channel switching

[rathmann]

Anyone have experience, opinions, thoughts, warnings, etc. on using Vactrols for channel switching?

The project idea will already incorporate some VCT5C3s for the vib/trem LFO.

My last project was a channel switcher which I did with relays. That seemed to go ok and the switching is fast and silent (ground shunting, "flyback" snubbers, and no "hanging DC" on caps). But for a number of foolish reasons the relays are big 4PDT mechanical socket jobbers. (One of them switched a number of elements in the power amp so it needed to handle some current.) It was a bit of a pain when I got to the layout stage of the design process having to deal with finding a place to mount them and then route everything there and back.

In hindsight I realize that a bunch of SPST or SPDT relays can just be switched in parallel to achieve the same effect, and then the individual relays can be physically located near the components they're switching.

The preamp I'm going to do will be heavily inspired by the Halfer Triple Giant which uses VCT5C1s for switching. I was just wondering if there would be a further space or layout advantage to using these. Two thoughts on the matter are 1) I've already done relays so I know how to do it right and can improve on my first experiment by using more discrete reed relays, whereas 2) I've never done this with the optoisolator method and it might be a cool learning experience while also saving space in what is growing into a pretty complex build.

IIRC - Kevin O'Connor kind of poo-poos optos for channel switching in favor of relays or transistors, but I don't remember the details. I'll have to dig out my TUT to revisit this.

Any thoughts on the matter appreciated!

Thanks

Steve

[defaced]

Personally, I have not used them. I like relays for marginally meaningless reasons. But the SLOClone forum guys use optoisolators alot for channel switching because that's what the original SLO used. If you don't get alot of info out of here, that may be a place worth looking.

That aside, it's interested to see the price difference between Newark and Allied (the only two places I found that carried the VCT5C3), $7.38 and $2.50, respectively. Looks like it's worth looking around for a solid price.

[alice]

I have used them with good success. It is important to get the right model with low "on" resistance and quick turn on (which is slow by relay standards). VTL5C1 I think. They are a little of a pain to put on the board and a little. The biggest drawback is that is that they are only SPST.

[rathmann]

yeah, while looking around for other info I found a post in a thread on another forum where someone pointed this out. there was a photo of the vactrol next to a DPDT relay. obviously the relay was bigger but to get the same functionality would have required a number of optos which would have negated any space savings.

I'm thinking of simplifying the preamp now anyway and eliminating one tube. I've been using a series/parallel switching scheme (of the first two stages) in three other amps with good result so I might just do that, but make it footswitchable. A single dpdt relay should work fine here.

Thanks for the replies.

defaced - for some reason the phrase "marginally meaningless" stuck with me all day. Sounds like something out of Hitchhiker's Guide To the Galaxy - reminiscent of Ford's entry about Earth, "mostly harmless". I'm intrigued by what those marginally meaningless reasons might be.

~s

[defaced]

More than anything, simplicity. I can visualize a relay, I can't visualize varying resistance. I like that the relay is either "on" or "off", not "well, I'm almost a wire" or "well, I'm almost an open circuit".

Also, in a schematic, it's easy to see what's going on with a relay. With an optocoupler, at least the schems I've seen, it's like pulling teeth, especially when simple tonal things are happening like cathode resistors/caps or bypass caps. What's on? What's off? I don't know because nothing opens or closes, it just changes state but it's always in the circuit.

Footprint helps too. Relays have their contacts under them (reduced footprint) and can be socketed, optocouplers can't. If you blow a relay, pull and replace. Blow an optocoupler, and grab the iron. For a turret board that's a non issue, but for a PCB, that can a nice feature. All of my switching stuff goes on PCBs I designed, the amp circuit goes either PTP or turret boards.

And lastly, I like the click of a relay. Simple diagnostic method, and it sounds cool, like shifting gears or something.

[Enzo]

Then there are JFETs. Unlike a DPDT relay, you can put the two JFETs here and there, right by thier switched circuits. And you can wire together as many gates as you want for one control signal to switch as many as you like.

[defaced]

What kind of JFETs are typically used for this application?

[Don Moose]

FWIW, the Yamaha T50 I have (Soldano design) uses Optos.
In fact, Mike used a switching voltage high enough to put all the LEDs from one 'throw' in one series string, and the LEDs in the opposing 'throw' in another series string.

The advantages to this approach are that the current you're flinging around to execute the switching never spikes or exceeds the 20mA If of the Opto, and you can daisy-chain the wiring instead of running pairs everywhere.

Hope this helps!

[Enzo]

JFETs - J112, J175, etc. ANy of that family will work.

[rathmann]

Hmm...the plot thickens. I'm going to have to dig out my TUT now. One of the other advantages of relays is that I know how to make a 6VDC supply for one. JFETs are totally foreign to me, but I can recognize the benefit of their compactness and flexibility.

[Enzo]

If you are familiar and comforatble with relays, then design with relays. Nothing wrong with relays.

Crate uses JFETs a lot, especially in their SS amps. Look at Q6, Q7 in the GX1200H as classic switching elements selecting channel A or B to feed into IC4.

Look at the output of the reverb circuit in Peavey Classic 30. That is a "classic" example of using a JFET as a switch - it turns the reverb off and on.

[tedmich]

here are some cheap ($0.50) audio optocouplers:
Audio Opto Coupler-The Electronic Goldmine


I tend to use DPDT relays, I have some that are ~7mm square from a job a while back...

[d95err]

JFETs are (according to Kevin O'Connor) close to ideal switching elements. Easy to control, low current requirement, no noise, cheap, etc.

The problem with JFETs is the voltage rating. A typical JFET has a maximum rating of 40V or so. This means you have to be careful about where in the circuit you put them. It's important to consider the "worst case" scenario for the JFETs (i.e all knobs on 10 and amp being pushed by a booster hard enough to clip the first gainstage).

[lowell]

could a mosfet be used for switching needs above 40v?

[DIY_Guy]

Originally Posted by lowell

could a mosfet be used for switching needs above 40v?

I used Vishay MOSFET SSR's in a few of my builds to handle switching duties.

I'll probably use them in my next also.

[wizard333]

Anyone have a sample schematic using JFETs for chanel switching?

[lowell]

I don't but can tell you that they can be used quite simply for shunting. Connect the source to ground, drain to the signal line (e.g. tube grid - in parallel w/ grid leak via a capacitor), and a negative control voltage to the gate that will turn the jfet on and off. Make sure there is a cap between the drain and the signal line so the jfet only affects AC. You will need to taylor the cap's value to surrounding resistors using the Fc equation so that ALL the AC signal is shunted. (1/2pieRC) It'll probably actually be fine if you use the same value as whatever coupling cap comes from the previous stage. I believe -15v will suffice for a gate control voltage but not really positive about the control voltage. When the negative voltage is on the gate the jfet is off, therefore your signal is going into the grid of the tube. When the gate has no negative voltage (0v) it will be like a closed switch and the signal will short to ground.

[lowell]

I just reread KOC's info on this. He says to use a series and shunt jfet together. I don't see how this is necessary if you're using a shunt arrangement. Anyone know bout this?

Also, can someone suggest a suitable mosfet for switching needs? They will need to be able to switch an amplified zener in and out of the screen and plate supplies.

Finally, can a simple supply for switching jfets and mosfets on/off be tapped of the bias supply w/ a reverse biased diode, 2 resistors (voltage divider), and 1 capacitor? I'd think this'd work but maybe there are unforseen issues with it's simplicity.

[Jazz P Bass]

Here is a link for N channel fet function.
P channel is the opposite function.
Link:SCEN103 -- n-channel MOSFET switch