Ad Widget

Collapse

Announcement

Collapse
No announcement yet.

Channel Switching

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Channel Switching

    I'm about to build a Mesa Mark IIC+ clone, but I'm thinking about avoiding the optoisolators this time... (too pricey).

    Anyway... What relays are people using for this sort of thing these days? Any hints on what I need to do to avoid pops & other noise?

    Thanks.

  • #2
    I've seen 5v Omron relays in a hand full of channel switching high gain amps I've looked at recently. I can't help too much on the pops since I have limited experience actually implementing relays for channel switching.
    -Mike

    Comment


    • #3
      I use 12v relays with a dedicated power supply (usually a Radio Shack 12v transformer). There is more on this style of switching than you could read in a day in the Dumble section of the ampgarage. You can go 5v off the power transformer too but I found that 12v gives me a little more slop if the voltage regulation is iffy.

      There are a couple rules of thumb on popping and other noises:

      1) You want a ground path for grids if the switch cuts off their usual source. Hanging a 4m7 resistor from the grid to ground won't audibly cut the signal but it keeps the grid from developing any potential to pop on the way out. This especially matters if the channels you are switching are out of phase.

      2) Most of the guys I've chatted with keep the relay grounds completely self-contained (never grounding on the chassis). I've done both without any trouble. On the former, use XLR jacks. On the latter, a regular mono or stereo guitar cord works fine. My favorite power supply is a 12v trannie off the on-off switch to a FWBR and a 1k uf/25v filter, through a 7812 regulator and another 1k cap to the relay hot. The trannie is going to put out around 14v with no load and the regulator keeps the volts up even when you have several relays on at once. If the volts are still too high, you can hang an ordinary diode in reverse series on the regulator ground to knock a volt or two off. Works like a zener.

      You can put panel switches ahead of the footswitches. SPDT with the relay ground on the pole, footswitch on one side and ground on the other. Some guys like putting the indicator LEDs in series with a dropping resistor but I prefer running a 50R to 100R resistor in parallel (for a 1.8 forward volt LED). The resistor puts up enough of a fight to light the LED but if the LED croaks, I still get my solo in.

      3) Finally, keep the power supply wires away from signal wires.

      Hope this helps, Skip

      Comment


      • #4
        Thanks to you both, this helps!

        Comment


        • #5
          You also want to give the coil's flyback current someplace to go - usually a reverse-biased diode as close to the coil pins as you can get it.

          (reverse-biased --> stripe to (+))

          Hope this helps!

          Comment


          • #6
            Originally posted by Don Moose View Post
            You also want to give the coil's flyback current someplace to go - usually a reverse-biased diode as close to the coil pins as you can get it.

            (reverse-biased --> stripe to (+))

            Hope this helps!
            It does, thanks!

            Comment


            • #7
              you can even parallel the diode with a .01 disc. sh

              Comment


              • #8
                I use 5V relays, powering them by rectifying the AC heater filament circuit using the circuit below, again incorporating a voltage regulator.

                To reduce background noise/pops, I agree with much of the above poster's observations. personally, I use the following "rules", which work for me. i get absolutely silent, instant switching.

                1/ Always give the unused grids a path to ground.

                2/ Use a separate grounding point to the chassis for relay circuit grounds, away from audio signal grounding points.

                3/ put 0,1uF cap across the footswitch contacts - "softens up" the switch a bit, reducing the chance of "pops".

                4/ Don Moose's point of "You also want to give the coil's flyback current someplace to go - usually a reverse-biased diode as close to the coil pins as you can get it."
                Attached Files

                Comment


                • #9
                  I use 5V relays, powering them by rectifying the AC heater filament circuit using the circuit below, again incorporating a voltage regulator.

                  To reduce background noise/pops, I agree with much of the above poster's observations. personally, I use the following "rules", which work for me. i get absolutely silent, instant switching.

                  1/ Always give the unused grids a path to ground.

                  2/ Use a separate grounding point to the chassis for relay circuit grounds, away from audio signal grounding points.

                  3/ put 0,1uF cap across the footswitch contacts - "softens up" the switch a bit, reducing the chance of "pops".

                  4/ Don Moose's point of "You also want to give the coil's flyback current someplace to go - usually a reverse-biased diode as close to the coil pins as you can get it."
                  All this is nice and true. Unfortunately avoiding pops and click is not always possible no matter what you try especially in high gain amps and especially if you switch tube grids that's why almost every amp now has a simple mute circuit to completely eliminate relays pops and clicks. If you're somewhat into uCUs it's even easier.

                  Comment


                  • #10
                    Gregg -- tell us more. sh

                    Comment


                    • #11
                      Gregg -- tell us more. sh
                      For mute circuits check out all recent Mesa schematics, also Peavey, Engl E530 and others.

                      Comment


                      • #12
                        For your mute, you can use a JFET as a shunt element, driven by a one-shot (NE555-equivalent, google 'monostable multivibrator') in place of an LDR.

                        Powering the one-shot in such a way that it can make/switch the negative gate voltage the JFET needs to turn OFF may be a trick. The N-channel JFET is ON unless the gate is pulled below the other two terminals by some few volts. Since the lowest swing may be near ground, you have to be able to feed it a negative voltage to get the thing to turn OFF and stop shunting your signal to GND.

                        Hope this helps!

                        Comment

                        Working...
                        X