In article , Joe McElvenney 
 writes: 
 
   Has anyone a pointer to the RF output impedance of the 74HC 
series gates (at about 10MHz)? I have a few inverters/buffers 
left over in a package and thought of using them stacked to drive 
an output port via a simple matching network. I've seen this done 
before and have simulated it myself but would like it spelled 
out. 
 
Fairchild's application note AN-393 is an informative one from the 
standpoint of _transient_ switching analysis.  Analysis or simulation 
for impedance matching, presuming matching at a single frequency, 
leans towards _steady_state_ analysis. 
 
"Impedance matching" of very fast transition time rectangular 
waveforms must be done over a very wide bandwidth to insure 
fidelity of the rectangular waveform at the far end of the transmission 
line.  Digital devices are characterized under near-total On or Off 
current source drivers...working into very non-linear active device 
loads. 
 
To get into the theory of a narrowband source impedance condition 
is possible.  Some of that has already been done with the "Class E" 
amplifiers using certain SCR devices at 40 meters, such as several 
examples at Cal Tech, Pasadena, CA, groups.  A search on Class E 
material might turn up something that would fit your application. 
 
Class C tube or transistor operation can be considered a sort of "soft 
switching" condition similar to On-Off current sources but that is 
complicated by the fact that the On state of the active device varies in 
current carrying capability depending on a narrowband load... hence the 
"soft switching" label.  Triode to pentode vacuum tubes were never very 
good switches in On-Off operation compared to transistors.over the 
very wide bandwidths necessary for fast-transient digital waveforms. 
 
While it might be an interesting intellectual exercise in analysis, it will 
take considerable time.  Far more time than geting up a small 
breadboard test circuit and TRYING IT OUT.  Nearly all of the "74" (or 
"54") family components are relatively cheap...small-scale integrated 
circuit packages cost less than a dollar each in single quantities. 
 
Len Anderson 
retired (from regular hours) electronic engineer person 
		 
		
		
		
		
		
		
		
		
	
	 |