[Lowfer] Matching transmitters to transmission lines

[Bill Ashlock]

Johan,

I ran a very interesting experiment last night to prove my point about the 
matching issue, and guess what? I disproved it!

I used my modified version of Lyle's stage which is the same as the one I 
use for my loop final. This has four base diodes instead of two and they 
limit the degree of collector saturation in order to maximize efficiency. I 
connected 100ft of RG-59 with a 70 ohm load resistor on the end and set the 
Vcc for 20v. The waveform at the input to the coax was a near perfect square 
wave and the TO-92 transistors were cold to the touch! The measured input 
power was only a couple of percent greater than the power at the coax 
terminating resistor. There was a slight glitch, maybe 10% of max (not 
ringing), on the waveform at the receiving end of the coax about 10% beyond 
the leading edge but this would be of no consequence for most applications. 
Installing a series 70 ohm resistor at the sending end eliminated this 
glitch but of course reduced the voltage by 1/2.

These test results are totally different compared to what I see with my 
beacon's exciter and I shall continue the testing to determine what causes 
the large ringing. BTW this circuit uses a complementary emitter-follower at 
the output. Don't emitter followers have a tendency to oscillate when 
driving inductive loads because of a negative resistance effect? Maybe the 
'fix' is to simply change to a common emitter design like the final.

Thank's for insisting that a coax can be driven with a near 0 ohm source! 
I'll let you know what I find out with the exciter.

Bill


> > I think I am going to have to respectfully disagree with your statement
> > that there is no problem in driving a coax with a 0 ohm square wave.
> > I have tried this numerous times and large ringing occurs at the
> > transmitting end even though the signal at the receiving end looks 
>fairly
> > close to a square wave.
>
>Are you sure the load is purely resistive 50 ohms at all frequencies?
>Square wave contains a lot of harmonics and a "tuned" load will
>probably be mismatched on all frequencies except the fundamental.
>
>
> > I just did a search in the ARRL handbook and I can find no example of a
> > class D PA driving a coax unless there is a impedance matching network
> > between.
>
>That is probably because the goal is to produce a sine wave. "Voltage
>switching" class D amplifiers, such as Lyle's complementary "totem pole",
>a Decca style H-bridge or even a HCMOS gate all need a series resonant
>circuit at the output. The fundamental frequency passes straight through 
>the
>series circuit while the harmonics will see a very high reactive impedance
>(depending on Q). There will be practically no harmonic current in the
>finals, only voltage.
>
>
> > In fact in every example I can find relating to transmission line
> > theory the source impedance is always matched to the
> > transmission line with the same impedance.
>
>I disagree. That would limit amplifier efficiency to 50% maximum, wouldn't 
>it?
>
>73
>Johan SM6LKM
>
>
>_______________________________________________
>From the Lowfer mailing list
>Send messages to: Lowfer@mailman.qth.net
>To sub/unsub visit: http://mailman.qth.net/mailman/listinfo/lowfer




_________________________________________________________________
MSN Photos is the easiest way to share and print your photos: 
http://photos.msn.com/support/worldwide.aspx