Hello all and apologies for what may be a very simple question.

I have a micro controller that produces a very short (say 2 or 3 millisecond) 5 volt pulse every now and then. The signal occurs only upon the occurance of certain events by design, so is not cyclical or otherwise predictable.

I would like to capture and log when these pulses occur, but my data logger only samples 50 times per second and therefore misses some of these short pulses.

My question is-- how can I extend the length of the pulse to about 1/50th of a second to ensure my data logger captures each pulse? Each pulse occurs infrequently, so I'm not worried about missing one during the time the pulse is extended. Also, the voltage of the pulse can decay over the extended pulse time, as long as it remains above 2 volts for at least 1/50th of a second.

All help appreciated!

 

You need a 555 multistable multivibrator circuit, but the TOS here forbid automotive topics. This thread will be closed as soon as a mod notices it.

 

Thanks for the heads up. The signal is coming from a micro controller (8052 based if I recall). I edited accordingly as the question pertains to a non
-automotive project.

 

To clarify wayneh's post, you need a 555 mono-stable (one-shot) circuit with about a 30ms period.

 

You don't need to lengthen the pulse. What you need is for your data logger to be responsive to an edge-triggered interrupt.

 

You don't need to lengthen the pulse. What you need is for your data logger to responsive to an edge-triggered interrupt.

Well that would be nice but I am using a commercial logger that doesn't work that way. I'll read up on using a 555 timer (i should have thought of that in the first place) and revert as I come across specific issues with the circuit.

 

If the input impedance of your data logger is high (>1megOhm), then a simple passive pulse stretcher like this might work:

 

Well that would be nice but I am using a commercial logger that doesn't work that way. I'll read up on using a 555 timer (i should have thought of that in the first place) and revert as I come across specific issues with the circuit.

Here's a circuit that I apparently built that I think performs the function that you want. The reason I say "apparently" and "I think" is that I can't find the actual schematic and my notes about it. All I have left is what you see, but maybe it will be some help to you. Of course, you will have to adjust R1 and/or C1 to get the pulse length that you want. With the values shown, the pulse length should be about .385 seconds; for 1/50 of a second, try changing R1 to 1.8k.

 

Here's a circuit that I apparently built that performs the function that you want. ...

Be advised that a 5V pulse would not trigger the posted 555 circuit. It might work if the supply voltage of the 555 was reduced to ~5V.

The posted circuit is similar to Fig 1. in the applications section of the LM555 data sheet (LM555.pdf).

 

Thanks all for the input. Mike, would the passive circuit you posted follow the basic t=r*c formula? Does the diode simply prevent the cap from draining in both directions? I assume the logger is high impedance -- if not then would it simply throw off the discharge time? How much current is draw charging the capacitor without restriction like that? The circuit creating the pulse presumably cannot supply much current. Sorry for all the questions -- I'm reading up but and starting with very little base knowledge.

 

The discharge time will be proportional to (R1//Rin)*C1, where Rin is the data logger.

The charge time might be slowed depending on how whimpy the pull-up on the pulse source is. Try it and see. You might be able to trigger the data logger even if the stretched pulse height is only a volt or so.

If the pulse source is very whimpy, you could try this:

 

Just a quick note to say thanks. The last circuit (with the transistor) works like a champ. I'm in the process of fine-tuning the cap/resistor values for most consistent logging results, and may add a few unrelated bells and whistles, but this basic approach works great. Thanks again!