1. Home
  2. Electronics Design
  3. Re: Wheeler inductance formulas revisited - Wheeler82.zip (0/1) - Wheeler002.gif (0/1) - Wheeler_ProcIEEE_70-12(1982).pdf [1/2]

Re: Wheeler inductance formulas revisited - Wheeler82.zip (0/1) - Wheeler002.gif (0/1) - Wheeler_ProcIEEE_70-12(1982).pdf [1/2]

> W>>>

>>> Very good. Here's a pdf file. I'm a big fan of Wheeler's 1928 IRE >>> article, and am happy to see this more obscure work of his propagated. >>> Clearly, 54 years later, he thought it was an important improvement >>> on his early famous article, and it's good to see that now people can >>> actually read all about it, just as he wished. Twenty-three years ago >>> Harold Wheeler paid $190 to have these two pages published, and it's >>> nice to see his wishes and payment are not dead. Thank you, Phantom. >> >> This should be the definitive answer to the question what is a good >> formula and how accurate is Wheeler's formula. The best of Wheeler's >> last formulas are good for 0.1%, about 10 times better than the older >> formula. With the tweaks I showed, we're good for 0.02% for all values >> of D/L, so there shouldn't be any more doubt about what accuracy we >> can get with a simple formula. Notice that to get that accuracy one >> needs physical measurements good to about 4 digits; not easy to do. > > With all due respect to your Mathematica ramblings, I question your 0.1% > and 0.02% assertions. That's because all these Wheeler formulas, good > as they are, are missing a correction for wire size. So, except for the > impractical case of a large fine-wire coil, a 0.02% accuracy, etc., won't > be available to the users of these formulas. It's back to Grover for us.

With all due respect, you are quibbling and missing the larger point. I find it simply astounding what a skilled practitioner is now able to accomplish with today's affordable engineering tools (PC based Mathematica for one).

Can you imagine how long and carefully Grover and Wheeler must have labored to fill out tables and work through algebra? Another poster noted how small mistakes (such as in formulae or tables) would propa- gate from tome to tome because subsequent authors rarely had the time, energy or inclination to redo weeks, months (or sometimes years) of laborious hand calculations.

Now that has all changed. A few hours or maybe a few days to set up the problem, push a button, and the computer spits out a mistake free

*exact* symbolic answer in milliseconds! Who needs tables? (or even numeric solutions?) The paradigm has clearly evolved.

Once he is reasonably well practiced with the tool, the average (well maybe somewhat above average) engineer can now check the heretofore uncheckable. Even though it has been a standard that has passed under the gaze of countless engineers for well over fifty years, The Phantom's "ramblings" have managed to uncover errors in Grover and suggest improvements to Wheeler. Although I certainly admire The Phantom's prowess with Mathematica, what's really wonderful is that almost any reasonably sharp engineer could routinely unravel these knotty old problems - or, better still, do the same sorts of things with new problems. It is a new way of looking at how to get things done.

Don't worry if you don't know how, it's enough to know what "how" could do, and that you could learn how if need be.

I'll get off my soapbox now. :) -- analog

Reply to
analog
Loading thread data ...

I've been a big fan of Grover's book (and his earlier writings) since discovering them a decade ago, although I've not made much use of his massive laboriously-calculated tables. You make an excellent point, Grover's painful formulas are child's play for Mathematica, MatLab, Maple or MathCad, etc., any of which opens them up to routine use and could even be said to obsolete Wheeler's simple formulas.

BTW, here are two interesting posts, re Mathematica vs MatLab:

"I am a user of mathematica and I can imagine why novices/matlab find the syntaxes a bit of a letdown. Heres what i do all the time.

" (1) No need to remember function names All mathematica functions start with capital letters, so you always know if you want to integrate you need to begin with the letter "I". Next hold down the control key and press "k" (ctrl + k). All the functions starting with I appear right under your fingers! Keep typing (e.g: integr...) and you will see that the list keeps on filtering. Or you can just select the function from the drop down list you were looking for and hit enter. You don't even need to type the name of the function yourself.

" (2) No need to remember syntaxes After completing the the name of the function press "shift+control+K". The whole syntax will pop up! As a bonus dummy parameters are in place too! :surprise:

"Also, you can always type ?FunctionName to get help on a function (right inside your notebook).

"I don't remember when I last left a Mathematica notebook on which I am working to open the help and look for syntaxes (literally). Everything is presented to you under your cursor."

To which the response was,

"umm.. you dont need any of that stupidity in Matlab. in matlab you just press int and click tab twice, and it will give you all possible functions ... and if you dont know how to use it, type 'help int' and it will give you the syntax with examples."

--
 Thanks, 
    - Win
Reply to
Winfield Hill

All of which adds up the reason why I use Mathcad. I just type what I would normally write. I can place the equations wherever I want them on the page, and they look like real maths. Sure you can use shortcut keys, or you can take functions off floating menus. It even uses units sensibly - even if I get silly with them. For example, if I want the resonance frequency of a cartridge, I can put in stiffness in dynes/cm, mass in tons, run the calculation and it gives the answer in Hz.

d

Pearce Consulting

formatting link

Reply to
Don Pearce

Dear Robert,

thanks for looking, 'tis frustrating when you're near enough to smell the solution! I had a search around the SLAC website for some time but could not find a reference to such a paper.

Your description sounds very much like the ciruit with which I started this thread,

regards, Colin

Hello boys, I'm baack

Robert Baer wrote:

that which does not

transistor pair with reference current and gain

it

still be slow at low levels?

to look in my "bonded storage" for that

over a wide current range.

ago,, but this is what i remember about the scheme.

opamp was primarily for keeping the logging

volts, and the second one was to give

had its own compensation, and the resultant Bode plot (as i

something like this:

cannot vouch for the accuracy of this, but this may be useful.

about improved and/or wideband logamps, published by

Stanford Linear Accelerator Center.

ask that person to do a

someone that works there, but he is never available and

Pope is easier to communicate with.

Reply to
colinsmithwork

I'm looking at the circuit in your original post.

You say it is unstable. What freq is the oscillation?

Is there any significant C loading at Vout? Even a scope probe might be too much.

Mark

Reply to
Mark

Missed the original post...

The general problem with log amps is that the effective resistance of the emitter-base (1/gm) varies over orders of magnitude as the logged current changes. This moves poles around quite a bit and makes stabilization somewhat subtle. There's a good discussion of this stuff in Roberge's book (still one of the best, out of print for way too long) and also in the Analog Devices _Nonlinear Circuits Handbook_. I believe ADI's book is still in print - I should know this being an ADI employee, but I don't.

I'd have to look up the details myself as I haven't dealt with loggers in ages. Both of my books are at work, perhaps some kind soul will look this up and post something before I'm able to do so tomorrow night (East Coast USA time).

Steve

Reply to
Stephan Goldstein

Tim Wescott wrote

th

b

Doubling the phase would be adding 3 dB. Thi is more like quadrupling the phase noise

Does anyone have a simple math derivation o

how they get 20log(N)

more

10MHz
30MHz
a

So it's really just a mixer in this case. But it's neve really a 100% pure tone anyhow

S
Reply to
radio913

Well, with taking the output of the envelope detector and passing tha to a difference amp as described in my OP, I am amping ~150 time bringing the output of my amp to ~3-4 volts. It isn't very noisy and am able to sample and get the resolution I need. The sensor i actuall functioning perfectly. My only issue is in pratice, I won't have $4000 programmable PS available to null out the DC component at th difference amplifier

Being a digital guy, what I was thinking was to take the steady stat

out of the envelope detector, passing it through a A/D, and storing i a small micro. Then I could use a SW programmable pot or similar t "null" the effect of the DC component into the difference amp. Thi approach will work, but I figured I would investigate an analo equivelent that would be cheaper as well as take up less real estate

Dav

Reply to
Dave

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.