Good candidates would ask you if it was a Ge or and Si transistor. Better ones would worry about semi-conducting polymers as well.
If you insulted me by asking that kind of question, my reply would also be satirical.
NPN.
"Too trivial" means a lot less than it used to.
"It's all about me! Me! Me! Me!" You've defined the fundamental asshole >:-} ...Jim Thompson
There's not a lot of "me" in ""Too trivial" means a lot less than it used to."
Jim is coming across as more like the orifice in question than even John Larkin, not that there's a lot to chose between them.
On a sunny day (Sat, 22 Mar 2014 17:02:05 -0700 (PDT)) it happened Bill Sloman wrote in :
Triggered Bil's glowball circuit
Something wrong with your idents > needed 2 >> here, I wrote that
I was addressing manned spacecraft, ISS is just a few km up and they have never been more than a few hundred km away from mama earth since Apollo was scrapped.
I just did read Buzz Aldrin actually peed on the moon right after he landed, dog sort of thing to leave your mark. At least a proof we were there.
BTW scrapping any unique working energy generating source creates a risk. This week I was reading about a paper that found (by experiment) that the DNA repair mechanism adapts to stronger radiation, and makes the next generations more immune to it. Sorry no reference, German teletext IIRC, if I come across it again I will let you know[1]. Does not help you or me, but for persistence of humanity it does. Traffic .. and lost planes .. take more lives than nuclear plants ever did, coal mining even more. But the greens are working on mining ever more coal, you don't give s*it about miners do you? Dust lungs, did you know that was why the coal mines in the south of the Netherlands were closed?
[1] Could be government planted to get back [prepare the public for] the nuke plants going on line again.
I have, but not for me :-)
But is is relatively easy to fix that, if people get of their arse.
What I find is that people won't bother how to learn how to use their basic tools of trade. What would we think of a carpenter that could not sharpen his chisels?
A competent Analog designer should be expected to be able produce serviceable models to run in Spice. It isn't rocket science! Most can't, but it should be part of their basic job. The simulator is already written for them.
Kevin Aylward
I disagree. Have you tried hand calculating the effects of colliding rotating black holes using General Relativity :-)
As I noted, modern analog asic designs are way too large to confine simulation to just checking devices. Simulation allows true invention and design of new topologies.
Its a ahmmm... a totally new paradigm, that for the most part, most modern analog asic designers engage in.
Its just a fact that that many aspects of circuit behaviour is ONLY understood by by running simulations. Its an ego based facility that argues that to understand something, one can understand everything by paper and pencil. A clear example is Phase Noise. The correct theory is just impossible to actually design with.
Large chunks of physics was only achieved by experimental methods. The complexity of many electronic systems are such that there is an "as if" there is no or little understanding of what is really happening, despite many golden brains claiming that they do understand "by inspection". There needs to be honesty here. By and large, asic analog simulation is at the stage where the simulation shows up real effects, that were never anticipated prior to simulation. A classic is a particular issue in 1/f phase nosie up conversion. I actually discovered this independently of Ken Kundert analysis below, and it was quite refreshing to understand that this was universally not understood!
Only brute force complicated non-linear stochastic equations using phase noise simulations identified this issue.
Sure, you need some brain power, but computing does most far better.
That was very good then, and quite an achievement for its day, but times really have moved on. Typically, I do one year development for one asic at
100 blocks X 50 transistors per block of of various, LDOS, amplifiers, function generators, oscillators, heater loops, etc.... with 1st time pass. A 10 transistor LM3080 would be a small detail at an hours work to fully design and worst case via simulation. I would say the joint probability of 100 blocks all working simultaneously compared to 50 separate chips, is a tad lower :-)I think most here know me for doing sums, as in General Relativity
Typically, I truly do millions of simulations a year.
Kevin Aylward
Sure, but many people don't use simulation because they don't understand electronics either.
Yes. Yes. Yes.
But still a very, very, low probability though.
30+ years on, to be blunt quite, and I am still truly stunned at the (in) competence of alleged analog designers both 30y and novices alike.Typically,
Q. How do you compensate this op-amp A. ...poles... zeros...mutter... Q. What do you actually do... A. mutter blah. roll off.. 90Deg... Q. Physically, where do you place what and where... A. .ahmm... blank stare...
I was looking for "put a series RC across a node", any node would have done...
Over a few years, I interviewed around 50, from around 300 Resumes. Apart from 2, all were completely and utterly, useless.
Kevin Aylward
I disagree. Its trivial to sweep several capacitors, inductors and stacked transmission lines all at once, from 50ff to 50p, 1nh to 1uH, delays from
50ps to 10ns etc in AC and Transient runs.When I started my SuperSpice around 1997, Mike Englehardt touted the speed of LTSpice, he had a point then. In 1997 it was $30k per GFlops. Its now about 22c for the PlayStation. My home machine is 4 years old and does 30 GFlops. I can easily do full behaviour sims of Deleta-Sigma ADCs
Sure, and this can most easily be done via simulation. It tells you what the limits have to be for it to work. It does matter that you don't know the value is. Its just as easy to but in 1uf parasitic as it is to enter 1ff. Do them all!
Sure, and without worst case simulations of all reasonable component values, it is luck that it will work. Its simply not feasible to test real hardware with all anticipated variations.
People need to move on. They need to spend the time to actually model the system and components, rather than simply claim, it is no value because of this and that.
All PCBs can be effectively modelled with all practical parasitics.
Kevin Aylward
Jim Thompson scribbled thus:
People in power get very hostile when their ivory tower is threatened particularly if they realise that they could be wrong !
Well, that surprizes me because this was in my uni courses. Its basic E&M.
Of course, I don't manually use those equations, I just use a spice model with:
L1 1 2 {L1} L2 3 4 {L2} K1 L1 L2 {K}
Which does all the M=K.sqrt(L1.L2) messy stuff automatically.
Add Rs and Cs as necessary, and vary K to evaluate just how tight to unity the coupling really needs to be.
You need to understand what is going on, then do the bulk in simulations.
Kevin Aylward
Can I take lessons ?>:-}
All I said was, "No, that's not correct, it would look like this..."
And I could instantly see in his eyes that all was lost :-(
Yes, I learned the appropriate word for these common occasions. Schadenfreude.
Kevin Aylward B.Sc. snipped-for-privacy@kevinaylward.co.uk
It involves Chebychev functions.
Kevin Aylward
On a sunny day (Sun, 23 Mar 2014 10:31:41 -0000) it happened "Kevin Aylward" wrote in :
I was thinking about your si-mulations how deep do they go, I mean at the quantum level it is well known that now you are here, but there is a finite probabilliity you could appear at the south pole all of the sudden. I am sure your si-mulations also acount for that, and how often has this already happened to you?
Maybe that's what happened to that aircraft? ;-)
There's got to be a catch here somewhere, but I'll take the risk:
Vb= +5v Ve=+4.3v Vc=+10v Ib=43uA (assuming Beta 100)
Come on, then. Where's the catch?
Perhaps. One of the precision questions that was passing through my mind was modelling coupling in ferrite-cored transformers.
SuperSpice had rather natty transformer models, early on. I haven't had it running on my computer in recent years - Norton Anti-Virus doesn't like the Anasoft web-site for some reason, and I haven't been motivated to chance my arm by over-riding it.
Meanwhile LTSpice lets you use the John Chan hysteretic core model in modelling inductors, but not transformers.
And I wonder - from time to time - whether I should use the same coupling factor for two coils which are actually bifilar wound as I use for the multi-layer secondary I'd get wound on top of it.
If I was doing serous design, as opposed to retiree fiddling, I guess I'd find out.
On a sunny day (Sun, 23 Mar 2014 14:13:33 +0100) it happened Cursitor Doom wrote in :
Na... its in ORBIT!
I doubt if 90% of carpenters even HAVE chisels. ;-)
OK, write me a good model of a switching power supply. I don't have the information to do it, if I knew how.
about miners do you?
The greens aren't actually working on mining more coal. They'd prefer to se e more natural gas burnt - at least until we can get our wind and solar pow er act together - and natural gas is fairly easy on the people who extract it.
Netherlands were closed?
Not really. I'd though that it was because Dutch underground miners got too expensive. The DNA repair mechanism may well adapt to higher levels of ionising radiat ion - if you expose organisms to enough ionising radiation to kill half of them before they reproduce, you will select a population that is marginally less susceptible to ionising radiation. It's not a program you'd find easy to sell to the population being weeded out by the higher levels of ionisin g radiation, any more than Dutch coal miners liked being selected for highe r resistance to coal-miners lung.
Have something to add? Share your thoughts — no account required.
Ask the community — no account required