Does Millennium Bypass really affects high gain circuit or tone?

[rogeryu_ph]

Guys,
Is this true? Well I usually order DPDT switch co'z it's much cheaper than 3PDT and also I do have bunch of FET to use. I usually include mil circuit on all of my project yet I also never try using the 3PDT.

Any experience or comment

Thanks,
Roger

[R.G.]

If it's constructed and wired properly, the Millenium Bypass has no effect whatsoever on gain or tone. It can't. When the effect is not bypassed, the Millenium circuit is disconnected from it. The Millenium is only connected to the circuit when the effect is bypassed and at that time the effect is disconnected from the signal path.

It not only doesn't, but it ...can't... change either the effect sound or the bypassed sound.

[rogeryu_ph]

For further clarification R.G. 
When the effect is not bypassed the Millenium circuit is lit and it is disconnected from the circuit and it will not affects the gain circuit even though it share the same ground....?

Roger

[R.G.]

That is correct. If it's constructed and wired properly, which means that the grounds are also wired properly. "Properly" means that the Millenium and the effects circuit do not share common ground wires where the return currents can be picked up from Ohmic voltages in the ground wires.

If you build and wire it correctly, the Millenium makes no change to either the effect gain/tone/sound or the bypassed gain/tone/sound.

Here's an interesting thought. 99.9999% of the current used by the Millenium bypass, all but a few nano-amperes, goes through the LED. In the 3PDT, 100% of the current goes through the LED, the only difference being the high leakage diode current. So essentially the same current flows in the ground wire of the Millenium as does the ground wire of the LED with a 3PDT. So the Millenium and a 3PDT have almost exactly the SAME chance of causing some change to tone/gain/etc. They both share power and ground with the circuit. They both have about the same current, whatever goes through the LED.

No difference.

[rogeryu_ph]

I usually include the mil circuit into the PCB. Before itching the board I draw the mil circuit to the same +V pad of the project PCB then tap connect to the nearest project PCB ground. Is this proper or not by what you meant Sorry for being such a noob..
Also thanks for your patience R.G.
Roger 

[Faber]

So you were saying that both 3PDTs and DPDTs with Millenium have the same chance at affecting the tone, and the 3PDTs have the LED wired to the same ground as everything else (unless I've been doing something horribly wrong).  If my thinking is correct, then should it really matter where the grounds are connected?  I thought all the grounds went to the same point anyway?

I'm adding a Millenium to my first pedal within the next week, and I just want to make sure I'm doin in right. 

Yes, R.G. thank you for your patience.

[R.G.]

I usually include the mil circuit into the PCB. Before itching the board I draw the mil circuit to the same +V pad of the project PCB then tap connect to the nearest project PCB ground. Is this proper or not by what you meant

Drawing it to the nearest ground trace is OK most of the time, but sometimes will cause a click when the LED turns on or off. It is best if the ground for the Millenium - or the 3PDT, for that matter - goes back to where the power ground comes onto the PCB.

So you were saying that both 3PDTs and DPDTs with Millenium have the same chance at affecting the tone, and the 3PDTs have the LED wired to the same ground as everything else (unless I've been doing something horribly wrong).  If my thinking is correct, then should it really matter where the grounds are connected?  I thought all the grounds went to the same point anyway?

And this thought is right at the fundamental issue for understanding why some layouts are better than others. You're close to a better understanding of electronics all around.

I make a distinction in my mind between the "Perfect World" and the real world we live in. In the "Perfect World" everything is exactly as the math, geometry, calculus, etc. concepts says it is. All lines are perfectly straight, so straight they can be used to measure the crookedness of a laser beam. All measurements are exact. All surfaces are perfectly flat. All parallels are in fact parallel.

In the Perfect World, copper is truly zero resistance, and all circuit traces have no capacitance or inductance. In the real world, we know that copper has a very low resistance, and we decide to ignore what small resistance it does have. That works most of the time. But copper does have some resistance, and it obeys Ohm's law. Remember this, we'll come back to it.

At one time, "ground" literally meant "drive a metal stake into the dirt and connect your wires there". It still means that in AC power wiring, for safety reasons. In signal electronics, we have relaxed that to mean that ground is a place we decide to call "ground", and all other voltages are measured from that spot. Literally, it's a place we decide is a source of 0.000000000V so we can clip our meters and oscilloscopes there for reference purposes.

Here's the critical part: THERE CAN ONLY BE ONE GROUND POINT IN ANY SYSTEM. Since our best practical material for connecting things electrically is copper, and all copper is a resistor, we can only connect things to ground with resistors. And all resistors cause a voltage drop if there is any current running through them. So all other points in a circuit except where you have your voltage reference lead hooked are not at ground - they can't be because they're separated from ground by the voltages in resistors leading to the reference point.

This is the thinking that leads to the idea of star grounding. Each section of the circuit that has a "ground" symbol in the schematic replaces that symbol with a copper wire running to the one, true ground point. All ground wires radiate from the single point ground like a star.

By now you're thinking "R.G., you're off playing around in the Perfect World again." Not so. I'm being very practical. The point here is - no, all of those places on the schematic diagram which have a ground symbol are NOT at the same voltage. They are separated by resistances of various low values. If you don't know what currents flow through the resistances, you can not predict what voltages appear at the ground symbol in the schematic. And if that ground symbol is the ground for a sensitive input, you WILL have funny noises, clicks, hiss, crosstalk, hum, and possibly self oscillation; this is especially true if your circuit is a high gain distortion circuit.

The key here is that you must know what current flows in what wire. When you switch an LED, the LED pulls a big (compared to the rest of an effect circuit) current, and it pulls it suddenly. That causes the voltage at the LED from the battery to droop a little because of the wire from the battery. It also causes the "ground" at the negative end of the LED switching circuit to raise up above the one, true ground by the amount of the LED current times the resistance of the copper leading to the ground reference point.

If you have connected your LED ground to the closest ground in the layout, and by some coincidence that happens to be the signal reference ground for the input stage, then the ground for the input stage jumps up and down as the LED turns on and off.

If you have an opamp driving a lot of current into an LED/LDR module and it's being driven by the full wave rectified signal voltage for an envelope filter, then a tiny replica of that current signal appears on the "ground" wire connecting to the opamp/LED. Connect that "ground" to something sensitive, like the bias voltage generator ground, before it goes to the true "ground" point, and now you have the unwanted signal on the bias voltage of ALL of your circuitry.

I mentally classify grounds into:
- safety ground; this is a wire which connects to the AC power line safety ground and keeps me from dying when I mess up. It connects only to the metal chassis of AC powered things
- true ground; this is where you clip your meter or oscilloscope lead for accurate readings. It is usually the ground point of the power supply, but may be the ground side of the input jack.
- signal reference ground; this is the "ground" that you get by running a separate wire from the true ground point to some place where you want no interference on ground, like the input jack, or the bias voltage generator.
- sewer ground; this is how the used-up electrical power that runs into the + lead of ICs, transistor circuits, etc. gets back to the power supply. There are all kinds of things in most sewers, usually ugly if you look at them, and that's true here. You do NOT want to contaminate your signal references by connecting them to "look into" the sewer.

All of this is of critical importance in something like a studio. But it's worth a few thoughts in an effect PCB. Circuits as small as two transistors can oscillate uncontrollably if you get the ground wrong. If you want your indicator circuit to NOT affect your audio, run its power and ground wires separately to either (a) where the power comes onto the PCB or (b) the main power supply filtering capacitor on the board or (c ) its own local decoupling cap, and then just the ground separately to the PCB power.

That will prevent contamination of your signal.

[rogeryu_ph]

R.G. thanks.
Very well explained indeed

Roger

[Drake120]

So, for example in this layout:
http://tonepad.com/getFile.asp?id=7

not WHOLE of the green trace is really "ground"? Only the place closest to "GND" pad?

Thanks for all your work, R.G.

JS

[R.G.]

Here's the way to understand it.

Replace all of the copper traces with a bunch of little resistors in series. That's what's really happening.

To help us learn, we make up simplifying rules. One of those rules is "Things connected by copper wire have truly zero ohms between them."

This one is almost true, for the little circuits we use in effects. But it's not completely true, as there is a little bit of resistance. And sometimes even that little bit matters.

The only way two places can have the exact same voltage is if (a) they are connected by a conductor  and  (b) there is absolutely zero, no current at all flowing between the two points. So when the power is on and current is flowing, there can only be one place that is really, truly 0.00000V. All other points on the "ground" trace are separated from the true ground by slight voltages caused by the current flowing and Ohm's law.

Most of the time, the voltages are so tiny that it doesn't matter. But sometimes it does. We have to remember with each special simplifying rule that sometimes the simplification makes it too simple.

This reminds me of Albert Einstein. He said "Everything should be as simple as possible. But no simpler."

[Drake120]

Thanks a lot. Another thing is now clear to me

JS