CD4066 Bypass w/buffer..the results

[dschwartz]

Hi all..
As i said in my previous post about stompswitches going bad, i designed a Bypass system with a Flip Flop using a 4049 IC, a 4066 for switchingn and a fet buffer to avoid any loading of the signal..this is the schem:


as you see, is fairly simple..i based the design on one of Tone God´s articles, but i changed the switching adding a shunt for FX in..i didn´t couple the FX In because most fx boards have dc blocking anyway (if not, add a coupling cap). For the buffer, i used a Fet for it´s high input impedance, litlle board space, and simplicity..i know opamps do a better job, but i wanted the board to be as small as possible.

- question..what if i use a bjt for buffer? will it load the guitar signal?..

you can use any decent spst momentary switch...

this is the silk screen (warning: 2.2M resistors looks like 22M..they are 2.2M...on the schematic are the 1M):


and here is the pcb ..very small...:

http://diynoise.googlepages.com/bypassPCB.pdf

please , comments, improvements and corrections are welcome..this has not yet been verified..

[Ripdivot]

On the schematic I think R5 should be 100K or R6 should be 10K in order to get the 4.5 volt reference...

[dschwartz]

oh, you´re right!..i updated the schem....and note that the IC is a 4069, not a 4049...livewire doesn´t have 4069 model ...

[R O Tiree]

Daniel - did you create terminals for those "4.5V" tags, or did you get them from "voltage rail" in the Power Supplies menu? New Wave Concepts used the same symbol for both, which was lazy of them and is confusing for users.

In that software, a "Voltage Rail" is merely a perfect voltage source, so there would be no "connection" in the software between R5/R6 and R4, for example. It wouldn't matter  what values you had for R5 and R6 - the junction would always be a perfect 4.5V, as would the voltage at R4.

If you used "Terminals" (from the Connectors menu) then all is well.

It's not explained very well in the Help file.

[slacker]

I'd breadboard that before you build it. You might find it pops because you haven't got a consistent DC voltage on all the CD4066 inputs and outputs.
I would add a cap to decouple the source of the FET from the junction of IC2a and IC2b. Then add a resistor from there to 4.5 volts, then add another resistor from "fx in" to 4.5 volts. That way all the switches have the same DC voltage on them so you shouldn't get pops caused by changes in DC levels.

[tempus]

question..what if i use a bjt for buffer? will it load the guitar signal

I would say yes - I don't think you can set up a BJT to have high enough input impedance to prevent loading.

[earthtonesaudio]

I'd breadboard that before you build it. You might find it pops because you haven't got a consistent DC voltage on all the CD4066 inputs and outputs.
I would add a cap to decouple the source of the FET from the junction of IC2a and IC2b. Then add a resistor from there to 4.5 volts, then add another resistor from "fx in" to 4.5 volts. That way all the switches have the same DC voltage on them so you shouldn't get pops caused by changes in DC levels.

+1

question..what if i use a bjt for buffer? will it load the guitar signal

I would say yes - I don't think you can set up a BJT to have high enough input impedance to prevent loading.

A BJT configured as a follower will have very high input impedance, so high that the base bias resistor typically dominates the equation.  However, FETs are lower in noise at high impedances, which makes them more attractive as input buffers.

[dschwartz]

ok, by part:
- R O Tiree: I used the voltage source icon..but actually i´m taking it frtom the 9v source..I.E, i´m using them as connectors..if you check the board, there´s a voltage divider with 2 10k´s (they are not together..there´s a jumper wire called Vr that makes the connection).
- Slacker: maybe the design is too optimistic expecting a 4.5 V bias from the output of the buffer..i´d think a BJT will give a better reference?.. adding other cap and resistor will make the board bigger than what i want..

[R O Tiree]

OK, maybe this will make clearer what I'm talking about. I put together 2 voltage dividers. In both, I used a "Voltage Rail" of +9V for the power source. In the one on the left, I also used a "Voltage Rail" for the junction of the 2 resistors. In the one on the right, I used a "Terminal" to denote that junction (it's not unusual for us to label this point in our circuits as "4.5V" even though it often isn't). As I said, it was lazy of NWC to use the same symbol for both, as it gets confusing.

As you can see from the resistor values, the junction should be at 3V but using a "Voltage Rail" (see the drop-down menu) this forces the junction to be 4.5V (red trace). The one on the right is sitting at 3V, as we'd expect, even though I labelled it as "4.5V" (blue trace).



If however, you use a "Terminal", as I did in the right hand one, then you can call it what you like - it will simply connect all the points similarly labelled and the voltage will be whatever the circuit dictates. You can find it in the menus here:



From your answer, I know you used a "Voltage Rail" for your 9V supply, but it looks as though you might have used "...the voltage source icon..." for the 4.5V connections as well. If so, then each of those "4.5V" points will be a perfect 4.5VDC source, not a connection.

[head_spaz]

R.O.Tiree,
Excellent point !!! And very well demonstrated too.
Your little tip will help me tremendously when using livewire. I was ready to chuck it!

[dschwartz]

R O, i know exactly what you are talking about....i changed them for terminals...

[dschwartz]

OK OK OK...
I applied your comments to the design..i added the cap and reference voltage to the buffer output, so..check again:

Adding those 2 components made the board 1mm bigger..so, no biggie..but, man, it´s a tight fit!
By the way..looks like 1uF coupling caps are overkill..even with 100nF and 2.2M resistors, the corner frequency is like 1.7Hz....so i think any cap bigger than 100n that fits in the board will do ( like those little multilayer caps..)

schem:


silk screen:


PCB:
http://diynoise.googlepages.com/bypassPCB2.pdf

[R.G.]

In looking at your schemo, I would expect bad results from using an MPF102. The MPF 102 can have a vgsoff of up to -8V. SOME MPF102s might work, but I think many would not, because the source of the JFET will stabilize at some voltage higher than the gate voltage. The gate voltage is already half the power supply, leaving only 4.5V for the source to rise. The MPF102 could quite easily need that much. Try a J201 instead. It has a very , very low Vgsoff, of -01. to -0.3V. The source will then ride down nearer the middle of the power supply. You may also have to increase the value of the source resistor to ground to let the source pull it high enough. This resistance and the DC level at the source must draw less than Idss; the JFET can't supply more than Idss.

[dschwartz]

hmm.. youre right (as usual)
i simulated it with a j201 on LT spice, and a 33k Ohm looks better (nearer to 4.5V , at least)

[dschwartz]

I build one of this last night, haven´t tested yet..I used sockets for the IC´s so the total size of the board was 25mm width, 40mm heigh, and 12mm deep. pretty easy to build..

After all the trouble designing it, etching , and building, i´d say that i rather use this things on complex builds, since the board is far more complicated than a fuzz face board

I like the 24mm height because that way you can put it on any side of a 1590b box, and not make much trouble.

I´m still on the search of a decent actuator though..
These look nice..although is chromed plastic:



I bought a couple of this ones...they look sturdy, but the shaft radio is like 20mm..:

[Ben N]

Are those colored pushbuttons the ones hongkongsuperseller sells for like $0.99 (and $9.00 shipping-LOL)? Also, would it be easy to add more 4066s in order to switch multiple effects?

[dschwartz]

Are those colored pushbuttons the ones hongkongsuperseller sells for like $0.99 (and $9.00 shipping-LOL)? Also, would it be easy to add more 4066s in order to switch multiple effects?

well, i just bought them in a local store..no idea..

and yes, you could switch multiple effects with more 4066s

[brett]

Nice project, Dschwartz!  Long needed.  If this can be made low-pop, it will revolutionise stomping.
(DIYeffectsboxes.com if we lose our stomp?  )

[dschwartz]

hahaha, thanks brett..but i think this kind of project were around for a while.. i only missed a decent PCB.
now i´m biting my tongue because the board and schem look more complicated than most of the fx projects out there..but, well, it´s another option for bypassing i may use sometime (this are perfect for amp switching by the way)

[Lurco]

A little suggestion for a next version: switch-to-ground activation control. And pulldowns at the extremities.