Circuit selector

[Rocksteady]

Hi. I am quite new to electronics so I am not even sure whether "circuit selector" is the correct terminology. Imagine multiple circuits, let's say 10 circuits, you want to switch on one at a time (i.e. only one circuit will be on at any given time) using only two tactile buttons, "up" and "down". "Up" toggles through all the circuits and when you get up to circuit number "9", it goes back to "0" with the next press. The "down" button simply goes the other way. How to do this without using a microcontroller please? I was thinking with mosfets or a 555 timer or a combination of both BUT I've racked my brain and can't figure it out. Surely there must already exist such a circuit? Thanks!

 

[Bluejets]

Up down counter.
Discrete ic's available.
Set to count to 10 and then reset.

More information required on what you want to switch exactly to give a more detailed answer.
Alternative would be a 10 position rotary switch.

 

[crutschow]

Do you want to switch the power to the circuits, or their output signals?

 

[kellys_eye]

In the simplest possible sense use a 4017 decade counter driving 10 banks of relays who's contacts switch-over the relevant signal lines of the 10 circuits you want to control. Pulse the 4017 using a de-bounced push-button and (potentially) a 555 monostable.

Of course the 'relays' can be some other form of switch - solid-state - depends very much on what you're switching and how fast, voltage, current etc

 

[Harald Kapp]

n the simplest possible sense use a 4017 decade counter

Problem with the 4017 is that it counts only upwards. A CD4029B or MC14516B (for example) could be used to count up or down, but here an additional BCD-to-Decimal decoder is required.
Also the counters I know usually use a single clock input plus an up/down control input. This makes it a tad difficult to control up/down with two separate buttons. On will need an additional circuit that conditions the inputs from the buttons such that first the control line is set correctly and next a clock pulse is generated.

How to do this without using a microcontroller please?

A µC is probably the simplest solution.

 

[Martaine2005]

It is possible with two CD4017s and a 555. One for ‘up’ and one for ‘down’. But it would obviously be better with a micro.
Or depending on your circuits (which nobody knows yet) you could use diode steering to achieve the same result. Tell us what kind of circuits.


Martin

 

[Bluejets]

I am quite new to electronics

How to do this without using a microcontroller please

Why no uC..?
Simple enough and chances are there would already be code written out there somewhere to do what you want.
If not, learn how to do it.
Kids in primary school do it every day
Not that difficult with say Arduino Nano for example and plenty of help available on their forum.

 

[Harald Kapp]

This is how the up/down pushbutton logic could look like:

In the idle state (pushbuttons not pressed) clock is low, the control signal up/down is high (count up).
pushbutton up pressed: after debouncing by R2/C1 a clock signal is produced at the output.
pushbutton down pressed: after debouncing by R4/C2 the output of A3 goes low and the control signal is set to low = count down. After an additional delay by R5/C3 the output of A4 goes high thus crfeating a clock pulse at the output of A2.

Correction: The gates shown in the schematic should be inverters!

 

[Rocksteady]

Why no uC..?
Simple enough and chances are there would already be code written out there somewhere to do what you want.
If not, learn how to do it.
Kids in primary school do it every day
Not that difficult with say Arduino Nano for example and plenty of help available on their forum.

Thanks blue jet. I am building an ohm meter calibrator and I do not want a uC that will interfere with the signal. Thanks

 

[Rocksteady]

Up down counter.
Discrete ic's available.
Set to count to 10 and then reset.

More information required on what you want to switch exactly to give a more detailed answer.
Alternative would be a 10 position rotary switch.

Thanks, that is very helpful! Only problem is I have 28 circuits and I need to be able to switch on and off multiple circuits at a time.

 

[Rocksteady]

Do you want to switch the power to the circuits, or their output signals?

One output signal but I need to select between multiple circuits and combination of circuits that are all connected in series, kinda like a mega voltage divider. All circuits will run off the same, low voltage power supply. The main thing is that any switching must not interfere with the signal, not even half a percent. I am building a device that simply provides multiple accurate resistances between 1 ohm and 10Mohm using a total of 28 different resistors.

 

[Rocksteady]

In the simplest possible sense use a 4017 decade counter driving 10 banks of relays who's contacts switch-over the relevant signal lines of the 10 circuits you want to control. Pulse the 4017 using a de-bounced push-button and (potentially) a 555 monostable.

Of course the 'relays' can be some other form of switch - solid-state - depends very much on what you're switching and how fast, voltage, current etc

Thank you kellys_eye, that is extremely helpful. Actually the idea is to build my own decade switch. Sorry, I should have been more clear, I need to control 28 circuits in total. I am building a 1ohm to 10 M ohm calibrator BUT I also need a display that will output the selected resistance. I need to switch between 28 different resistors in series OR combinations of the resistors (like a mega voltage divider) and without disturbing the signal so I get a precise value on the output every time. It has since occurred to me to use LND mosfets for switching except I am not 100% sure one can get any that won't change the resistance on the output, not even a bit. I also need everything to fit into a small, handheld box so I am not sure whether using 28 separate relays will work either.

 

[Rocksteady]

In the simplest possible sense use a 4017 decade counter driving 10 banks of relays who's contacts switch-over the relevant signal lines of the 10 circuits you want to control. Pulse the 4017 using a de-bounced push-button and (potentially) a 555 monostable.

Of course the 'relays' can be some other form of switch - solid-state - depends very much on what you're switching and how fast, voltage, current etc

Thank you kellys_eye, that is extremely helpful. Actually the idea is to build my own decade switch. Sorry, I should have been more clear, I need to control 28 circuits in total. I am building a 1ohm to 10 M ohm calibrator BUT I also need a display that will output the selected resistance. I need to switch between 28 different resistors in series OR combinations of the resistors (like a mega voltage divider) and without disturbing the signal so I get a precise value on the output every time. It has since occurred to me to use LND mosfets for switching except I am not 100% sure one can get any that won't change the resistance on the output, not even a bit. I also need everything to fit into a small, handheld box so I am not sure whether using 28 separate relays will work either.

 

[Rocksteady]

This is how the up/down pushbutton logic could look like:
View attachment 56032
In the idle state (pushbuttons not pressed) clock is low, the control signal up/down is high (count up).
pushbutton up pressed: after debouncing by R2/C1 a clock signal is produced at the output.
pushbutton down pressed: after debouncing by R4/C2 the output of A3 goes low and the control signal is set to low = count down. After an additional delay by R5/C3 the output of A4 goes high thus crfeating a clock pulse at the output of A2.

Correction: The gates shown in the schematic should be inverters!

Thank you Harold, that is a really excellent diagram. The only problem is I can not have ANY interference on the output signal. I need a precision of less than 0.1 % on the resistance.

 

[Rocksteady]

Do you want to switch the power to the circuits, or their output signals

Hi crutschow, thank you very much indeed for your reply! I am building an ohm meter calibrator so power going through the resistors will all come from the power of the ohm meter that is being calibrated. I am going to the trouble of using high precision metal oxide film resistors SO, to further ensure the highest possible precision, any type of switching I use may not interfere with the resistance on the output. I will have 28 resistors (7 decades, 4 per decade) connected in series, a bit like a voltage divider, and in combination they must, very precisely, provide between 1 and 10^7 ohms resistance (actually I think it will be between 1 and 1.9996x10^8 ohms but who is counting?). I also need a 4-digit display that will output the selected resistance and will probably have it's own small button battery). I would prefer NOT to use relays because the unit must be as compact and lightweight as possible SO I have been toying with the idea of using LND (normally closed) mosfets instead of relays EXCEPT I am not sure by how much that will interfere with the selected resistance, especially when several are switched on in series at the same time.

 

[Bluejets]

high precision metal oxide film resistors

So, they are still around plus or minus 1%....unless you are making a special order request.
Most are satisfied with a rotary switch and metal oxide standard resistor.
Any addition circuit will only add tolerance variations I'm sure.
Apart from that, any precision ohm meter would be a constant current injection type, again with certain tolerances one cannot avoid.
Any certified test on instruments with high tolerance requirements certainly would not be recognised if built by someone"new to electronics" as you profess to be.

 

[Rocksteady]

So, they are still around plus or minus 1%....unless you are making a special order request.
Most are satisfied with a rotary switch and metal oxide standard resistor.
Any addition circuit will only add tolerance variations I'm sure.
Apart from that, any precision ohm meter would be a constant current injection type, again with certain tolerances one cannot avoid.
Any certified test on instruments with high tolerance requirements certainly would not be recognised if built by someone"new to electronics" as you profess to be.

Thank you Bluejets. I would be building one for my own bench to "calibrate" my own non-professional test equipment and also purely as a learning exercise so it would not be a "certified" test.

 

[Bluejets]

Regardless, the same applies.

 

[danadak]

One possibility to get rid of all the switching and severe MOSFET requirements :

Active Resistor Circuits

Active resistor circuits represent complex structures, implemented using exclusively MOS transistors that are functionally equivalent with a classical resistor, being able to simulate, between two input pins, a linear current–voltage characteristic. The goal of...

link.springer.com

A compendium of articles on active resistors (see references at page bottom) :

https://link.springer.com/article/10.1007/s10470-009-9351-5

One example using OTA :

https://ijecce.org/administrator/components/com_jresearch/files/publications/IJECCE-1382_Final%20N.pdf

Regards, Dana.

 

[Rocksteady]

Regardless, the same applies.

Well, not exactly.