IC based bypass switch question

[ivle]

Huh... this is the story: a friend of mine gave me one of those Digitech death metal pedals that sound something like an electric shaver (no offence to anyone). after playing a while with the capacitors i turned it into a nice "less gain" unit  , but still    .... so i decided to build a tube screamer clone and put it in the digitech's enclosure ( i'm in love with the solidity of those tanks). And all tube screamer schematics show how to use and wire a heavy duty 3PDT switch but the digitech is built to use a tiny button and an IC to switch to bypass.... SO THE GREAT QUESTION: is there a way to use that tiny bastard with something like an analog CMOS switch like the analog devices ADG888 (OR ANYTHING ELSE) for bypass instead of drilling a hole for a Heavy duty 3DPT switch    Thanks

[R.G.]

Read Geofex.

[mth5044]

Check out the Tone God's Wicked Switches

http://www.thetonegod.com/tech/switches/switches.html

There is another place on his site with layouts of these switches.

[amptramp]

You have several choices here that permit the switching to be slow enough that there is no audible "pop" from switching.  If you use a D-type flip-flop, you can take the outputs through R-C filters to slow the switching transients.  You have three choices that enable you to do this simply:

1. Use a CD4007 as the analog switch using the connections shown in the Motorola/ON 14007 datasheet.  This gives independent control of the gates of the FETs used as analog switches.  This is better than the 4016/4066 analog switches or the 4051/2/3 multiplexers because in these devices, there is an internal inverter for the gates that prevents slowing the switching response time.  See Figure 1:

http://www.datasheetcatalog.org/datasheets/270/397423_DS.pdf

2. JFET switches can be used as long as the pinchoff voltage is low enough and the gates can be slowed down in the same way.  The gates should be driven from an active pulldown and tied by a resistor to a buffer that follows the source voltage (with the sources tied together as the multiplexing point).  Pinchoff voltage may be a significant part of the DC supply, so this method can only be used with low-threshold FETs.  See the National AN-32 app note image "DTL-TTL buffered analog switch" on page 9:

http://www.national.com/an/AN/AN-32.pdf

3. Diode switching can be used with resistors from the flip-flop outputs to the common anode and common cathode of a diode bridge.  Signal can be applied to one cathode/anode section of the bridge and taken out of the other one.  These can be driven through resistors and the slowdown capacitor can go from the common anodes to the common cathodes.  I don't have a link for this, but diode bridge demodulators were common in early FM multiplex adaptors.

In general, stompswitches are the only way to get true bypass operation, but both electrical and mechanical noise is expected if the effect has to be switched on or off during a song.  Simple switches followed by logic are usually cheaper than stompswitches (which cost $9 locally).  A 74C74 dual D-type flip-flop is about $0.50 (plus taxes) and a CD4007 may be $1 here but diodes are $0.01 in quantity, the LED and driver used to indicate which state the unit is in will be no more than $2, so the stompswitch is usually more expensive.  Most effects use buffers for the input and output, so there is a convenient point to take the bypassed signal from and recombine it to.  Note that stompswitches do not have a controlled priority - which switch section switches first and last may be different for different switches, so the switching "pop" may only become evident once the unit is in production and another switch lot is used.  Note also that a 74C74 can be initialized to come up in a defined state on turn-on whereas the stompswitch does not.

[ivle]

Thanks for the info guys.... This is really helpful

[R.G.]

3. Diode switching can be used with resistors from the flip-flop outputs to the common anode and common cathode of a diode bridge. 

Diode switches and modulators are generally limited to signals of below 50mv, and below 20mV for low distortion.

I've recently reverse engineered some of the circuits from the Vox Supreme and Thomas Vox Beatle amps for their tremolo which I liked a lot. This used a diode ring modulator, but it necessarily uses the signal at about a 25mV level and then amplifies it back up, so it's noisier than other methods.  RF signals are generally in the millivolt or below range, so diodes make great switches and modulators for RF.

In general, stompswitches are the only way to get true bypass operation,

Well, there's relays. 

[Mark Hammer]

Here is an old POLYPHONY article elaborating on the fairly standard 4007-based switch used by DOD on a great many of their pedals:  http://hammer.ampage.org/files/dodswitch.gif

[amptramp]

3. Diode switching can be used with resistors from the flip-flop outputs to the common anode and common cathode of a diode bridge. 

Diode switches and modulators are generally limited to signals of below 50mv, and below 20mV for low distortion.

I've recently reverse engineered some of the circuits from the Vox Supreme and Thomas Vox Beatle amps for their tremolo which I liked a lot. This used a diode ring modulator, but it necessarily uses the signal at about a 25mV level and then amplifies it back up, so it's noisier than other methods.  RF signals are generally in the millivolt or below range, so diodes make great switches and modulators for RF.

In general, stompswitches are the only way to get true bypass operation,

Well, there's relays. 

A diode bridge should work at the summing junction of an inverting op amp.  The signal going into the amplifier stage may be on the order of several volts, but the voltage at the inverting input would be within a few millivolts of the non-inverting input, which should be fixed.  The trick is selecting a resistor pair that does not load the signal but ensures adequate drive to bias the diodes.