Author Topic: Pulse Peak Detector  (Read 1066 times)

0 Members and 1 Guest are viewing this topic.

Offline jayeshTopic starter

  • Newbie
  • Posts: 1
  • Country: in
Pulse Peak Detector
« on: February 22, 2018, 02:54:03 pm »
Hi guys,
can anybody suggest a pulse peak detector circuit for pulse widths in nano seconds.
thanks.
 

Offline danadak

  • Super Contributor
  • ***
  • Posts: 1875
  • Country: us
  • Reactor Operator SSN-583, Retired EE
Love Cypress PSOC, ATTiny, Bit Slice, OpAmps, Oscilloscopes, and Analog Gurus like Pease, Miller, Widlar, Dobkin, obsessed with being an engineer
 

Online BrianHG

  • Super Contributor
  • ***
  • Posts: 7725
  • Country: ca
Re: Pulse Peak Detector
« Reply #2 on: February 22, 2018, 03:15:01 pm »
You can use an MMBTH10 transistor, common emitter configuration, with your source signal fed to the base via a series input cap and a weak pull up resistor right on the base to VCC.  On the collector, you will also have a pull-up to digital VCC and optional small cap in the pf range in parallel to GND.  This simple switch amp will extract the peaks of your source, creating a digital IO so long as the source signal is 200mv or stronger.

You will need to tune the input cap and pull-up resistor to tune the circuit's peak selectivity.  The pullup resistor here will be around 100k and you may also want a smaller value pull-down resistor as well.  Note that with this circuit, with no input, the output will always say that there is a continuous pulse unless you tune this pull-down resistor to keep the transistor off by default, but you loose some sensitivity.  A stronger source signal above 200mv would mean no problem here.

The output pull-up resistor should be around 2k to 10k depending on how wide/fat you want the output captured pulse to be.

This amp is and active low output.  When there is a peak, the output will be low.


2SC5066-O,LF may be better than MMBTH10 for detecting pulses at 1ns or less as it is a 7GHz transistor, but still in the 25cent price range.

Very strong pulses will be detected immediately, as the signal rises past the previous peak.  Detecting exclusively after the peak has been reached, especially sub ns sized pulse is out of the range of such a simple circuit.  Even danadak's posted suggestions is similar to my 1 transistor solution and wont wait until after the peak has been reached to be assured that the source has switched direction and the true peak for that pulse has been reached.

You'll need a good scope to work on this circuit.  Also, you'll need J-fet amplified low capacitance probes, or, you'll need extra transistors to buffer the signal typically in an emitter-follower configuration.
« Last Edit: February 22, 2018, 04:31:22 pm by BrianHG »
 

Online BrianHG

  • Super Contributor
  • ***
  • Posts: 7725
  • Country: ca
Re: Pulse Peak Detector
« Reply #3 on: February 22, 2018, 04:43:14 pm »
https://www.edn.com/design/analog/4347624/High-speed-peak-detector-uses-ECL-comparator



Regards, Dana.
This component is no longer available.  You can find a few if you only want to make a few of your project, but, do not go into production with this IC.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf