4066 bypass idea...

[gez]

…that didn’t quite work.

I was looking at the bypass done in some of the older commercial FX units, i.e. flip-flop and two FETs, and thought it might be ‘fun’ to try doing the whole thing with a 4066 chip.

By connecting one side of a switch to ground and the other to the +ve supply via a resistor you end up with a inverter (output taken from the junction of the resistor to +ve supply and switch).  Connect up two of these with a few extra components and you get a flip-flop (see Andrew’s excellent ‘Wicked Switches’ article or RG’s equally excellent one on the 4053) .  The remaining two switches  are used for the audio part, one switch’s input connecting up to one inverter’s output, the other switch’s input to the other inverter’s output.

The disadvantage of this method is that one switch is always on so there’s always a drain on the supply via those resistors to V+.  However, this current can be put to good use and used to turn on a LED (or two!).  I used a tri-colour common anode LED so that each LED is wired in series with one of the inverter’s resistors to V+.  When the effect is bypassed the LED is red and when on it’s green - a sort of Standby/On indicator.  This slaps a couple of mA onto the total current draw when the circuit is bypassed, but looks cool so what the hell!  You could probably get away with using one LED for a indicator and making the resistor connected up to the other switch of large value to reduce current if this is a problem for you.

I was worried that the drops across the LEDs would mean higher resistance for the switches (inputs are pulled up a couple of volts short of the +ve supply), but my fears were unfounded.  I measured resistances of around 63 Ohms from a 9V supply, rising to 99 Ohms @ 5V.

The whole thing is triggered from a momentary switch and works great… except it pops! (sigh).  I followed RG’s advice of biasing inputs and outputs at half the chip’s supply and coupled all the audio via caps.  Reducing the value of the resistors to Vref  didn’t cure the problem either (went as low as 100k).  

I remember reading in a Penfold book that he subbed in a 4016 when he had switching clicks from a 4066 and this cured the problem.  Next time I put in an order I’ll get some and sub one in, but was just wondering if there are any other tricks I can try to eliminate the popping?  Although this isn’t the best method for bypass, I’d still like to get the thing to work (even if I don’t ever use it!)  Ideas anyone?

PS  I don’t claim to have invented this method or anything - if I thought of it many other have I’m sure, plus it’s just a ‘simplification’ of what has already been done - so apologies if I’m treading on anyone’s toes here (just haven’t seen a schematic for the above yet).

[Tim Escobedo]

I've never really tried it, but if you snoop around, you should be able to find PAIA schems floating around for the Motion Filter and Axe Grinder (I think). They use a SPST switch controlling a 4066 for bypass.

[Nasse]

:? There was a preamp project in Elektor few ( :cry: i´m getting old, i´m old fart circuit junkie) years ago, the 4066 switch was put in virtual ground, like between input resistor and inverting input of op amp, maybe arranged for unity gain. But that was mostly done to reduce distortion, and my memory is weak but I have a faint memory that it helped to reduce clicking, perhaps. But I really dunno :roll:

[gez]

I've never really tried it, but if you snoop around, you should be able to find PAIA schems floating around for the Motion Filter and Axe Grinder (I think). They use a SPST switch controlling a 4066 for bypass.

I had a look at the Motion Filter. They don't use momentary switches, which makes things simpler (though I specifically wanted to use momentary as I don't like the mechanical noise of latching switches), but I notice that a trimpot is used to set Vref - cool idea!  I should think this will help get cancellation of any feedthrough (it mentions in the text that it's used to get rid of switch popping).

I'll give this a try, thanks for suggesting those schematics Tim.

Anything else I should know about anyone?

[The Tone God]

R.G. has done something similar with the 4053 at GEO and theres the "Wicked Switches" arcticle at the site.

http://www.geocities.com/thetonegod/

Andrew

[gez]

R.G. has done something similar with the 4053 at GEO and theres the "Wicked Switches" arcticle at the site.

http://www.geocities.com/thetonegod/

Yeah, I know (both great articles).  The novelty here was using two switches in the 4066 as a flip-flop to control the other two switches, plus you get a LED indicator built in - lower parts count!

The 4053 would probably be the better way to go.  I'm wondering if something can be done with the spare switch?  :wink:

[gez]

The 4016 didn't help matters and neither did the trimpot.  I think the cause of the clicking might be the current that flows through the two switches configured as a latch when they're switched.  Perhaps the spikes are being coupled through via gate capacitance?

Another of my 'bright' ideas that doesn't work! (I've got a folder full of them).  Thankfully, they sometimes do work otherwise I'd have given up long ago!

[Nasse]

:shock: Just curious does anybody know if you could somehow marriage high freq pulse with modulation with lowpass filter to remove control ripple control and simple on/oof (voltage hi/lo change). I mean if you add short burst of needle pulses at ultra high over audio freq at on off transient, could this mask the switch capacitance bleed?

Maybe I am talking bullshit and dont know how these chips work, but just a thought. Anyway something like this measn you must say bye bye to simple low parts count and practical circuit. Probaply easier means exist :?