led true bypass with a dtpd questions

[cakeworks]

I am a little (very) confused as to the benefits of the simple circuits over the even simpler circuits used for getting an led to indicate on or off.

Okay so what's the difference between the millenium bypass and the led trick? if the led trick is simpler, why is the mil bypass so popular???
Why couldn't i just put a resistor in between a positive 9v current and the led?

thanks if you took the time to read and answer this.

[Stephen]

Because as soon as you plug in the pedal the light will be on... All pedals are on when in bypass too...You want it to come on when the effect is on!!!   Enter millenium bypass...3PDT has an advantage of an extra set of connections..

[The Tone God]

Your suggestion will tell you when the effect is powered up but as you have no connection to the switch you will not have an indication of effect state.

Andrew

[R.G.]

The differences are mostly in how much of a pop the LED circuits cause.

Reading "The Technology of Bypasses" and the articles on the Millenium at GEO would help as background info.

(1) True bypass does not use all of a DPDT. It uses 3/4 of it. There is one switch closure or opening left over from true bypass with a DPDT.

(2) The leftover portion of the switch can be wasted by using a wire to connect two same-way throws of the switch

(3) The leftover portion of the switch can be used to shunt the input or output of the effect to ground when the switch is in bypass.

(4) The Millenium uses the leftover part of the DPDT to switch the output of the effect to the LED circuitry.

How (4) works is critical to understanding the advantages of the Millenium versus its precursor the Ratt bypass and others like Joe's LED trick. All of these variants use the fact that the output of an effect can be pulled to ground by a resistor, and that this is in fact almost always done. The resistor is used to pull the input to the LED switching circuitry to ground, causing it to switch the LED.

The resistor to ground has to be some size other than zero ohms, or the effect could not drive it, since it also loads the effect when the effect is engaged. So practical values of this resistor are somewhere from 10K on up to many megohms. Lower is better for switching the LED, but worse for loading on the effect.

Given that the pulldown resistor must be a resistor and not a hard contact to ground, how much current the resistor pulls out of the LED switching circuit determines how low the input to the LED switching will be pulled. And the output capacitor to the effect will be pulled to that same small voltage by the bias current of the LED switching.

So the critical thing in variants of the fractional-DPDT indicator circuit is how low the current from the LED switching circuit can be. That determines how big the pulldown resistor can be, and also how big a pop you will get when you re-engage the effect from bypass. The input circuit from the LED circuit will pull the effect output capacitor up to that current times the pulldown resistor, and that voltage is heard as a "pop".

In my work on the Millenium, I realized that this pop was inevitable. The only thing to do was to make it inaudibly small. The original Ratt bypass used a darlington transistor with a 10M pullup and a 10K pulldown resistor. The 10M was enough to turn the darlington on, and the 10K was enough to pull its 9V bias down to roughly 1/1000 of 9V - 9mV. Guitar signals are on the order of 100mV, so that's smallish, but it can be heard under some conditions.

My innovation was to realize that I needed something that supplied less than one uA of bias current and that was realiable at that. After a lot of navel-staring and meditation, I remembered that a reverse biased diode leaks a pretty constant amount. For silicon it's nano-amperes or less. But there is a class of signal switching diodes which are deliberately doped to leak more as a side effect of fast switching. The ordinary 1N914 and 1N4148 among others are gold doped, and leak something between 10nA and 25nA. This is perfect for a bypass if you have a switching device that will switch on it. As we know, both JFETs and MOSFETs will do this. So a reverse biased signal diode will turn a FET on and a resistor will have an easy time pulling that 25nA to ground. The Millenium was born.

The Millenium will hit that 9mV of the Ratt bypass at a pulldown of 360K ohms if you have a full 25nA leakage. Most modern diodes ahve far less than that, but the makers spec it at 25nA because they don't want to test for such a small, esoteric specification. Most of them are actually 3-20 times better than that. So the Millenium is good in almost all cases up to a pulldown resistor of 1M.

With that long winded explanation, the differences in the LED-with-DPDT are in the amount of bias current they need to switch and how it's pulled to ground. If your effect has a low output resistance to ground, 10K or less, they'll all work great. If your effect has a problem with loading, or you want even quieter switching, the Millenium offers about the best there is in this style.

You'll hear me say this a lot  - the devil is in the details.

[cakeworks]

i'm copying and saving that tasty morsel of information right now. thank's everyone. especially R. G.