Are hybrid parameters useful for design?

Hey, you're wrong by a bit lower than 1.5% (at room temp, which of course I don't remind what it is :-)

Another useful value is 1K = 4nV/rtHz

or that 3 uA of photocurrent has a 1-Hz

So are PhD's ;-)

Seriously, VAF is just a parameter in another STYLE of modeling. And math is math is math. "Textbook curiosities" tweak the mind to think of alternate solutions to untenable problems. ...Jim Thompson

Sure thing. But it's a reasonable approximation over a wide range, because it's physics-based. h_re depends linearly on collector current, for a start.

And at least I'm not in a _pathology_ texbook. ;)

Cheers

Phil Hobbs

h-parameters can be complex math expressions, not just fixed numbers. ...Jim Thompson

Sure again. But then where's the benefit? A 2x2 matrix written out with all sorts of algebra and stuff sort of defeats the point of the exercise--it's far clearer just to write it out the usual way.

Cheers

Phil Hobbs

Phil wrote a great book on electro-optical design, with a goodly bit of low-noise electronics included.

He also serves as our reference LPTM (Large Physicist Test Mass).

The 'benefit' is the ability to easily model things that don't yet have models. Let the simulator cope with the Algebra. ...Jim Thompson

On Mon, 15 Oct 2012 12:16:19 -0400, Phil Hobbs wrote: A 2x2 matrix written out

If it is a text-book curiosity why bother to write it at all! Of course, Jim says he very occasionally uses it - which means he rarely uses it. Indeed, why should he? Oh, God! We got our time wasted on this!

Guys, now we are really discussing issues! Thanks!

I am so sorry, I was writing when JT was writing too. But do we require these simplified models at all?

I'm talking about the benefit of the H parameter format, i.e. organizing two (2) equations into a matrix. All it is, basically, is a way of collecting terms in an equation.

SPICE doesn't understand matrices in its input files anyway, AFAIK, so you'd have to write the silly things out longhand anyway.

Cheers

Phil Hobbs

It's not that hard. Remember orthogonality ?:-)

Here's how S-parameters are handled in Spice...

h-parameters can be done the same way. ...Jim Thompson

We're actually in violent agreement, then. No matrices that I could see. (I like the behavioural model for putting in S parameters. SPICE is just a glorified nonlinear sparse ODE solver, after all.)

Cheers

Phil Hobbs

Well, in a twisted way that kind of makes my point: h-parameters are introduced so that the student can use linear analysis to get a feel for how circuits work. You already knew.

One of the conceptual models that I find useful is a pair of back-to-back diodes, with a dependent current source around the B-C junction, possibly with appropriate base spreading resistance if that's going to matter. It does a good enough job of predicting most of the pertinent linearized parameters at any one operating point, and it makes it easy to visualize what's going to happen if anything reverse biases.

("Good enough job" is defined here as "kinda in the center of the variation you can expect anyway").

I honestly feel that one should not do SPICE modeling unless you _really_ understand the circuit: otherwise you'll miss something important.

are

Yep. My point, exactly. To ease my head-scratching processes I've developed a number of "gadgets" for PSpice that make solving (and displaying) such solutions a piece-a-cake.

And LTspice is next. I've finally cracked the 4046 (in terms of behavioral modeling) and will release a first pass on the VCO portion shortly. ...Jim Thompson

Spice is the "check" portion of a design. Spice doesn't pick, or place, any of the components I place on a schematic... I do.

Spice can do process corners and/or Monte Carlo in minutes. A human might _never_ be able to discover all possible situations. ...Jim Thompson

Ah! That's the point. But when teachers don't supplement the text books, you find yourself worse off than between the devil and the deep blue sea. ...

Yes, I thought asymmetrical diodes represented the circuit better.

Sorry I forgot to emphasise how important this comment is! So, it is better that instead of remembering the formulae of h-parameters, we do the analysis in the same way as h-parameters are reduced using simplistic logic. This at least sound logical :)

To get S11, I usually put the DUT in one side of a Wheatstone bridge, which is exactly the way some VNAs do it in real life. Apply 'AC 2' to the top of the bridge and get S11 as the voltage across. It's very similar to the approach in your reference, but it uses only one AC source. It's so simple I never bothered to encapsulate it in subcircuits.

Jeroen Belleman

(Today I realized that I answered this privately to the OP -and not to the group. Here are my thoughts on this:)

Hybrid parameters are just one of several representations for two-ports. Working with the conventional port variables V1, I1, V2, I2 there are 6 combinations of the independent variables, each one giving one specific family. In practice, one family of parameters may be easier to measure, or a given family may not exist.

So, your question should be rephrased as: is the two-port concept useful for design? And the short response is yes.

Pere