8 way switch drawing

[kvb]

  I'm starting to feel like this guy



I think I need some layout programming.


Any suggestions/pointers on the build to be?

I imagine that if things are OK on the breadboard, a rat's nest of wires might not be a big deal.
I don't really know yet how to simplify the wiring. I'm not looking forward to all of the stripping wires (again), soldering, etc.
But I am looking forward to what this thing is going to do.

[oskar]

Any suggestions/pointers on the build to be?

Thanks for the photo in the earlier thread...        ---   Can't you post it on the stompboxthread and say you need help debugging your fuzzface?
It looks like you want a lot of coffee to go with that build.
In this case I would substitute the switch debouncing and the logic part with a microprocessor.

skar

[kvb]

Any suggestions/pointers on the build to be?

In this case I would substitute the switch debouncing and the logic part with a microprocessor.

skar

Yes, I understand that this is not the most efficient way to do this, but I do not know how to do the uC thing.
That knowledge is possibly years away.

This switcher, a tedious as it might be, is not that far off.

As far as my thinking that this set-up might have more capability than an octal latch - I'm pretty sure that's wrong.
This (if I build it) is pretty much just a one of eight switch.

edited to say: I think I'll get an octal latch so R.G. can stop shaking his head.

[oskar]

This switcher, a tedious as it might be, is not that far off.

As far as my thinking that this set-up might have more capability than an octal latch - I'm pretty sure that's wrong.
This (if I build it) is pretty much just a one of eight switch.

edited to say: I think I'll get an octal latch so R.G. can stop shaking his head.

Not far off = best point ever. You actually have something that works...

There are many ways to achieve this. If you ever want to start programming microprocessors I recommend that you
return to this project again. It is a very good beginner project.

Latch would work to.

EDIT: Mayby this is the wrong forum for this machine... It's not DSP or microcontroller oriented and I think actually more would
be interested. There are a lot of threads on switching right?

[R.G.]

First off, let me say that I'm impressed. The sheer determination and manual skills to get that circuit working on a breadboard are great.

Let's talk about registers and latches. S-R latches are where I started with the footswitcher design. The logic for a one-of-two switcher is pretty simple. You take two inputs, A and B; you apply A to the set input of the A latch and also to the reset of the B latch. Likewise B goes to the set input of the B latch and to the reset of the A latch.

Now go to A, B and C. Three latches. The logic is that A, B and C all go to the Set inputs of their respective latches, and the OR of the other two go to the reset of each respective latch. That way, each input sets its own latch, and resets all the others. This can now be generalized to N-way; signal X goes to the set input of the X latch and to an N-1 way OR gate on the reset of all the other latches. For a one-of N switch, you need N latches, and N times an (N-1) input OR gate. For N=8, you need eight SR latches and eight seven-input OR gates. The latches and OR gates may be made up of the usual assortment of ANDs, NANDs, ORs, NORs and so on to get the right polarity of action and fewest packages.

A latch is set of flipflops which follow a data input when clocked, a set of D-flipflops. In the case of an 8-way latch, there is the latch and one clock generator. You apply 8 inputs to the latch and when the time is right, whack the clock input.

For a one-of-N switch, the time is right whenever any one of the inputs change. So we can generate the clock from the OR of all the inputs. That's what all that diode/resistor/capacitor stuff is in my drawings, an eight-way OR generated in diode logic. So for a one-of -eight with a latch, you need eight D latches and one eight-input OR gate. The rest of the R's and C's are to diddle with the clock pulse timing to generate a solid clock pulse from the data change that meets the setup, clock, and hold time requirements for the latch.

The only difference in the approaches is that the SR latch approach generates an active "Set one and only one" signal every time. The latch approach relies on there only being one input active, because all the inputs are latched in. If the latch has two inputs that are shorted, then every time you hit either one, both get clocked in. The SR is more logically robust because of the extra gating.

But they both work OK for footswitches.

[Chuck]

Dam, that breadboard is incredible.
I don't have the patience for that.  I whip out ExpressPCB and order prototype boards.
It is impressive.