|
Vintage Test Gear and Workshop Equipment For discussions about vintage test gear and workshop equipment such as coil winders. |
|
Thread Tools |
9th Mar 2019, 2:47 pm | #121 |
Heptode
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
|
Re: Q-meter questions
Sounds music to my ears.
|
9th Mar 2019, 3:00 pm | #122 | |
Dekatron
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
|
Re: Q-meter questions
Quote:
"If you can't describe something with a simple model then you don't understand how it works" I'm not an expert in solenoids and David Knight and all the people mentioned in his paper would destroy me if I tried to talk physics with them. But the old solenoid model I use at work actually does something they appear to have failed at. It actually models the behaviour of the solenoid across frequency including the gaps between resonances. I can't read that paper any more, it is a long disjointed mess but I still think all the people in that paper are playing the same game. They are trying to come up with equations that pin the tail on the donkey (in terms of predicting where the resonances occur). They aren't trying to 'completely' model the solenoid across frequency, they just want to know where the resonances are and I think that is why their equations (and theories) become more and more elaborate. Maybe I'll write a paper on how I do it. It would take one page and it relies on basic physics to predict the discontinuities in the structure. It seems to work across all frequencies and the model can be used in a simulator and it works if the inductor is left floating with high impedances at each end, it also works if one or both ends are low impedance. At work I did have access to a megabucks 20GHz VNA with a time domain option and this did help a lot when it came to understanding what goes on as waves try to propagate through the structure. I did all this work on tiny coils for use on a PCB but I think it will scale up OK to larger coils.
__________________
Regards, Jeremy G0HZU Last edited by G0HZU_JMR; 9th Mar 2019 at 3:12 pm. |
|
9th Mar 2019, 8:16 pm | #123 | |
Heptode
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
|
Re: Q-meter questions
Quote:
Last edited by regenfreak; 9th Mar 2019 at 8:35 pm. |
|
9th Mar 2019, 9:20 pm | #124 |
Heptode
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
|
Re: Q-meter questions
Disclaimer: my last post was referring to the academia world in general and nothing about David Knight's paper since I am not qualified and I don't have the knowledge to comment.
I work in academia and there are lots of bullshiting technical research papers published so that people can justify their posts. |
10th Mar 2019, 2:02 am | #125 |
Dekatron
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
|
Re: Q-meter questions
There's an earlier version of David Knight's paper here:
http://www.tuks.nl/pdf/Reference_Mat...id%20coils.pdf It's obviously a 'work in progress' version but I think it flows better and is easier to digest for anyone fairly new to the subject. On the subject of common source JFETs I did try and find some info online that describes the input impedance up at RF frequencies. I ended up digging out something from the 1970s. It's a chart from an old application note and it does at least try and model the admittance and it then goes on to plot both the input capacitance Cp and the input resistance Rp (1/G) vs frequency. This is the bottom graph. It shows how the input Rp drops rapidly, even by 10MHz. This is for a JFET with a 1000 ohm load resistor in the drain rather than 5000 ohm so there is less voltage gain and therefore less Miller Effect (compared to the Cymar circuit) and so the graph will be slightly different to my earlier version. But even so, at 10MHz Rp drops to about 15k ohm with a 1000 ohm drain load. It looks like a classic JFET common source amplifier and if you showed the circuit below to lots of people I think the vast majority of them would not be able to predict that the input Rp would drop so fast. I think most people would probably say it would be defined by the 100k ohm resistor even at 10MHz because they have be taught that JFETs are (always?) high input impedance devices.
__________________
Regards, Jeremy G0HZU Last edited by G0HZU_JMR; 10th Mar 2019 at 2:08 am. |
10th Mar 2019, 3:39 pm | #126 |
Heptode
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
|
Re: Q-meter questions
Out of historical interest, I have scanned through the translation of P. Drude's paper 1902 "On the construction of Tesla transformers" translated by David Knight and Bob Weaver. It seems David Knight developed his work from Drude's study.
Drude also used glowing vacuum tubes. His experimental rig is like a bipolar tesla coil replacing the top loads by a HV variable cap submerged in paraffin oil. His transmission line rig reminds me of Tesla's hairpin. BTW Bob Weaver is the same expert who wrote the highly sophisticated Superhet tracking design software that we exchanged posts in another forum and eventually discovered a difficult bug within his program. PS I don't understand the concept of EM wave scattering in David Knight experiment. I only heard of radar scattering used by the facetted surfaces of a F117 shealth bomber.. Last edited by regenfreak; 10th Mar 2019 at 4:06 pm. |
10th Mar 2019, 6:54 pm | #127 |
Heptode
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
|
Re: Q-meter questions
David Knight wrote Note also that the term
'scattering' refers to the scattering of radiation by the coil. It does not refer to the measurement of S (scattering) parameters (although an S21 transmission measurement using the two antennas is a very good way of finding resonances). Here is a quote from a technical paper "Introduction to the Basic Technology of Stealth Aircraft " on the definition of scattering: (a) Specular wave scattering is essentially a reflection of the signal. The main contribution arises when the signal vector is perpendicular to the local surface. There may well be side lobes to the reflected signal due to local diffraction (b)Travelling Wave. A travelling wave may result when the radiation is more or less aligned along the length of a long, thin body. The scattering arises when the wave encounters surface discontinuities, the end of the body, or changes in the electromagnetic properties of the surface of the body, for example different materials. Diffraction. (c)Diffraction scattering occurs when there is a discontinuity in the target geometry, for example at the tip of a conical nose of an aircraft. d) Creeping Wave. Some regions of a body may welt be in the shadow of the radiated signal, but even so there may be creeping waves in these regions which result in scattering. These are associated with body currents and can result in unfavourable interference with specular waves. e) Electromagnetic Boundary Condition Changes. Scattering will occur wherever an incident wave which is propagating along the surface of a body reaches a discontinuity such as a sharp change in slope, different material, a gap or an edge. A radar has frequency of 3GHz which corresponds to a wavelength of 0.1m. This is smaller than the physical dimensions of the facetted surfaces of a F117. Anyway I give up reading David Knight's stuff and go back to my homebrew superhet project Last edited by regenfreak; 10th Mar 2019 at 7:03 pm. |
10th Mar 2019, 9:12 pm | #128 |
Nonode
Join Date: May 2007
Location: Redruth, Cornwall, UK.
Posts: 2,573
|
Re: Q-meter questions
Hi.
I've just come across another Q meter project in Electronics Australia magazine, November 1975 issue which may be of interest. https://www.americanradiohistory.com...EA-1975-11.pdf Regards, Symon |
10th Mar 2019, 9:40 pm | #129 |
Heptode
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
|
Re: Q-meter questions
Lastly this is off-topic but for the radio engineers..As I mentioned earlier about the scattering of radio waves around a solenoid. Below is a completely different EM wave scattering problem:
This is an obscure maths paper written by a Russian mathematician Ufimtsev (completely ignored by the Russians at the time) who worked on the scattering/diffraction of EM waves on 3D surfaces and then it was picked up by a Lockheed Martin engineer that ultimately lead to the development of stealth fighter F117. It was translated by the US Air Force> The stealth aircraft remained a highly classified project for many years. This is an unclassified doc from the USAF: https://apps.dtic.mil/dtic/tr/fulltext/u2/733203.pdf Last edited by regenfreak; 10th Mar 2019 at 9:57 pm. |
11th Mar 2019, 5:21 pm | #130 |
Diode
Join Date: Nov 2009
Location: Buenos Aires, Argentina
Posts: 4
|
Re: Q-meter questions
Hi guys, take a look at this home project, I have build one and it works very nicely:-
http://www.users.on.net/~endsodds/qm.htm Complete information for construction is given. I hope this info be of yours interest. Thanks, bye bye. |
11th Mar 2019, 5:52 pm | #131 |
Heptode
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
|
Re: Q-meter questions
Ok this is my last post.
For those who is interested in measuring the unkonwn stray capacitance of an inductor without a Q meter. This is a very simple method described in Terman p922 and Howe (1911): The Calibration of Wave-Meters for Radio- Telegraphy referenced by David Knight's paper. https://iopscience.iop.org/article/1...4/24/1/339/pdf Either (1) plot Plot 1/f2 versus C ---> the negative x-intercept is the distributed capacitance. or (2) plot C versus 1/f2 ---> the y-intercept is the distributed capacitance Note that the paper used old fashion wavelength instead of frequency in the graph This measurement requires the use of multiple high-Q silver mica caps of different values or a few silver plated variable air capacitors to get as many data points possible to obtain the best fit straight line. The accuracy of the stray capacitance value extrapolated from the intercept can be improved by the use of the least-square curve fitting technique. I have tried it and it is a very simple method. However if you have only a few data points in your graph and the stray capcitance of the inductor is small, the error can be large. The stray capcitance of an inductor largely depends on the form factor L/D ratio. |
11th Mar 2019, 6:11 pm | #132 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,864
|
Re: Q-meter questions
That's a *very* elegant method of measurement. It uses the simple idea that an inductor without a capacitor and without stray capacitance of any sort should resonate at infinite frequency. You can't measure with zero capacitance, but the intercept of a graph mapped to make lambda versusC linear is a suitable substitute.
Terman was an astute cookie David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
11th Mar 2019, 7:43 pm | #133 | |
Nonode
Join Date: May 2007
Location: Redruth, Cornwall, UK.
Posts: 2,573
|
Re: Q-meter questions
Quote:
Thanks for the link. The article is very interesting and the circuit looks well worth building. Regards, Symon. |
|
12th Mar 2019, 2:27 am | #134 | |
Dekatron
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
|
Re: Q-meter questions
Quote:
The capacitive divider needs to be made with low inductance caps (especially if you want to use it at >20MHz) but even then this driving method will be a bit of a compromise. So I'd expect to see some errors in places but nothing major. I'm not sure how well it will work up at 33MHz though. The design for the detector buffer looks to have had at least some attempts to deal with negative resistance. I've only had a quick look over it but it looks reasonable so far. Calibrating the dial might require some care because of the non linear detector response but I suppose this will be OK. By contrast, some of the earlier magazine/homebrew Q meters on here weren't fit to be published in a Ladybird "Q meters for Boys (and Girls)" book
__________________
Regards, Jeremy G0HZU |
|