Delay tails question

[armdnrdy]

I've seen and implemented different circuit versions of achieving delay tails.

Last night I was reviewing a drawing when the thought occurred...

To allow the delay repeats to cycle through the circuit after the effect is bypassed, the output of the delay circuit could be left connected to the output jack. The remaining repeats would be able to "drain" out naturally instead of being cut when bypassed.

My thought is to wire the delay true bypass, and use a SPST toggle to connect the delay circuit output to the output jack. This will give one the option of TB or tails.

Now...the only thing that I see that would affect the clean signal is the 1M pull down at the circuit output. That should have very little affect.

Is there something that I'm missing that would cause issues?

I used a redraw of Scott Swartz's delay circuit, and altered the output switching for the example:

[Mark Hammer]

Sounds to me like it ought to work.

In the late 80's, I made a rack unit for myself with a bunch of Anderton projects as modules, including a couple of EPFM CMOS switches that I could control with a footswitch, and a 2-input mixer, which formed the master output stage.  I would send one copy of the signal through one of the CMOS switches to a clunky desktop analog delay unit (Yamaha, I think), and feed the delay output back to the mixer.  The switch let me engage or cancel input to the delay line, and whatever the delay line happened to have in its grips was part of the final signal.  "Tails/trails" well before it became a regular feature on pedals.

In my case, I was using the CMOS switch (which I arranged to be either latching OR momentary; whatever I was in the mood for) as you plan to use the SPST, but since I had the patch ongoing, it was as if the overall bypass switch was always "on".  That niggling detail aside, it is exactly what you describe.

[armdnrdy]

Thanks for that Mark!

This seems so simple that I thought....what am I missing here!

After some thought...it would probably make more sense to use a footswitch. A bit more "guitar in hands" friendly. 

[R.G.]

Just for reference, trailing echoes has a fundamental conflict with true bypass.

Trailing echoes by its very definition means that sound goes out on the signal line after the pedal is bypassed. True bypass by its very definition - no electrical connection to the effect when bypassed - forbids this.

The options for trailing echoes get much easier when you remember that true bypass is a solution to a problem that was an issue back in the 60s and 70s, but not so much today. Today, the biggest value for true bypass is to be able to claim it truthfully in advertising to non-technical guitarists who have read the earlier generations of ads.

Trailing echoes with true bypass requires some means of doing something to the signal instantly and then true-bypassing some variable number of delay times after the guitarist requests "true-bypass me NOW!"

It can be done of course, but it gets complicated. Going to a modern means of bypassing is much more easy technically.

[slacker]

I'm afraid that won't work very reliably. The output of IC1A is a low impedance and will be connected to the input via C31, so it will load down whatever is connected. If this is another pedal with a low output impedance it might work but if it's a guitar you'll get huge tone suckage^TM.

Like R.G said you can have true bypass or tails you can't have both at the same time, tails needs some sort of buffered bypass. If you want a compromise I would go with something similar to Merlin's idea in reply 2 here http://www.diystompboxes.com/smfforum/index.php?topic=95632.0 this gives you tails using buffered bypass or the option of true bypass for when you don't want tails.

[armdnrdy]

Thanks Ian..I'll take a look at that.

[Mark Hammer]

Just for reference, trailing echoes has a fundamental conflict with true bypass.

Trailing echoes by its very definition means that sound goes out on the signal line after the pedal is bypassed. True bypass by its very definition - no electrical connection to the effect when bypassed - forbids this.

The options for trailing echoes get much easier when you remember that true bypass is a solution to a problem that was an issue back in the 60s and 70s, but not so much today. Today, the biggest value for true bypass is to be able to claim it truthfully in advertising to non-technical guitarists who have read the earlier generations of ads.

Trailing echoes with true bypass requires some means of doing something to the signal instantly and then true-bypassing some variable number of delay times after the guitarist requests "true-bypass me NOW!"

It can be done of course, but it gets complicated. Going to a modern means of bypassing is much more easy technically.

Jeez, I hate it when you're right!    But you are.  By definition, you can't have a signal that passes through a delay chip at the same time as being a direct connection between input and output jack.

[induction]

I designed a tails switch for my All-Star Reverb circuit that you could probably adapt for your delay. It will usually work in a circuit that has buffered bypass tapped from the input buffer, as long as the effect output includes the (unity volume) buffered dry signal. This includes some reverbs and delays, and maybe some other stuff where the bypass is done by simply disconnecting the input buffer from the part of the circuit that causes the effect.

Here's my reverb schematic for reference:


Here's a switch that toggles between tails and mechanical bypass. It requires a 3PDT footswitch and a SPDT on/on toggle. In bypass, the output jack is switchable between the circuit output and the input jack, but the input jack stays connected to the circuit input at all times. Assuming your input buffer has a high input impedance, this will probably be indistinguishable from true bypass. You won't need a pull-down resistor on the input buffer because it's never switched.


If you want to go all the way to true bypass with the tails switch, you can do that with a 4PDT footswitch and DPDT on/on toggle switch. It's probably overkill, but some people demand true bypass. You might not need a pulldown resistor on the input buffer because you can ground the input in bypass. You can alway include one just in case because popping on a reverb is pretty awful sounding.

[armdnrdy]

After looking over a few options and letting it digest a bit, I ended up moving full circle to what I had used previously.

I looked at Merlin's two switch/FET tails solution and saw that it accomplishes the same thing as a tails mod (soulsonic) I had used when I built an AD-900 delay.

Merlin's tails switching:


To implement Merlin's switching:
Engage SW1 for delay effect
Engage SW2 to make ready for tails
Disengage SW1 for tails
Disengage SW2 for True Bypass

Soulsonic tails mod added to AD3208/5:


To implement soulsonic's switching:
Engage SW1 for delay effect
Disengage (open) SW2 for tails
Disengage SW1 for True Bypass
Engage (close) SW2 to make ready for delay

So...the same amount of "actions" required for both tails circuits.

Merlin's circuit requires additional FET switching circuitry and a switch.


The soulsonic mod involves a switch and a large resistor across the lugs to combat popping.