Author Topic: Let me get this straight once and for all.  (Read 1776 times)

0 Members and 1 Guest are viewing this topic.

Offline SimonTopic starter

  • Global Moderator
  • *****
  • Posts: 18118
  • Country: gb
  • Did that just blow up? No? might work after all !!
    • Simon's Electronics
Let me get this straight once and for all.
« on: March 15, 2019, 07:29:21 pm »
Right, Attached is some of my coursework. I am trying to work out haw to calculate harmonic values, For example I am asked for the 101st harmonic, I make that using the big formula on page 3 4.7x10^-32V for a 1MHz pulse lasting 10ns wise a rise time of 100ps.

I am troubled by the fomula at the bottom of page 3 that states that the individual sinusoides will be the....... and that is simply the formula for the amplitude at some point in time of any generic sinusoidal waveform and bears little relevance does it not?
 
The following users thanked this post: Chris56000

Offline dmills

  • Super Contributor
  • ***
  • Posts: 2093
  • Country: gb
Re: Let me get this straight once and for all.
« Reply #1 on: March 15, 2019, 10:55:07 pm »
10^-32 Really? That is small.

Hold on :

Fundamental is clearly 1MHz (1us period).
First breakpoint is at 1/PiTw = 1/(3.1 * 10ns) = ~32MHz.
Second breakpoint is at 1/PiTr  = 1/(3.1 *100ps) = ~ 3200MHz.

Thus the 101st harmonic is on the 1/f slope.
 
Low frequency amplitude is (from figure 4) 2ATw/T, Tw is 10ns, T is 1us, so below the first breakpoint amplitude is 2% of whatever the peak amplitude of the 'square wave' is.  And your 101st harmonic is only about 3 times the first breakpoint, so I find 10^-32V highly unlikely.

Of course you might have a combination of numbers that happens to give a deep null at that particular harmonic, and I really cannot be arsed to do the sums. 

Regards, Dan.
 

Offline SimonTopic starter

  • Global Moderator
  • *****
  • Posts: 18118
  • Country: gb
  • Did that just blow up? No? might work after all !!
    • Simon's Electronics
Re: Let me get this straight once and for all.
« Reply #2 on: March 15, 2019, 11:23:06 pm »
i made a mistake in the calculation. It is 3.56x10^-4.

But here is the thing. My calculations do not make up the graph they theorise and i proved this with my signal generator and oscilloscope in FFT mode. The harmonics drop off and then spike back up at the rise time frequency like my plot does but the material shows an entirely different graph.
 

Offline dmills

  • Super Contributor
  • ***
  • Posts: 2093
  • Country: gb
Re: Let me get this straight once and for all.
« Reply #3 on: March 16, 2019, 01:14:36 am »
I doubt that your AWG is managing 100ps rise time at the scope, being as it is only a 20MHz/250Ms/s unit (so must have a reconstruction filter cutoff below about 100MHz or so).

Then your scope will have a rise time measured in ns not picoseconds....

Regards, Dan.
 

Offline SimonTopic starter

  • Global Moderator
  • *****
  • Posts: 18118
  • Country: gb
  • Did that just blow up? No? might work after all !!
    • Simon's Electronics
Re: Let me get this straight once and for all.
« Reply #4 on: March 16, 2019, 08:13:51 am »
No it's a 6.6ns rise time which is the 50MHz peak. My point is that contrary to the theoretical graph the harmonics due to the principle waveforms frequency die down to near 0 and the Rise time frequency causes a renewed spike in the harmonic content.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Let me get this straight once and for all.
« Reply #5 on: March 29, 2019, 10:52:25 pm »
Probably too late, but have you tried using a simulation package such as LTSpice? I would have a go myself, but that PDF doesn't view on the PC I'm currently using at the moment. I'll try it when I get home on Sunday.
 

Offline SimonTopic starter

  • Global Moderator
  • *****
  • Posts: 18118
  • Country: gb
  • Did that just blow up? No? might work after all !!
    • Simon's Electronics
Re: Let me get this straight once and for all.
« Reply #6 on: March 30, 2019, 12:53:58 pm »
Yes i could have a go at what options LTSpice has.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf