Re: soft source question - TDR simulation

[RUMSEY IAN SCOTT <rumsey@ucsub.Colorado.EDU>]

andy-
	Try a hard resistive source.  Part of the magic of the soft source
is that it's always matched to any kind of line.  That effect may be more
significant than the series resistor, and I think that a hard source may
be more realistic for a voltage-source-in-series-with-resitor type of
driver.


-Ian Rumsey

On Mon, 11 Oct 1999, BYERS  ANDREW CLARK wrote:

> Hello all,
> 	I have noticed a funny effect when I am using a resistive soft
> source on a microstrip line in lc. What I want is a TDR-type of a source,
> which is a driver with a series resistor that is approximately matched to
> the impedance of the line. The initial voltage launched on the line is
> then one half of the voltage source. This initial voltage is simply a
> voltage divider between the line impedance and the source voltage. 
> 	For example, if you had a 1V source waveform, a 50 ohm source
> resistor (Rs) and a 50 ohm line (Zo), the initial voltage is 
> 	Vinitial = (1V)*(50/100) = 0.5 V
> This works fine in lc, for when I match the resistive source to the
> (estimated) line impedance, the initial voltage is about 0.5V. 
> 
> 	Another example is when the line is overdriven, or the source
> resistor is lower than the line impedance. If Rs=10, Z0=50, V=1V,
> 		Vinitial = (1V)*(50/60) = 0.833 V
> The same should be true for underdriven lines (Rs > Zo), where the initial
> voltage is less that the desired 0.5V.
> 
> 	Here is the problem: when I vary the resistance of the soft source
> in lc, I do not see this variation in the initial voltage launched onto my
> transmission line. I see some weak variation, but it appears that the
> resistor is not working like I had suspected. I have not delved into the
> theory too hard yet, but can anyone tell me if they have got this working
> and how? Or why mine is not working? I am source the end of a microstrip
> line that is open (ie not terminated into the PML). The source is 4-5
> cells from the boundary. The source is the width of the line, etc, etc. 
> 
> 	Thanks in advance, Andy Byers
>