74hc165 to 74hc595 communication without microcontroller

carl vianelli

Hello anyone, I am trying to to connect a 74hc165 parallel to serial shift register, to a 74hc595 serial to parallel shift register.
i would like to use a 555 timer for the clock. i am trying to send the position of switches or push buttons at least 7 minimum,
through a least amount of wires. if anyone has any other ideas how to do this i need help. It dosen't have to be shift registers, But I
thought the registers would use the least amount of ICs.
please help, thanks..

Pliers

Sounds like a fun project.
What is the length of you communication cable?
Is it "The least amount of wires" or "least amount of ICs"? Or is it both?

If you are concerned mostly about cost, what is the budget for this project?

A wireless system uses the least amount or wire.
Sorry for all the questions, I'm trying to understand the project better.

The speed of the communications and the distance of travel are big factors in any design.

Good luck!

Mike Green

This is actually very complex since you have to generate not only a shift clock, but a load clock as well. You have to produce 8 shift clocks, then you have to produce a load clock. This will require a counter and some control logic at the very least. You will need 3 lines to the 74HC165 and 3 lines to the 74HC595, a data line, a shift clock line, and a load clock line. If you're concerned about the number of ICs and/or the overall cost, you'd be much better off with a microcontroller. A PIC or other microcontroller in a 14 pin (PIC16F688) or 18 pin (PIC16F628A) package would provide 12 or 16 I/O pins that could be used for inputs or outputs and you could use a single line (plus power and ground) connecting one to the other (with standard asynchronous serial I/O used for the data). A Basic Stamp or Propeller could also do the job. The Stamps are really easy to use but relatively expensive. The Propeller is cheaper and has more I/O pins available (28).

Pliers

Here you go.
No micro controller.
One wire transmission one wire common.
One 555 as a clock.
Eight switches and eight LEDs.

Three ICs and three op amps.

It dosen't have to be shift registers, But I
thought the registers would use the least amount of ICs.

How it works: each switch has a different value resistor. The summation amplifier makes a cheap digital to analog converter.
The 555 runs the counter through all 255 values, and then the data is also summed up.
When both values are equal the comparator goes low causing the FLipFlops to latch the current data.

Obviously this drawing is not complete. The summation amps are drawn incorrectly and the comparator should be a window comparator.

Zoot

Pliers, not sure why you are using weighted outputs with a op amp. There are scads of "divide-by" chips out there that can divide a pulse train by 2, by 4, etc. Generally they are binary counters (counting input pulses in binary) that output the bits of the count on pins. This gives you divide by powers of 2 outputs. So if the 555 generates clock pulses, the binary counter's bit4 pin would output a pulse every 8 clocks. This can be used for load/latch. If the load clock needs to trail or lead the main clock, an inverter can do that. This is how it was done back before micros cost $1

Pliers

For Zoot.
I was trying to satisfy the low parts count requirement.
The original poster was vague on what they were doing. So I threw out the first thing I thought of.
Just having fun.... There are lots of ways of getting data from push buttons.

carl vianelli

pliers, thanks for the help I didn't think of going this way. all I had in mind was " pulses,pulses,pulses" or binary 000-111,this would give me 7 switches with 3 wires,1 common.
my cable is only 8', my budget is very little, speed doesn't matter,well not too slow but it's only human controlled.
this looks easy to expand. i don't know what a window comparator is I'll google it. I have 339 quad and 741 comps. I'll try breadboarding it.
using flip flops sounds good "push-on, push-off" this is better..earlier I have been looking at PIC starter kits, that mike green suggested using a PIC,I never used a PIC before,but the price
of PIC chips is good.The first starting part is what is going to cost...It won't be soon. I have a BS2 so I have a little headstart on microcontrollers.I'm going to look into the PICs in the future. I'm not sure what "zoot" was saying but I'm going to try to figure it out, I'll read his post again.
I'll look at your circuit and start to get something working.
thank you very much!!!.....................Did you see that space jump, long drop

carl vianelli

mike, thanks for your suggestion. I'm looking into PIC microcontrollers, they are much less expensive than a BS2. I have to get a starter kit because I never used one before.
pliers drawn up a circuit I'll try..
thank you for your seggestion...now I want a PIC starter kit,HA HA

carl vianelli

zoot, you got me thinking :thumb: I'll get some paper and start drawing lines.
thanks......

Pliers

Carl Vianelli.
I would not take my circuit to seriously.
My goal was to use the least amount of wire and ICs, and have some fun.
I think your goal is more about learning digital control.
You will find a lot of smart and helpful people here.

I hope you let us know what you learn, and post some circuit drawings.
Best regards.

Zoot

What's your final goal here? Will the switch positions/states be "read" by a micro? Or just circuitry to turn indicator lights on/off or something?

Pliers

Here is an all digital idea.
It is missing the pull up resistors and the buttons should be tied high, not low as shown.
Also, I'm not sure if the all the logic levels match. As in active high or active low or edge triggered inputs,

The CD4017 is a decade counter. One output at a time goes high with each clock pulse.
The 4049 and capacitor is used to delay the clock pulse. This gives the data time to settle.