Pulse detector

Not a typo. Those are PNP transistors and they need negative collector supplies.

It is a bit confusing--nowadays one generally tries to draw schematics with current flowing generally downward and signals generally to the right.

Cheers

Phil Hobbs

Edward-

The minus voltage appears to be correct for PNP transistors.

Minus 12 Volts might work for a lower amplitude pulse.

Fred

OK, i am used to drawing positive on top.

What if i replace them with NPN, fliping the diodes, lower the two 18K to 1K and tie to ground?

I think the first two Qs are just flip flop and the third integrator. I just need the single rail positive version.

What, exactly, is "this"? Do you really want to hold the peak DC voltage of an incoming signal for a few milliseconds, or until you reset, or do you want to measure that peak height?

Waveforms don't have 'a' voltage, and 'detector' is not very descriptive.

I want to measure the peak DC voltage of a 10% duty cycle signal.

How about diode+capacitor+discharge pushbutton?

How fast?

Is it repetitive?

The 'simple circuit diagram' smells very old.

Yes.

just need the single rail positive version.

Enter it into LTspice and try it. First I'd enter the circuit as is to mak e sure it works as expected. Then modify it as you are looking to do with the positive supply and eliminating the other supply.

What speed do you need from this? How high, often and wide are your pulses ? This is the sort of thing I would likely use an op amp for. You don't h ave to worry about biasing and other messy details. An op amp, a diode and a few gain setting resistors do the job.

At one kHz definitely use an op amp.

You could Google "op amp peak detector circuit" or here are a couple of examples.

or here's a 35 minute educational video by our favorite public speaker.

The original circuit (flip it horizontally makes more sense) might be good for negative signal. I am measuring positive signal; so, i want to change to NPN.

OK, that might work.

Yeah, that's a peak voltage detector.

Next for when to capture that voltage (DutyCycle = 10%), maybe an R-C LPF and a window comparator circuit... Use two copmareators to trigger the peak V capture (whatever you need to do with it) when the duty cycle is 10% of the full range voltage ?

Yes, that's what i need. Should have search on "peak detector" instead of "pulse detector". It's just a simple sample and hold of the input, then using several comparators to trigger on the voltage level.

A 10% duty cycle, 1 kHz pulse can be measured quite well using a single chip, as small as 8 pins.

The simplest approach for that (used with proportional counter pulses) is a peak detect/hold/convert-to-time-delay-and-count circuit called a Wilkinson A to D converter.

Basic circuit is an op amp follower on the input voltage, feeding a diode and capacitor (output through diode to capacitor, feedback from the capacitor). When a peak passes, the diode reverse-biases, and the op amp output goes to the negative rail. Then you disconnect the op amp (it's done its job) and either directly voltage-measure the capacitor, OR connect an accurate current-sink to the capacitor and count clock pulses as it discharges to zero volts.

rote:

is

and capacitor

a peak

ive rail.

Or you use a $0.60 MCU to digitize the 100 uS wide pulses at 100 kSPS, let software find the valid pulse measurements and you are done.

Why is everyone making this so hard? He didn't initially say he was measur ing the pulses, but still, not sure what that implies. Does he want data t o be sent to another computer? Does this need to drive a display? Is he l ooking for real time updates of each pulse? An average of some sort? Is t he pulse height varying? Sounds like he wants window comparisons to produc e a decision of some sort? All of the above is very easy to do in the same $0.60 MCU that is taking the measurement. It can send an output via RS-23

This is a classic case of, "No, tell me the problem you are really trying t o solve".