The stupidity of slimulations

But they make up for it in the numbers of very well paid administrators.

I think it's hilarious when they whine about what Obamacare is doing to them.

??

Krw does resent the existence of people who are smarter than he is - which is most of the population.

that

Vdc comes in straight or reverse polarities, Vac swings both ways.

?-)

r1 q1 q2 r2 q3 q4

q2 is an emitter follower.

r1 q1 r2 q3 is a potential divider.

q3 q4 is a mirror provding a pull-down on the q2 emitter

the divider biases q2 to provide vdd/2 at the output.

tc is very good due to the symmetry.

what's "Is" ?

If the ferrite does nothing, why even have it? I don't understand its use in this case.

It is the "saturation" current, JT was trying to get folks to wade through the Ebers-Moll model to get the same answer you and I got more intuitively ;-}

piglet

e transformers. Most of the high frequency current through them flows betwe en the two sides of the transmission line and never gets out to influence t he ferrite.

"Most" isn't "all". And "wideband" does include relatively low frequencies. I measured the first one I put together, and it was flat from about 50kHz to about 150MHz which - as Win Hill eventually pointed out (and I hadn't no ticed) was where the length of the transmission line in the transformer was down to a quarter wavelength.

The ferrite was working quite hard at 50kHz.

"Jan Panteltje" schreef in bericht news:lgjf7h$tk8$ snipped-for-privacy@news.albasani.net...

The stupidity of slimulations. I have noticed a lot of simulating people using slimulations as replacement for reality. I am a bit pissed so if you are offended you could be one of them :-)

We see that everywhere, NASA ... artist impressions of alien planets, complete with cows ? grazing, spacecrafts never build, like a bad dream. Awakening.

Slimulate it.

Unfortunately (or unfortunately depending on how you want to or are forced by the lack of a full working neural wired constellation) it is no good..

I hear about (I keep clear of them people) developing software with for example 'slimulators', I alway wondered if they ever make something that works, they do not normally release code and if they talk about programming it sounds dumber then a duck going quack quack.

And the same for FPGA, electronics hardware, what not. And they sort of force others to do the same.. that is not design, is not understanding, as just a few limited equations do not describe reality, parasitic capacitances, what not. And by the time you counted them all in in your slimulation you cannot even build it like that, Result: Bloat hardware like Jimp somson fixing problems that are there because of wrong approaches, making a market for yourself fixing things adding ever more crap to something that is crap to begin with, its bloat bloater and some bloat and more to fix the bloat as in software, and now hardware too, We do not normally know these days whats in them chips but for some I fear the worst.

W[here]TF is some basic understanding. We are neural nets and should use that aspect, we are no computahs and a bunch of equations, THAT can be automated Mr Hobbbbs Use your g*d*d*... soldering iron learn how to use a f*cking scope, 'I have this storage scope and still cannot see a 2 us pulse' looks to me like you really need a brain scan :-) That is like J.L. who has a collection of what seems to be the most expensive and fastest scopes to impress people and nothing it does. Anybody left out? I like Lieberman's story about the MRI, yes this is electronics and humanity these days. And that club wants to fight the Russians, the knicker black in white house with his pushbutton phone. wonder if he has a red black-berry. To talk to Putin, hell US needs Russian enjuns to even launch the spy sats, gimme a break. Your electrical grid is collapsing, you try to make war in Europe, why should I not support a round of nuking to help humanity with a free fresh start. mm maybe better get some coffee and actually put that clock code on the website,. been long time since I did see anybody here publish something that works, F8ch you

Skipping a lot of details, a simulater is just a tool. Like every tool it has its use for some people. Like every tool you need to know how to handle it and you need to be aware of its limitations.

As an example look at flight simulators (the real ones). They have there use for pilot training but you will never become a pilot by flying simulators only.

If for nothing else you can use LTSpice for distributing schematics in plain ASCII on text-only newsgroups. Much better readable then hand written scratches on ABSE :)

petrus bitbyter

On a sunny day (Tue, 25 Mar 2014 14:40:14 +0100) it happened "petrus bitbyter" wrote in :

Yea, all good and well. Actually wine crashes my x server. I have read Microsoft is one of the largest contributors to Linux these days. So they probably f*cked up wine, or the x server, or both, to force people to buy their crap. So I won't run LT spice on this machine, do not have wine installed on the laptop for same reason. Maybe the LT guys should do a true Linux port.

xfig is a nice program, but I cannot be bothered to use it. ABSE I do not have on these news servers (text only). And anyways ASCII art works OK. It is my strong suspicion some people here still run DOS in a 640x480 screen on an half dead CRT from the eighties with a viewer that resizes my 2852x1975 true color JPGs to 640x40 8 bits deep, So help is not possible in that case. BTW download those and use xv -nolimit in Linux. MS Windows is all over anyways.

Yes LT spice is a tool, I used a huge pair of pliers as nut cracker this week. HUGE, works fine.

In other words, build your transformer out of ideal transmission line segments, like SPICE thinks of them -- perfect differential lines, infinite CMRR. Then stack them up and put ferrite around it to sort-of realize the CMRR in reality.

People involved in tube hi-fi have the same problem: thousands of turns around a crappy core (laminated iron), yet wideband (20-20kHz). The solution there of course is to interleave the primary and secondary windings (and their subsections, if tapped) -- but not so much that you drive up the capacitance (i.e. drive down the impedance). Same rules apply, they're just harder to derive (lumped capacitance equivalent of a bank winding?) and usually approached by seat-of-pants (as is traditional for the field ;) ) rather than theory. Experienced designers know that the core is basically meaningless above 3kHz, and you get the last decade or two (for good distortion and phase response with negative feedback, you need well over 20kHz at the top end!) entirely from construction.

Tim

You don't get the job >:-} ...Jim Thompson

Nope. Intuitive was what I was after, but the intuitive answers I'm seeing are just a wee bit too loose.... indicating lack of understanding, showing only good "guess-man-ship". ...Jim Thompson

Actually I threw in "IS" to confuse those who _think_ they understand transistors... like people from Berkeley and Tulane >:-} ...Jim Thompson

Very interesting, indeed. Thanks for the explanation.

Transmission line transformers are amazingly useful building blocks in wideband signal processing, but besides RF gurus, hardly anyone seems to know and use them.

The classic references are Ruthroff: "Some broad-band transformers", Guanella: "Nouveau transformateur d'adaptation

(That last one is a book.) There are many more, of course. Well worth knowing, in my opinion.

In my case, the fact that they are rad-hard is an important additional feature.

Jeroen Belleman

(I was too chicken to post this originally)

It looks like it ought to be some standard analog chip building block, but I figured it like this:

left hand side of circuit is just two resistors and two diode-connected transistors. So transistor above Vout is just buffering Vcc/2, with the diode drops cancelling.

The lower transistor below Vout is harder to understand, but it must be working as a current mirror so that both sides run at the same current. Which must work to further balance the Vbe drops. Yay.

So it is a buffered 0.5Vcc generator, using identical but not well specified resistor values (like you get on chips I imagine).

The biggest pain of integrated circuits is the absolute value tolerance on resistors, typically +/-25% or more. The joy is that it's relatively easy to get 1% or better ratios, so ratio design is the magic bullet to get good chip performance.

Though, in recent years, automatic trimming has become so commonplace that precision chips are a piece-a-cake.

For economy, I usually try to get the customer to agree to one external resistor. I can apply a bandgap-related voltage to that, and get a bandgap-accurate current that I can ratio throughout a chip. Accurate, yet cheap-cheap-cheap ;-) ...Jim Thompson

The ferrite handles the LOW freuency and the air [ignore the ferrites] handles the HIGH frequency. That's why he referred to "...extending the high frequency performance..." and working with WIDEBAND transformers.