many years of wear and dirt accumulation.
There is no way to assess this other than personal experience, or the manufacturer giving lifetime data in the spec sheet.

Cleaning with a switch cleaner can extend life.
Treatment with a switch/contact corrosion protection fluid can extend life.
But some of these products can degrade plastics used in construction.

I see that as a really big 'pro' for a relay, you can much easier replace one or more relays, you don't have to search for a specific type or size of switch. To replace the relays or the (cheap) control switch is much cheaper and a lot faster. The best thing of it is though you have much shorter signal paths. And it's easier to introduce later a remote control for the input switches. Or a LED for the active input.
 
Since the discussion has moved on to talk about relays rather than switches the next logical step would be to go for a solid-state relay. In other words a pair of back to back FETs fed from a photo-voltaic isolator. Wouldn't that be an improvement on a mechanical relay in this application?

Bonsai is using such an arrangement for an output relay for his poweramp - Solid State Loudspeaker Relay - seems to me the FET sizes can be scaled down considerably for signal level applications.
 
The conventional switch option is still open...
While I'm persuaded about the advantages to the relay solution, I'm
just not sure if I'll be able to impliment (fit it) into the packaging...
Waiting on things to arrive, only then will I know...

Don't let that interfere with the question though......
Equally interested in seeing how that route plays out....
 
While I haven't used them in a project yet, I picked up a few NOS 4 position double pole mil spec Grayhill switches. They are fairly small, the body is only about an inch in diameter or so. They are extremely well built, very high quality. I see them listed on Ebay for reasonable money. One will be going into my Aikido octal tube preamp build. I have other selector switches but nothing near as nice as the Grayhill.

BillWojo
 
I always liked the input selector buttons on the Quad 44 preamp, so when I built a DCB1 preamp, which has relays for the 6 inputs, instead of using a rotary switch I used momentary contact push buttons and a logic circuit to energize the corresponding relay. I am very happy with the result.
 
Since the discussion has moved on to talk about relays rather than switches the next logical step would be to go for a solid-state relay. In other words a pair of back to back FETs fed from a photo-voltaic isolator. Wouldn't that be an improvement on a mechanical relay in this application?

Bonsai is using such an arrangement for an output relay for his poweramp - Solid State Loudspeaker Relay - seems to me the FET sizes can be scaled down considerably for signal level applications.

I think a relay would be the simpler approach and make better contact
 
Last edited:
Did you open a Thread to show us how to do that?

No, but I discussed it on the "Mezmerize DCB1 Building Thread" back in 2011, and you provided a lot of help. See http://www.diyaudio.com/forums/pass-labs/176723-mezmerize-dcb1-building-thread-11.html#post2432550

That was almost 6 years ago and my computer died somewhere in there, so I'm having trouble finding my schematic files, they might have been on a drive that never came back from the dead.

The basic idea I used was a D-type flip-flop logic circuit, which has an input, an output, and a clock. When the clock transitions from low to high, the state of the input is replicated to the output, which will stay in the same state until the next clock transition. You can get chips with multiple flip-flops sharing a common clock, and since the Mezmerize DCB1 has 6 inputs I used a hex flip-flop, a 40174.

The "trick" to making this useful is to use your data as the clock, so when you close one of the switches it does 2 things: it puts the corresponding input to a "high" state, and also sends the clock pin high. I used series diodes to isolate the inputs from each other. The chip needs a short delay for the inputs to settle before the clock transition, so I used an RC network on the clock pin. The outputs connect to transistors (via base resistors), and the collector of each transistor is connected to the corresponding relay coil, and to an LED+resistor (I should have used a single resistor in series with all the LEDs for simplicity). The emitters all connect to "ground". So, push button 1, input one goes high, all other inputs are low, clock goes "high", output 1 goes high (all others low) and stays that way until the next button press, which makes Q1 conduct, which closes relay 1 and illuminates LED1. Now press button 3, output 1 goes low, output 3 goes high, relay 1 opens, relay 3 closes, etc.

I found that decoupling on the supply pins of the 40174 was very important for stability. There is an additional pin ~CLR, which resets all the outputs to low when it is low; I connected that to V+ through a resistor with a cap to ground, so it stays low for a few milliseconds when powered on, otherwise the initial state is undefined until the first clock low->high transition. In theory it is possible to depress two buttons simultaneously and thus select two inputs at once, but I have tried and never been able to do it. Exactly once in 6 years it turned on in a strange state, but touching one switch got everything back to normal.

I found a schematic, but the labels are a mess and the layout isn't very good, but I was learning to use the schematic software at the time. I'm sure I had a better one but can't find it.

I mounted the switches and LED's on a piece of project board which mounts to the back of the front panel, and the rest of the circuit on a second board. The hardest part was drilling holes in the front panel that would line up with the switches.

201341_1950142317163_5767021_o.jpg
221221_2029957232486_3514030_o.jpg
 
Wikkid, I picked a few of these up from a member of AK and was really impressed with the build quality. I suggested buying from Ebay only because the price to buy new from Grayhill would be crazy expensive. If you look at there catalog it's pretty clear who is there target market, certainly not us. Mil spec parts are usually built to very high standards, it's the best our money can buy.

BillWojo
 
Bill...
I'm with ya 100% ....Absolutely!!!
MIL spec/surplus definately! For exactly the reasons you mention.
Can't fail quality, at overstock prices...

I'd pull the trigger right away on one, but I'm having trouble finding the right one.
I keep starting then running out of time before looking at the catalog (which I figure I would do, too)..

The one in the second link is only 3 pos... I do need the 4 pos..
The limited sweep... For 2 stereo channels of course..
and the shaft size concerns me a bit.

Hate to admit it, but I got a cheap case for the project, cuz I need to finish it fast...... it has a lot done already (comitted to certain things)... and the dang knobs are metric..
No idea how to easily/quickly adapt..
Some of the Grayhills listed don't specify shaft size..
Takes to much time to go thru the catalog by process of elimination, to find out the shaft size...

Also, for future projects, I'll dig deeper into the relay solution..
But for this one, I need to stay conventional..

Help/suggestions most welcome!!!
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.