Electronic by-pass switch

[j.c.]

I've found two web pages wth the squematics of a  Electronic by-pass switch.
In the two pages it is mentioned they are cheap and noiseless solutions.
the pages are:
    http://members.shaw.ca/roma/switching.html
    http://www.ee.washington.edu/circuit_archive/circuits/guitar.html


I have a doubt in the circuit. of page http://members.shaw.ca/roma/switching.html.
-Could anyone explain me the "Using the CMOS 4053" and the "Using a Push Button" ?  What are the points 1 and 3 for, what is the diference between the two circuits, and what is the input voltage?

-In your opinion what is the best circuit?

P.s.- Here in Portugal i can't find a DpDt switch and this look like a good solution.

regrads

[R.G.]

Have you read the bypassing info at GEO?

[j.c.]

Yes. i have pass my eyes on that article, but that talks almost about mechanical switchs.

What is the good things and the bad things of electronic/mechanical switch?

[R.G.]

Mechanical switches are simple and rugged, but expensive and hard to find. They have a high ratio of off to on resistance (infinite to millihoms), which is good. They do not have one of the pop-causing mechanisms, charge injection, because the "signal" that makes them switch is not electronic, it's mechanical.

Electronic switches are more compex, only as rugged as the rest of your circuit, but are cheap and easy to find parts for. They have a modest ratio of resistance, essentially infinite to a few hundred ohms, which is perfectly usable. They have charge injection, so you have to design to minimize that.

[gak]

As I can't found any good DTDP I always use the 4053 circuit in the page you linked with a STSP switch, it works perfect, it's a simple build and is cheaper and smaller than the DTDP switches I could find
The pad marked as 1 is where you connect your effect input, and the 3 connects to the effect output.

The difference between the first one, and the other using a push button is that the first one uses a switch with to fixed positions (like common DTDP switches) and the other one uses a normally open momentary button that changes the state of the logical gate every time gets closed, like in boss or ibanez boxes.

I would recommend you the first one, it has less component parts so it will be smaller and a bit cheaper. It works perfect, doesn't make any pop or delay when you switch. I'm so happy since i found this alternative to extremly hard to find and expensive DTDP switches. I etch a couple of PCBs for this circuit every time i'm etching, they are always useful  

[donald stringer]

check your private messages jc

[R.G.]

did you see the CD4053 bypassing article at GEO?

[j.c.]

No that article i din't see. were is it? Could you tell me the page?

[R.G.]

6/14/00 -http://geofex.com/Article_Folders/cd4053/cd4053.htm

[j.c.]

thanks for the page.
Looking at the two squematics, they are a bit diferent.
The one from geo uses a 1m resistor while the one from http://members.shaw.ca/roma/switching.html uses 0.1k resistors, and as far as i can understand the geo one uses a Vr while the other doesn't, the one from the page i said as alredy a led to indicate that is bypassing.
-what are the diferences between the two in terms of work, noise, etc.
-Using a SpSt switch whats the voltage in pins? 0 Volts and 9 Volts?
-Whats the best for me to build? (I don't have much experience in electronics)

Thanks

[j.c.]

does anyone know the anser to my question?

-The spst switch will be an on-off switch, is that right?

[zeta55]

I've tried them both. Had problems with the one at GEO. The other one I use quit frequently. It gives no noice and no pops.

[The Tone God]

The roma circuit has a major error(s) with the 4049 version in the control logic. It won't work with that board.

To add to R.G.s information read the Wicked Switches article.

The Tone God's Domain

There are boards availible for download which have been tested.

Andrew

[R.G.]

All electronic switches suffer from popping because of charge injection. Charge injection is simply the fact that the signal that causes the signal to switch leaks into the audio path a little bit. The whole game in electronic switching is to manipulate that "little bit" down under the audible (or irritating) level.

CMOS switches do the least charge injection when the switch inputs and outputs are biased approximately at the middle of the chip power supply. There is some point where charge injection from the positive and negative sides nulls out to zero, but it's different for each manufacturer and probably each batch. But dead middle is almost as good as perfect.

The charge injected is just that - a packet of charge, a little blip of current. If you think about it, the lower the resistance to signal ground of the signal line where the charge is deposited, the lower the audible voltage blip.

So you pick your resistances. I picked 1M for less treble loss. 100K is better for lower switch blipping, and you can use that if you can live with the treble loss. 10K is even better for popping, but you'll get bad treble loss.

It's a balancing act.

Another thing that works is to slow down the switching speed. This is what they do with JFET switching to keep it quiet. Unfortunately, the control points that you'd need to do that with CMOS are not available outside the chips.

[j.c.]

Finaly I've build this switch. I used the squematics from the 1ª page but i changed the 100k resistrors for 1m resistors.
I'm using this switch whit the lava rim 2 board.
 I don´t know if it is working because my guitar is being repared, but i connected the box to the amp to see if i could notice any diference, and i notice this things.
     -1--> I hear a "pop" when i press the switch.
     -2-->  When the effect is on the sound from the amp is more "clear", i can't here the "rain" anymore.
     -3--> In a certain value of the pots i can here a noise that i think its from the phone, Its that sound that we hear when we are near a sound sytem and our phone is about to ring.

For 1 and 3 What can i do to sotp this?
For 2 is this normal for this circuit?

[zeta55]

So you built the "roma" switch using 4053 IC with 1M resistors?
I always used mine with the 100K resistors, it does not pop when I hit the switch. I once made bad solder joint on the GND connection to the IC. Then it poped realy bad.

[j.c.]

Yes i've built with 1m resistors because i read the R.G. post and he said the he uses 1m resistors.
One thing you used a 200, 2k and 100 ohms resistors, right?
in the roma circuit he says one value for the resistors and the color code other.

[zeta55]

Yes i've built with 1m resistors because i read the R.G. post and he said the he uses 1m resistors.
One thing you used a 200, 2k and 100 ohms resistors, right?
in the roma circuit he says one value for the resistors and the color code other.

Ahh, I never noticed that with the colour codes! Here's what I used: 100K for all the grounding resistors, 200ohms for the switch and 2.2K for the the led. I never used any 100ohm resistor!

[petemoore]

Further looking at the GEO article, the Schema's became clear [i think], and so of course I have questions that already have answers?
 Could a Pot connected:
 lug 1 to V+
 Lug 2 to CV [would this reliably provide <3v / >6v
 Lug 3 to V-
 provide control voltage to activate the switches in a CD4053?

[j.c.]

zeta55 you are right, i forgot the "k" on the value of the resistor is 100K insted of 100 omh