Zener diode below-threshold reverse current

lable up-thread? That wasn't tactical, just economical.

Wasting the reader's time by leaving text in your post that they don't need to read is an issue.

old.

Sadly, the RC is much more sensitive to stray capacitances than the LC, and correspondingly less designable. In your initial LC design you neglected t he non-trivial parallel capacitance of your inductor (which is at least con sistent and often documented) and now you are neglecting the stray capacita nces around the inputs to the LTC1441, which are likely to be of a similar magnitude to the 1pF capacitor you explicitly show.

nitially assumed - in the LTC1441.

e a 50% duty cycle output, but what I should have done was to have looked a t the tp HL and tp LH times at the top of page 6 of the data sheet - the LT C1441 turns off more slowly than it turns on - see also the top of page 5 o f the data sheet, which make it clear that a typical device can barely make it to 100kHz.

"Typical" means that it sometimes barely makes it, and my implication was t hat there is a risk that a slower example of the LTC1441 might not.

I did. The BFR92A-based two transistor multivibrator has got plenty of spar e bandwidth in hand.

ndency to make the 555's >particularly horrible resistive voltage divider e ven worse. In the 555 the "resistors" are diffused >into silicon, their tem pco is vile, and their voltage coefficient of resistance even worse.

The one time I tried to use the NE555 in a tolerably precise oscillator, it wasn't remotely good enough. A classical emitter-coupled two transistor os cillator did a lot better, but still fell down on the temperature coefficie nt of the Early effect. It wasn't my daft idea, but I got it to the point w here we had something that was only marginally inadequate.

o make nearly as many as you think you discover.

But you didn't.

ting any of the stray capacitances around your explicit 1pF capacitor, this is less of an achievement than you seem to imagine.

And their tolerances.

You'd have to lay out the circuit first, which is more work than a hypothet ical exercise can justify.

apacitance transistorswould draw most of the OP's 10uA current budget at 10

Wrong. I'm not making any hidden assumptions. I'm just pointing out that it was a pedagogic exercise, which should have been obvious anyway.

ance (here several) that would bugger it up in real life.

If you'd explicitly specified that your "1pF" was a placeholder for decided ly uncertain heap of stray capacitances, you might be able to make that cla im, but your simulation looks rather more like one of John Larkin's tweak-u ps, which simulates okay, but doesn't represent anything that would oscilla te all that close to 100kHz if you built a real example

her level error of exactly the same kind. "Substitution" should have figure d in your list

So what? It was a typo, and not worth getting excited about.

titution of a small number of characters"

us.

The thing about typographical errors is that professionals make them just a s often as everybody else. In a professional environment they usually get c aught by the checking processes that are there because everybody does make this kind of mistake.

This isn't a professional environment.

As for it not being a typographical error, think again. I obviously wanted to pull the word "comparator" from my mental lexicon, but got the adjacent word "capacitor". It was a substitution, which does happen to be one of the classic forms taken by error of action, even if you managed to leave it ou t of your list.

lly used for anything quick, they may not have bothered.

xample isn't particularly inspiring.

Nobody is going to build any of our 100kHz oscillators.

s missing from the model you are wasting your time simulating.

Reality is often crass.

On Wed, 7 Sep 2016 13:47:06 -0700 (PDT), snipped-for-privacy@ieee.org wrote:

Or - from my point of view - nothing that John Fields wanted to take seriously.

iously.

I've just equated your way with simulations to John Larkin's. That's a fair ly nasty insult.

That's because you don't want to recognise quantified data when I dig it ou t and throw it at you. You imagined a 500mH inductor with no parallel capac itance, and are still unwilling to admit that it was an unrealistic assumpt ion (and fatal to any practicable version of the circuit that you initially put forward).

Says the dullard, who prefers to think that I haven't shown him up.

e:

seriously.

airly nasty insult.

You haven't seen some of John Larkin's simulations, which looked superficia lly fine if you ran them for exactly the period John Larkin specified, but fell to pieces if you ran them any longer.

He's technically competent, but uninspired. His occasional irrefutable comm ent tends to be a cliche, and much of his output - for example everything h e says about climate change - is in dire need of refutation.

We are both cantankerous and elderly, but you haven't got the patents and t he cited paper, and I haven't got your enthusiasm for the 555,or your skill s at making it do things that are better done by practically anything else.

out and throw it at you. You imagined a 500mH inductor with no parallel ca pacitance,

The real inductors I dug out had both series resistance and parallel capaci tance. The one with a self-resonant frequency below 100kHz didn't work in y our circuit, while the one with a self-resonant frequency above 100kHz coul d be made to. Self-resonant frequency depends on parallel capacitance rathe r than series resistance, so I did emphasise that defect in your simulation . Infinite Q tanks circuits can pose other problems, but the practical limi tations of the LTC1441 meant that you didn't run into them.

put forward).

My post making the original point did cite the part numbers and manufacture r of the two example real inductors, and the relevant numerical data (which is also available from the manufacturers data sheet).

Plug the numbers into your simulation and you'll see my point proven in spa des.

Your failure to take parallel capacitance into account was a qualitative er ror. My illustration that it was an error did depend on specific quantitati ve numbers. The confusion is all yours, and it's a pretty dramatic confusio n, as these things go.

rote:

ote:

ke seriously.

a fairly nasty insult.

cially fine if you ran them for exactly the period John Larkin specified, b ut fell to pieces if you ran them any longer.

Your simulation wasn't crap in that particular way. The 2008 simulation tha t included a 500mH inductor with zero parallel capacitance and zero series resistance was unrealistic crap, but not quite so obviously dishonest - tho ugh John Larkin may not have realised that his simulation would fall over i f he'd run it for longer.

The recent simulation with a 1.0pF capacitor as a place-holder for a bunch of stray capacitance exhibits much the same imperfect grasp of reality as y our 2008 effort, but it's harder to rescue. In 2008 I could find real, off- the-shelf inductors that could be plugged into your circuit, and one of the m would have worked in real life.

The current offering would need a printed circuit layout before you could g et a realistic idea of the stray capacitances, but we don't need to do that to know that those stray capacitance would be big enough to matter.

He's essentially providing a service to physicists who are smart enough not to try to design their own electronics. I did much the same job for the sc ience faculty workshop at Nijmegen University from 1993 to 1999, and most o f the job was steering the scientists towards products they could buy, rath er than spending a lot more on developing special purpose gear. We did do o utside work - a board I designed was supposed to end up in orbit - but the Dutch market wasn't remotely big enough to encourage me to go into business on my own.

ange

But true.

d the cited paper, and I haven't got your enthusiasm for the 555,or your sk ills at making it do things that are better done by practically anything el se.

I boast about mine, and you don't boast about yours. My dog is thus probabl y marginally bigger, but since my point was to emphasise the differences, r ather than make value judgements, why get excited?

It's certainly inadequate now. It was very handy in a lot of applications w hen it was first invented, and stayed useful for about a decade thereafter. I didn't happen to be involved in any of those applications back ten.

There are still legacy designers around, and amateurs who copy old circuits .

it out and throw it at you. You imagined a 500mH inductor with no parallel capacitance,

acitance.

And I got to reveal it because your grasp of the obvious hadn't been quite as firm.

circuit, while the one with a self-resonant frequency above 100kHz could be made to.

You've got the .asc files. If you haven't worked it out for yourself by now , you come across as remarkably incurious.

With the inductor with a self-resonant frequency below 100kHz the simulatio n would oscillate, but at less than 100kHz - I do know how to make such cir cuits oscillate faster, but not within a 10uA supply budget.

With the smaller inductor, with the higher self-resonant frequency, I had t o add parallel capacitance (not a lot) to get it to oscillate at exactly 10

te:

te:

ote:

take seriously.

's a fairly nasty insult.

rficially fine if you ran them for exactly the period John Larkin specified , but fell to pieces if you ran them any longer.

that included a 500mH inductor with zero parallel capacitance and zero seri es resistance was unrealistic crap, but not quite so obviously dishonest - though John Larkin may not have realised that his simulation would fall ove r if he'd run it for longer.

My agenda is entirely personal. I haven't sold my birthright for a post of message. I post stuff here that I find interesting - for a variety of diffe rent reasons - and I react to stuff that I find irritating, again for a var iety of different reasons. If there is a common thread, it's anti-nonsense.

Your 500mH inductor with zero parallel capacitance (and zero series resista nce) was not only nonsense, but misleading nonsense.

John Larkin is a renaissance man when it come to nonsense - he's got silly ideas about a whole range of subjects, and resents being identified as bein g ill-informed on any of them.

ch of stray capacitance exhibits much the same imperfect grasp of reality a s your 2008 effort, but it's harder to rescue. In 2008 I could find real, o ff-the-shelf inductors that could be plugged into your circuit, and one of them would have worked in real life.

It's speculative, rather than fatuous. I've made quite a few of those kinds of speculations through my career, and been right quite a bit more often t han not.

If your circuit hadn't worked in reality - with a properly selected real in ductor - the most likely problem would have been a lack of gain-bandwidth i n the actual LTC1441 selected to realise it. The LT simulations are of typi cal parts, not worst cases.

I posted that information back in 2008, and I've provided a link to the rel evant post. Go for it.

d get a realistic idea of the stray capacitances, but we don't need to do t hat to know that those stray capacitance would be big enough to matter.

I'll do a layout for my low distortion oscillator first - I've got KiCad on my Linux partition, and I've started the KiCad version of the schematic, b ut I haven't touch it for the last six months. Editing the next issue of th e quarterly newletter of the NSW branch of the IEEE has a higher priority, and I don't have to crank up my Linux partition to do it.

Your silly idea is further back in the queue, and would serve no useful pur pose at all.

not to try to design their own electronics.

That would be Winfield Hill too. We both know we can design our own electro nics. Win had the sense to drop out of a Ph.D. in Chemical Physics, while I persisted with mine in Physical Chemistry, and only moved over to electron ics after I'd got the degree. By moving early, he ended up in a classy elec tronics environment rather faster than I did - it took me six years post-Ph .D. to make it to EMI Central Research.

versity from 1993 to 1999, and most of the job was steering the scientists towards products they could buy, rather than spending a lot more on develop ing special purpose gear.

No. The academics did their own purchasing, once they'd been made aware tha t there was stuff out there that they could buy. The professional publicati ons they read have ads for that kind of stuff, but they needed to have thei r attention drawn to the existence of the ads. Graduate students don't need much hand-holding, and tend to over-estimate their capacity to get by with out it.

We occasionally had to help them dig themselves out of the holes they had d ug, but mostly they did fine.

If you define success as looking like John Larkin, I may be a failure. I've spent a career doing stuff that I'm good at, and doing it well.

I'm not doing much at the moment, which means that I can imagine a better o utcome, but I can't see a way to get to that imaginary better outcome, and suspect that there never was one.

Your imagination is less restricted by my reality, and John Larkin can't im agine that anybody could be in a better place than the one he currently occ upies - narcissists are like that.

That's modern science for you. It's a scheme for presenting hearsay in a wa y that allows the stuff the gets it right to reinforce other stuff that has got it right. Seems to work well.

and the cited paper, and I haven't got your enthusiasm for the 555,or your skills at making it do things that are better done by practically anything else.

You would say that, wouldn't you.

hasise the differences, rather than make value judgements, why get excited?

It may feel that way to you. I regard myself as more into anti-nonsense tha n into self-promotion, but I would say that, wouldn't I.

s when it was first invented, and stayed useful for about a decade thereaft er. I didn't happen to be involved in any of those applications back then.

I wasn't averse to quick and dirty. My particularly brilliant colleague was the one who earned the nick-name "Puker" by coming up with such revolting effective kluges that the rest of us wanted to throw up, but the rest of us did learn by imitation.

Sadly, the NE555 was a bit too dirty even for that environment. EMI Central Research was a high-tech place, but they mainly wanted stuff that would wo rk, and there was one high-tech project (which I managed to avoid, largely by being very rude about it) which was a theoretical nonsense though sponso red by a guy who had at least one very impressive high point in his CV, bes ide a whole bunch of self-indulgent idiocies.

its.

Something that works beats nothing at all, even if it works a lot less well than a better design would.

dig it out and throw it at you. You imagined a 500mH inductor with no paral lel capacitance,

capacitance.

te as firm.

Not that I recall. Since it was the absence of capacitance that caught my a ttention, I might have skipped the rest of the waffle.

ur circuit, while the one with a self-resonant frequency above 100kHz could be made to.

now, you come across as remarkably incurious.

tion would oscillate, but at less than 100kHz - I do know how to make such circuits oscillate faster, but not within a 10uA supply budget.

d to add parallel capacitance (not a lot) to get it to oscillate at exactly 100kHz, as I spelled out in 2008.

It was a total pain to find - I had to make several forays into the archive before I could nail it down - but it's further up this thread somewhere, w here it is a lot easier to find

"On Wednesday, August 31, 2016 at 10:58:13 AM UTC+10, snipped-for-privacy@ieee.org wro te:"

is your post, and

was my response.

How come you couldn't manage that? Are you one of the Luddites who won't us e Google Groups? It has its weaknesses, but it is very accessible.

Not that I think it's time to call Guinness, but is there a flame war record?

George H.

Not hot enough to qualify as a flame war. I think we aren't accusing one another of anything worse than being slightly senile, and we are both old enough that that probably doesn't even count as an insult.

Don't rush them. Give them a couple more years to work on it.

Recycling lame, pointless insults does seem to amuse them.

Designed anything recently?