555 frequency divider to create low frequency flashes at fraction of input rate.

Thread Starter

Walt Fricke

Joined Dec 21, 2013
6
I am making a tester for automobile diagnostic ports. It checks for ground and 12V (easy, green LED and resistor for correct hookup, red if + and - are reversed). But also for 12V square waves which go to the tachometer (systems before around 2002 had those, now it is all coded packets). Those are going to be, in a car at idle, in the 33-67 Hz range. Any light would flash to fast, and I don't trust concluding that an LED across the tach to ground being on means it is being pulsed.

Aha, just divide these frequencies down quite a bit, so the light flashes about every second or half second or so. How about using 12V power. a 555 in monostable mode as a frequency divider. I have a handful in a bin along with Rs and Cs and LEDs of the sort which would be used. So I bread boarded a standard circuit: the square wave into pin 2 (I've played with a pull up resistor of 100K to +, and a pull down to ground, don't think I should need either). The output from pin 3 to an LED and a resistor to ground. Fiddling with values, 100K resistor from + to pins 7 and 6, with a lowish microF (at first, 2.2, as it was in the bag on my desk) from there to ground. Pin 5 with 0.01 to ground. Pins 8 and 4 to +. Pin 1 to ground. I had to read up to remember this stuff.

But it didn't flash. It would if I ran my square wave generator at 1 Hz - gives a nice 1 Hz flash (which is probably what I want anyway, rather than something shorter and challenging the persistence of vision and whatever linger time - maybe none? - in an LED).

But when I turned up the frequency up beyond 3 Hz (flashes faster at 2, and faster yet at 3), it gives the appearance of being on all the time.

Upping the capacitor to 22 microF has it flashing with 5 Hz input at about once per 3 seconds for five flashes, and then a pause while it starts in again

With no capacitor, it flashes at the input speed, becoming a blur at about 30 Hz.

Which has me wondering if this is happening: The falling edge of the input turns the LED on for a period defined by the R/C. But when it has ignored (if it has) some pulses, and is reset and ready to flash again when triggered, the trigger comes so close after the change of state that it appears continuous? Maybe a scope or frequency meter could measure this, but it is too fast for the eyeball?

I know I could use a CMOS divide by 12 chip, or a variety of other chips, flip flops and so on) to put faster pulses in one end, and get slower ones out the other. But the 8 pin DIP 555 is just the size I want for something perhaps housed in the back end of an OBDII connector. No SMDs, but small.

So can I get the 555 to do what I want? Or do I need to try something else. If F is between 25 and 100 Hz, turn on the flasher LED? This is stretching my cook bookery.

Walt Fricke
 

GopherT

Joined Nov 23, 2012
8,009
Ok. That was a lot of words of what you did, but, what is the end goal?
I understand a square wave of engine speed but what good does it do? You already know the engine is running. What do you plan to do next? What instrumentation do you have?
 

crutschow

Joined Mar 14, 2008
34,280
Here is an LTspice simulation of a 555 circuit that outputs a 0.2s pulse every second with an input of 30Hz to 70Hz (or more).
D1 carries a pulse from the output to the input voltage divider to raise the trigger voltage and prevent a retrigger until about 0.8s after the end of the pulse.

555 one-shot.PNG
 

Attachments

wayneh

Joined Sep 9, 2010
17,496
You could also use a counter such as 4017 to divide the signal down by 10. Or, you could just use the signal as audio input and listen to it. You can't see 60Hz but you can sure hear it.
 

Thread Starter

Walt Fricke

Joined Dec 21, 2013
6
Thanks for the suggestions.

I apologize for using words instead of a diagram. I don't know how to whip up a nice diagram, so would have to draw one and scan it. What can I say, I'm a lawyer and we use words. And I figured that many denizens here have a full understanding of the 555, and not the kind of partial one I have. I'll try harder with the next request for help I may have.

As noted, it works off of the 12V which is required to be present. Using 5V circuitry would require more components, which I'd like to avoid if possible.

This circuit is used at automobile race tracks as part of checking for rules compliance, including RPM limits and gear ratios. They are done on track with a compliance data system plugged in (high rpms), and in impound with the rear of the car jacked up (low rpms). All depend on the signal being present at the data connector. This device just checks for a suitable signalbefore hooking up a different test device. Time is wasted if the real test device isn't getting a signal, typically because a wire to the data port has been cut to attach the racer's own data system, or because street car data readers aren't always useful for the owner. And someone can tote this around the paddock to see which cars can and can't be tested with the other gizmos. It is used on 4, 6, and 8 cylinder engines, with idle speeds in the 800-1,000 RPM range. Real world application.

So this circuit just plugs in and indicates that tach signals are present at idle (plus power/ground). If so, odds are good things will work at high RPMs, so we can go ahead and attach the real testers. If not, naughty naughty.

I'll try crutschow's trigger delay circuit. I'd never in a thousand years of monkey typing have figured out something like that on my own, although it confirms my thought of a 555 as a likely approach which could be made to work. Many thanks.

Walt
 

wayneh

Joined Sep 9, 2010
17,496
Note that this will give the same output flash rate for any input rate within range. A divide-by-X arrangement with a counter (or two in series) would flash at a rate in proportion to the input rate. That seems useful to me, but it's your call.
 

GopherT

Joined Nov 23, 2012
8,009
@Walt Fricke

I would use a 555 timer wired as a flip-flop. Connect output to LED. At 1200 RPM, that is 20Hz. The flip flop cuts this in half. You will have 10 flashes per second. It seems fast but you can easily see the flash and it will appear to be full on as the car is rev'ed. Keep the current to the LED in the 1 to 5 mA range so you don't saturate your eyes and stop seeing the flash.

For a design, just google, 555 as flip-flop.

If 10 Hz seems too fast, get a 556 timer (essentially 2 555 timers on one chip). Then wire them both as flip-flops and use the output of the first to trigger the second. you will have a 5Hz flash. Once reved to 5000 rpm, the flash will go to 20 hz or so.

Flash rate will be slower for those cars idling at 800 rpm.
 
Top