CMOS switching brain teaser

[Scruffie]

I like 2 stage phasers, I like 6 stage phasers, but I also like 4 stage phasers, I don't like having 3 phaser pedals on one pedal board... and rotaries are big and a pain to PCB mount.

With a DPDT on/on toggle you can switch out 2 or 4 stages of a phaser much like true bypassing them, to get the desired amount of stages in circuit but, with a DPDT On/Off/On and a single CD4053 (3 SPDTs) is there a way of switching between 2, 4 & 6 stages of phasing with a single toggle?

[R.G.]

There is a way to do almost any switching if you have a means to tell it what to do. You have three switches, and a three position switch. Each SPDT can be used to gate one of the three phase line positions into the output mixer. One position of the switch is "off" so one of the spdt switches has to be passing signal when it's in this position, off when it's in either of the others. No biggie here.

Both of the other switch positions have to turn off two switches, the center off one and the non-selected one. So you need two diodes per switch to do the controlling with MML. (Mickey Mouse Logic)

Next.

[Scruffie]

I thought you were taking the 'mickey' with the mickey mouse logic part but see it's a real thing, I don't quite follow but i'll start reading up on MML and see if it becomes clearer.

Would an On/On/On make it any simpler?

Switching is a big achilles heel for me, I can follow it on paper but in my head it takes a long time for the gears to get in to place.

[armdnrdy]

You can switch between different phase stages with a rotary switch.

Here is a thread with the modified Phase 100 schematic including phase stage switching.

http://www.diystompboxes.com/smfforum/index.php?topic=106895.0

[R.G.]

OK, enough with the veiled hints.

The trick is that each of the SPDT switches in the 4053 has a control pin that's either up or down. The common pins, let's call X, Y, and Z. So for section X, when the control pin for X is high, X is "on" to X1. When its control pin is low, near zero, X is connected to X0. Same for Y and Z.

Let's say you hooked this up with the three signals into your mixer from the X, Y, and Z pins. You can connect either X1 or X0 to the phase line anywhere you like, likewise Y0, Y1, Z0, Z1. So all three taps could possibly be selected by a high or a low on the control pin by whether you connected it to the n0 or n1 pin of the n switch in the CMOS.

If you take your three control pins and pull them all high with resistors, then use your toggle switch common pin grounded, you could pull down, for instance, X control pin with one side of the toggle switch, and Z control pin with the other side of the control switch. If you connected a tap to X0 and Z0, then when the toggle was on the "X" side, X would be enabled to the mixer, and when the toggle was to the "Z" side, Z would pass through to the mixer. In the middle neither X nor Z would go through.

But what about Y? If you make Y enabled to the mixer when Y control is 1, and use two diodes on each of the X and Y side of the toggle, so the X side pulls down the X control through one diode and the Y through the other diode; then the Z size pulls down Z to turn it on and Y to turn it off, in the middle neither X nor Z is low, so they're disabled, but Y control is high so Y goes through.

Pull up on each control pin. X and Z enabled with X control = 0, Z control = 0, and Y enabled with Y control = high. Diodes to pull X and Y control low on one side, diodes to pull Z and Y low on the other side.

[Scruffie]

You can switch between different phase stages with a rotary switch.

Here is a thread with the modified Phase 100 schematic including phase stage switching.

http://www.diystompboxes.com/smfforum/index.php?topic=106895.0

End of the first sentence i'd much rather fit a toggle and bit of CMOS circuit in to a pedal than a rotary

OK, enough with the veiled hints.

The trick is that each of the SPDT switches in the 4053 has a control pin that's either up or down. The common pins, let's call X, Y, and Z. So for section X, when the control pin for X is high, X is "on" to X1. When its control pin is low, near zero, X is connected to X0. Same for Y and Z.

Let's say you hooked this up with the three signals into your mixer from the X, Y, and Z pins. You can connect either X1 or X0 to the phase line anywhere you like, likewise Y0, Y1, Z0, Z1. So all three taps could possibly be selected by a high or a low on the control pin by whether you connected it to the n0 or n1 pin of the n switch in the CMOS.

If you take your three control pins and pull them all high with resistors, then use your toggle switch common pin grounded, you could pull down, for instance, X control pin with one side of the toggle switch, and Z control pin with the other side of the control switch. If you connected a tap to X0 and Z0, then when the toggle was on the "X" side, X would be enabled to the mixer, and when the toggle was to the "Z" side, Z would pass through to the mixer. In the middle neither X nor Z would go through.

But what about Y? If you make Y enabled to the mixer when Y control is 1, and use two diodes on each of the X and Y side of the toggle, so the X side pulls down the X control through one diode and the Y through the other diode; then the Z size pulls down Z to turn it on and Y to turn it off, in the middle neither X nor Z is low, so they're disabled, but Y control is high so Y goes through.

Pull up on each control pin. X and Z enabled with X control = 0, Z control = 0, and Y enabled with Y control = high. Diodes to pull X and Y control low on one side, diodes to pull Z and Y low on the other side.

Once again (and I very much doubt for the last time) thank you! I should be able to draw that out so I can see what's happening now and then work my way backwards to understand it

[armdnrdy]

"and rotaries are big and a pain to PCB mount."

Was that part there originally? I didn't even notice it. 

I guess I should actually read a post before I reply? 

[Fender3D]

End of the first sentence i'd much rather fit a toggle and bit of CMOS circuit in to a pedal than a rotary

Then have a look at my suggestion on the same thread http://www.aronnelson.com/DIYFiles/up/Phaser_output_mixer.pdf

confirmed working with phasers and switching of course...

[Scruffie]

"and rotaries are big and a pain to PCB mount."

Was that part there originally? I didn't even notice it.  :icon_wink

It wasn't, it's all just part of a larger scheme to... slowly drive you insane

Easy to miss 

End of the first sentence i'd much rather fit a toggle and bit of CMOS circuit in to a pedal than a rotary

Then have a look at my suggestion on the same thread http://www.aronnelson.com/DIYFiles/up/Phaser_output_mixer.pdf

confirmed working with phasers and switching of course...

Hmm, interesting, would certainly be a space saver, guessing the CMOS route is a little more solid but depending on what phaser I go for that might be just the ticket.

[PRR]

If you have enough impedance-difference to get tricky, this will do 1-of-3 (2/4/6) selection with just a double-pole on/off/on switch:



I'm not familiar enough with all the many phaser techniques to know if this can work without additional buffering. At an extreme, if the unshift/shift mixer is virtual-earth, and the shifters are opamps, it probably will work with little compromise.

[Scruffie]

If you have enough impedance-difference to get tricky, this will do 1-of-3 (2/4/6) selection with just a double-pole on/off/on switch:



I'm not familiar enough with all the many phaser techniques to know if this can work without additional buffering. At an extreme, if the unshift/shift mixer is virtual-earth, and the shifters are opamps, it probably will work with little compromise.

Elegant! I'll give this a go first, likely it will be based on the MXR phase 100 topology (schematic attached) discussed so will be using opamps for the phase shift.

Just to be sure, the resistors in the schematic I take it form the 'wet' signal mixing resistor, or are they there purely to divide the signal? So if I were to require a buffer I can just stick an opamp buffer at the end of the resistor network or would I need a buffer per stage group?

[PRR]

You said 2-4-6 stage but show a 8 stage.

I started marking-up the plan and just got tangled.

I *really* think you should get a 1P4T rotary. Five wires and a hole has to be easier than CMOS contraptions.

[R.G.]

Five wires and a hole has to be easier than CMOS contraptions.

Oh, ye of little faith.   

[Scruffie]

You said 2-4-6 stage but show a 8 stage.

I started marking-up the plan and just got tangled.

I *really* think you should get a 1P4T rotary. Five wires and a hole has to be easier than CMOS contraptions.

No, no, sorry for the confusion, I showed it as an example, if I use that anyway, i'd tap off from the 8th , 6th and 4th anyway as I find fixed stages to not give quite the... depth that swept do.

I'm actually thinking now rather than optical phase shift elements more along the lines of using the Bad Stone phaser shown here (sorry for the large image);


So would what you drew not work in this application?

If I must use a rotary I will, but CMOS seemed like it could be the solution here, in fact, i'd rather CMOS over the rotary, replacing a board mounted toggle is a cake walk next to a rotary and i'm not one for off board wiring.

[StephenGiles]

Another way perhaps - enable switching with foot connector, hit string softly - 2 stages, hit string harder - 4 stages, hit string very hard - 6 stages. Peak extraction circuitry fed to window comparator network sending appropriate hi or low voltages to switch in number of stages required.