Menu

Extracting the 5th Harmonic

Search this Thread

Prev

Previous Post

Next Post

Next

#34

Report Post

Old

March 15th 04, 10:20 AM

Paul Burridge

Posts: n/a

Default

On 14 Mar 2004 23:43:57 -0800, (R.Legg) wrote:

The reactance of the C2 part is almost 3K at the fifth harmonic.

Input impedance of the biasing network is 300 ohms - this is about
half the small signal input impedance of the 3904 @4mA.

Even with bypassed emitter, only 1/3 of C2's AC current will enter the
base of Q2.

Understood. I chose a very low value for C2 because it was the third
harmonic that was dominating and most needed cutting down to size.
Though the reactance at the fifth is high, it's a lot higher still at
the third. Although I suppose I could have used a simple series L/C
tuned circuit tuned to pass on the fifth and attenuate everything
else. Do you think that might work?

If the resonant circuit used lower L and higher C values, C2 could be
increased without as severe an effect as it has here.

Good point.

As previous posters have stated, if the input is squarish then the
harmonics are already there.

Theoretically, perhaps! Damned if I can find the 5th, though.

There is a +/- 3% window on all the optimum duty cycles (ie 10, 30,
50, 70, 90%), including risetime, for which the 5th harmonic amplitude
is relatively constant, at about 10% of the initial peak amplitude.
Note that the 30/70% period is a median quasi-minima for both 3rd and
4th harmonics, possibly reducing LF filtering problems in the first
stage, as the 1st and 2nd are farther away. 50% being available, you
should stick to it.

I'm as close to 50% as I can get. If square waves are so finicky about
producing the 5th if their mark/space ratios are a tiny bit out then
maybe I need to consider some other way of generating such harmonics.

I don't know if you're doing any actual physical breadboarding. The
100mW power dissipation suggests not. Pre-apps it's time.

I'm using my 'leopardboard' technique: a layer of PCB with liittle
island pads etched out in copper and component ends soldered to those
islands.

If this is a physical breadboard, then perhaps you might let us know
what you are actually using for your 2uH inductors. You wouldn't want
the relatively hefty classA bias to have any effect on them, so there
should be a lot of air in their flux path - not a couple of turns on a
bead, I hope.

Nope. They should be okay. 5mm dia, air-core, 20mm winding length
--

The BBC: Licensed at public expense to spread lies.

Reply With Quote

Reply With Quote

Thread Tools	Search this Thread
Show Printable Version	Search this Thread: Advanced Search
Display Modes
Switch to Linear Mode Switch to Hybrid Mode Threaded Mode

Posting Rules
Smilies are On [IMG] code is On HTML code is Off Trackbacks are On Pingbacks are On Refbacks are On

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Shorted 1/4 wave stub ?	Henry Kolesnik	Antenna	87	June 17th 04 11:04 PM
A Simple Harmonic Generator.	Reg Edwards	Antenna	12	March 23rd 04 08:16 AM
Frequency multiplication	Jim Thompson	Homebrew	108	February 25th 04 04:54 PM

All times are GMT +1. The time now is 07:43 PM.

About Us

"It's about Radio"

Contact Us RadioBanter forum home Privacy Statement

Copyright © 2017