Turning Narrow Pulse Widths into Even Square Waves

[debay]

I have a signal coming from a sensor on a rotating wheel that is at 12 VDC most of the time but will drop to 0 V for about 10 ms as the sensor is activated. Rather than these narrow negative pulses I would rather have even square waves.

I had an SN74ALS112AN dual JK negative edge triggered flip flop, but it runs on 5 V DC. Therefore I took my signal and used two 10K ohm resistors as a voltage divider to put it in the proper range. I tied the J and K inputs as well as the clears and presets the power to the chip to a 5 V supply. The chip ground was connected to the 5 V supply ground. This should've created a T type flip-flop.

Unfortunately, toggling was not consistent. I'm wondering if my pulse width is too narrow for my set up.

Would I get better results building a T type flip-flop from transistors that are designed to work with 12 volts? If so, how would I construct it? My guess is that I would want to use a design that doesn't have capacitors, so the RC time constant is not a negative factor in the design.

Thank you very much,

Dave

 

[blueroomelectronics]

A CMOS CD4027 can handle 12V.
What are you trying to do? What's it for?

 

[debay]

I am trying to capture rpm of the wheel and send the info to a PLC. If the pulse is too narrow then the PLC may or may not recognize it.

 

[KMoffett]

debay,
I think what you need is a monostable multivibrator to "stretch" the pulses to a usable length. Since this is a negative going pulse it seems just right for an LM555 timer chip.
**broken link removed**
Replace the switch with your signal. Here's a calculator to help pick component values to get the desired pulse wideth.
http://www.csgnetwork.com/ne555timer1calc.html

Ken

 

[debay]

Thanks to KMoffett and blueroomelectronics.

I liked the idea of a T-Latch to toggle states rather than just extending the state. With a CMOS CD4027 the system is working just fine and the PLC is responding to the signal.

Thanks for your help.

Dave

 

[Boncuk]

1/2 duty cycle

the easiest way would be to use a decimal counter CD4017 and have it recycle after each count. It doesn't take care of the pulse length and just counts on every rising pulse edge. That way the output is symmetrical.

 

[Roff]

the easiest way would be to use a decimal counter CD4017 and have it recycle after each count. It doesn't take care of the pulse length and just counts on every rising pulse edge. That way the output is symmetrical.

A JK FF (CD4027) or a D FF (CD4013) can do the same thing, and you get 2 per package. Can your circuit tolerate 1/2 the original frequency?
BTW, your original circuit didn't work because 74ALS requires more input current than your 10k resistors could provide.

 

[Boncuk]

A JK FF (CD4027) or a D FF (CD4013) can do the same thing, and you get 2 per package. Can your circuit tolerate 1/2 the original frequency?

So what should he plan to use the second half?

What's wrong with dividing by 1?

 

[Roff]

So what should he plan to use the second half?

Who knows? Maybe nothing, but having an unused circuit can sometimes be handy.

What's wrong with dividing by 1?

That would be great! If you know a way to get a 50% duty cycle, divide-by-one with an all-digital circuit, you should probably file for a patent before you divulge it to the public.