I need to build a controller 8 bit that can activate at least 50 switches
For example :
The signal 00000000 gives 5 volt in terminal number 1
The signal 00000001 gives 5 volt in terminal number 2
The signal 00000002 gives 5 volt in terminal number 3
Like this until at least 50​

 

Use a series of latching shift registers. Each will probably be good for 8 bits, so you will need 7 of them in series. Your uc will need only 3 bits.

 

The shift registers are a good idea.

It would help if you could tell us what the 50 +5v outputs are supposed to do.
For example if there is any particular current requirement, like driving LEDs?
Or is it just a logical high 5v level?

 

can i use for this application 74138 decoder?

 

We have no idea of the switches. Are they relays that are electrically activated? Do they get turned on one at a time, or controlled in groups? If individually controlled, do several or all get turned on at the same time?

 

Might look at a 74151. 3 inputs + enable and selects among 8 outputs.

 

Might look at a 74151. 3 inputs + enable and selects among 8 outputs.

A 74151 is a data selector/mux. It selects from 1 of 8 inputs to provide an output. A demux/decoder is what would be needed if taking that approach. The 74138 is a decoder/demux; it could be made to work with some extra gates/inverters.

Going along with Russ Hensel's idea, a CMOS 4021 8-bit latching shift register would be very easy to implement; part numbers would be HEF4021, CD4021B, or MC14021 for most of the major manufacturers.

However, we don't know if the OP has a requirement for a direct selection of an output, or if it can be clocked in over time.

Is this a homework assignment for a class or course?

 

Beenthere, I don't understand all the differences, can you explain​

​

My case is :​

​

Solenoids​

​

turned on one at a time​

​

Individually controlled​

​

 

OK, it helps a lot to know that your load is a large number of solenoids.

Right now, I will assume that your solenoid coils are rated for 5V DC. If that is not correct, please tell us what their rated voltage is.

We also need to know what is their resistance in Ohms, or their wattage, or their current requirement.

Also, are you going to be using a microcontroller (uC) to select these solenoids?
Or will the input be a series of switches?
Or will it be some other kind of control?

 

Beside the steering logic, this could control off a pair of thumbwheel switches - a bit pricey, but they look so good.

 

Beside the steering logic, this could control off a pair of thumbwheel switches - a bit pricey, but they look so good.

Yep, thumbwheel switches are cool - but that would preclude the use of shift registers, unless there were a uC in between.

 

Yes, but the OP has said only one will be energized at a time.

Maybe we will get a bit more information about the solenoids.

 

Had that backwards. Mux/demux is confusing, maybe I'll get it now. I look forward to how this circuit will turn out. I have no idea how a shift register would fit into this.

 

OK, a 4094 latching serial-to-parallel shift register circuit is pretty easy to explain. It'll help if you have a datasheet handy. HEF4094B is a good one to download.

Earlier, I mentioned a 4021; unfortunately that's a parallel to serial; we need to go serial to parallel.

These shift registers basically require three inputs:
1) A clock pulse
2) Data input
3) A Strobe to latch the serial data to the Parallel output.
They also require the OE (Output Enable) pin to be held high, as these have tri-state outputs.

All of the shift registers receive the same clock pulse and Strobe input.
The 1st shift register gets it's Data input directly from a uC output.
The QS1 or QS2 output (depending on how the clock rise time looks) of the 1st shift register goes to the Data input of the 2nd shift register. The remaining shift registers are cascaded in a similar manner.

There are 8 output bits per shift register; to have 50 controllable outputs would require a minimum of 7 shift registers. 56 solenoids could then be individually addressed and controlled.

 

OP still has not posted enough information. He said turn them on one at a time. But will more than 1 need to be on at a given time?

Also, what is the distance between the solenoids and the controlling circuit? If more than a few inches, serial to parallel data latches won't work.

The 4094 can only sink about 1.5 ma on each output. The 4094 would have to drive something like a ULN2803 darlington transistor array. The ULN2803 can handle up to 50V on each output.

If the solenoids are a substantial distance from the controller, the only way to do this would be to use some type of addressable switch with an RS-232 or maybe RS422. With RS422 5 pins of the microcontroller could drive 5 transmitters with 10 receivers assigned to each transmitter.

Another good choice if the solenoids are located some distance from the controller would be the Maxim DS2413 1-wire parts.

 

OP still has not posted enough information. He said turn them on one at a time. But will more than 1 need to be on at a given time?

We are painfully aware of the dearth of information provided.

Serial-to-parallel latching shift registers would provide for the capability of multiple solenoids on at a time. Decoders/demultiplexers would not.

Also, what is the distance between the solenoids and the controlling circuit? If more than a few inches, serial to parallel data latches won't work.

The shift registers could be placed in one small area on a board; wires/cabling routed to individual solenoids.

The 4094 can only sink about 1.5 ma on each output. The 4094 would have to drive something like a ULN2803 darlington transistor array. The ULN2803 can handle up to 50V on each output.

Yes, that's a given. CMOS would be better off with ULN2804's and a Vdd of 10v-16v. However, at TTL levels, they could directly drive logic-level MOSFETs such as IRLD014 or even abuse 2N7000 MOSFETs.

There really isn't much point in continuing this if we don't get more info from our original poster.