EMRFD Message Archive 13001

Message Date From Subject
13001 2016-07-06 15:48:02 k1trb Question on Shorted Turns

In most antenna tuners there is a movable or switched tapped inductor (Z-match excepted).  

The tap is often grounded.  The end of the coil is also grounded.  For most positions this means 

there are shorted turns.


Isn't this a problem?


The shorted turns have some resistance.  The "working" part of the coil couples energy into

the shorted turns which the resistance dissipates.  


thanks

tom

k1trb

.


13002 2016-07-06 16:43:17 k1trb Re: Question on Shorted Turns
Sorry.

I found the discussion in January 2013, including the flash bangs.

Even McCoy and DeMaw show schematics with shorted turns.  There are no flashes
with QRP, but still ...

Garland, April, May, June, 2002 QST shorts turns and he's running real power.

Hmmm.

thanks
tom
.

13003 2016-07-06 18:12:03 K5ESS Re: Question on Shorted Turns

Tom,

I’ve wondered about the shorted turns as well.  Nearly all the old vacuum tube boatanchor transmitters used progressive shorting switches for the final band switch.  These types of switches are hard to find and expensive these days if you do find one.  I’ve wondered what the penalty for using a single switch that shorts out all the unused turns without shorting out the intermediate taps.  I suppose leaving the unused end open rather than shorted would potentially result in high RF voltage that might cause arcing problems.  Anyway, you mentioned the discussion in Jan, 2013.  Was this in the EMRFD forum or some publication?

Thanks & 73

Mike K5ESS

 

13004 2016-07-06 18:22:08 K5ESS Re: Question on Shorted Turns

Tom, I shoulda looked before asking.. I found the discussion of Jan. ‘13

 

13005 2016-07-06 23:43:25 mvs_sarma@ymail.c... Re: Question on Shorted Turns
we can remove the short and work it,
 but I fear even open circuited coil section.
 I compare it with a transformer with open secondary, supposedly causing mismatch. we should thank TOM for the brain teaser.
13006 2016-07-07 00:16:10 Ashhar Farhan Re: Question on Shorted Turns
Shorting of individual toroidal inductors is alright though. this is because there won't be any induced field in the toroids from those that are switched in. i have used this in my qrp version of the balanced balanced tuner.

13007 2016-07-08 22:10:17 Andy Re: Question on Shorted Turns
Shorted turns on coils have always bothered me.  I don't know why so many designs use it.  Doing it soaks up some power.  Maybe it's not so bad because the coils are almost always air wound so the coupling to the shorted section is low?  And so what if you lose a few watts?

Open turns are no problem!  (Except for the possibility of HV and arcing from the end.)

Andy


13008 2016-07-08 22:47:14 Bill Carver Re: Question on Shorted Turns
You answered your own question, Andy. With airwound coils the coupling between sections isn't high, so the current induced in the shorted sections isn't destructively high. And consider: if the shorted coil has very high Q, the loss due to that current will be low, and the main effect of the magnetic coupling will be a small reduction in the effective inductance of the unshorted part of the coil.

Open turns ARE a problem, especially if their inductance and distributed capacitance happen to be resonant at the band of operation, or maybe a harmonic of it. You develop very high voltages, arcs, and even hefty contacts of a model 86 Radio Switch in a QRO amplifier can melt. I speak from experience!

If the coil is a tapped toroid then you would NOT short unused turns on the same core. Because the coupling is much higher on a toroid, very high current would flow in the shorted section and its switch contacts and the effective inductance of the unshorted portion will be darned near zero. That doesn't mean you can't use toroids in multiband tanks, but each shortable section of inductance needs to be wound on its own core. With no magnetic coupling between cores you can short any unused toroid (the whole winding) without any effect, as Fahran has already said.

I've used -2 powdered iron toroids for 160m inductances, shorting it when on higher bands. And -6 powdered iron for 80m inductances (yes, -6 is higher Q/lower loss on 80m than -2), also shorted when on higher bands. The remaining bands can use a modest sized tapped airwound coil up to perhaps 10m, where an edgewound ribbon probably has lower losses. When size is an issue you could use a toroidal core for 40m or even higher, where -10 (black) powdered iron produces high Q, up to -17 material on 10m.

W7AAZ



13009 2016-07-09 09:50:12 Mvs Sarma Re: Question on Shorted Turns
I had a discussion with a senior friend of mine,
 the so called short doesn't really matter so much, he opioned as the winding has its own capacitance   across each turn, and the wire that shorts is ling enough from the coil tap to switch and back to one end of the coil.  I took the analogy of J-pole and slim-jim antennas of VHF, where there is a short (almost) across the cable termination.
but it moderates the impedance.



13011 2016-07-10 11:20:48 K5ESS Re: Question on Shorted Turns
Interesting and informative experiment. However, I’m somewhat mystified by the results of Qu measurement for the air core inductor for shorted vs open as I would have expected the difference in Q between the short and open circuit configurations to be much greater. So I did some calculations as shown below. The calculated inductance for the 12 turn coil came out very close to the measured value (<1%). I then calculated the inductance for an 8 turn coil assuming no additional four turns present either open or shorted. The calculated inductance value for the 8 turn coil (3.42 µH) turned out to be around 7.5% larger than the measured value of the 8 turns with the tacked on open four turns. I then calculated the series resistance of the 12 turn coil based on the measured Qu values resulting in 0.422 ohms at 5 MHz. and 0.648 ohms at 10 MHz and extrapolated these values to an 8 turn coil. Then calculating the Q for the 8 turn coil I got Q’s of 382 at 5 MHz and 497 at 10 MHz, considerably higher than the measured values with the open 4 turns attached. So, obviously the open 4 turns have a fairly significant impact on the remaining 8 turn coil although the mechanism is obscure to me. It’s somewhat intuitive to understand the loss mechanism from the circulating currents in the shorted coil but not so in the open circuit case. I’d be grateful if someone could shed some light on this.

Mike, K5ESS






Frequency

Coil Config.

Inductance (µH)

Qu

Rs (Ω)






Measured

Calculated

Measured

Calculated

Calculated or Extrapolated


5 MHz

12 Turns

6.11

6.15

455

-

0.422




8 Turns (SC)

2.97

-

240

-

-




8 Turns (OC)

3.18

3.42

246

382

0.281


10 MHz

12 Turns

6.11

6.15

592

-

0.648




8 Turns (SC)

2.97

-

336

-

-




8 Turns (OC)

3.18

3.42

345

497

0.432



.



13012 2016-07-10 11:29:30 K5ESS Re: Question on Shorted Turns
Table didn’t show up like it should have. I’ll try this,



Mike, K5ESS



13014 2016-07-10 12:01:39 nothdurftm Re: Question on Shorted Turns
Just uploaded the table to the files section.  Couldn't get it posted any other way.
13015 2016-07-13 10:25:37 davidpnewkirk Re: Question on Shorted Turns
W7AAZ wrote:

That doesn't mean you can't use toroids in multiband tanks, but each shortable section of inductance 
> needs to be wound on its own core. With no magnetic coupling between cores you can short 
> any unused toroid (the whole winding) without any effect, as Fahran has already said.

Obliquely related to this, a Hartley oscillator built with a resonator consisting of two separate ferrous-toroidal-cored inductors will hartle just fine. Magnetic coupling between the sections above and below the "tap" is not necessary.

A datapoint for any lurkers harboring that special case of The Irrational Fear of Toroids we might call The Weird Aversion to Using Toroids with Tubes: I had a 10-W-output tube-based transmitter--built in a candy tin, as is my way--upside down on the bench, and because my lab notebook looked lonely I temporarily replaced its pi-net inductor (4 uH on a T106-2 powdered-iron core) with a pill-bottle-wound (and therefore air-core) solenoidal equivalent wound to exhibit a 1:1 diameter-to-length ratio (the value reported to be more or less optimum for Q). As measured with the diode-peak-voltmeter technique described in EMRFD,  the sender's tuned-for-max outputs with the toroidal and solenoidal coils matched to within 0.05 V.

At W9BRD the only application for which I now routine use air-core solenoids--aside from my antenna tuner, which uses big plug-in coils with soldered-on taps for interaction with my ladder-line feeder--is that of LC-tuned-oscillator resonator. I learned the hard way that the magnetic field around non-toroidal power transformers all too easily may noticeably frequency-modulate nearby such oscillators by interacting directly with their ferromagnetic cores. (The non-varying field around that nearby station speaker is doing it, too, but you don't notice because _that_ magnetic biasing is "dc." As a fun experiment, set your fave non-PLLed ferro-cored-resonator LC oscillator running on the bench, tune it for a nice beat note in a receiver set up for CW or SSB reception, and then see how soon you can hear that beat note change as you slowly bring a strong refrigerator-tchotchke magnet toward your oscillator from a few feet away. In the EM equivalent of a mine, a ferrous-cored LC oscillator is the perfect canary.)

Best regards,

Dave
amateur radio W9BRD
 
13016 2016-07-15 08:39:51 swift_glen Re: Question on Shorted Turns
Dave,
Enjoy your sense of humor. Near as I can tell, only reference to "Hartle" is to a state of the universe before time, so that kills an oscillator on a particular frequency.
You mention (as does EMRFD) that a mutually-coupled tapped coil can be swapped with two un-coupled coils, and still oscillate in Hartley configuration. This should be obvious when you consider the alternate oscillator configuration "Colpitts" where two capacitors are driven at their junction. These two capacitors are equivalent to un-coupled coils of a Hartley.
Here's a mind-bender for y'all:
Split the inductor part of the resonator into two series parts like a Hartley, and ALSO split the capacitor part of the resonator into two series parts like a Colpitts. Make the ratio of reactance (upper/lower) the same for capacitor branch as for inductor branch. So you have equal part Hartley and equal part Colpitts. A Colpartley or Hartpitts ;-/
This one does interesting things, like amplify as an over-coupled double-humped dual resonator. But I don't think it wants to oscillate.
An oscillator seems to want feedback to either inductive branch, OR to capacitive branch, but not both in equal measure.
13020 2016-07-16 00:04:56 davidpnewkirk Re: Question on Shorted Turns
Swift_glen wrote:

> This one does interesting things, like amplify as an over-coupled double-humped dual resonator. 
> But I don't think it wants to oscillate.

A device with high-enough transconductance to winnow sufficient input negative resistance out of that melange might oscillate _somewhere_, probably not in a good way. :-) In effect you've built two parallel tuned circuits in series, a config used some vintage "AM/FM" broadcast receiver designs to make IF amplifiers capable of operating at 10.7 MHz and 455 kHz without switching. (The 10.7-MHz tank pretty much disappears at 455 kHz, and vice versa. Circuit simulator, here I come...)

Based as they are on a cathode/emitter/source follower, the classical Hartley and Colpitts oscillators not-so-hiddenly instruct us that the lowly follower may fill a lofty role. In the ham literature we sometimes see common-anode amplifiers used as buffers, but buffering is not really one of their functions, as there's direct C between grid/base/gate and cathode/emitter/source. In my regenerative-detector-at-IF-based superhets, I find that a detector oscillating at IF and coupled directly, however lightly, to the output of a mixer can "feel" tuning shifts _even at RF_ through the mixer devices--even a diode ring followed by resistive attenuation--and so after experimentation comparing triode cathode followers and grounded-grid amps have settled on using a grounded-grid triode, with resistive plate load, between my regenerative detectors and whatever precedes them. Sorry, solid-staters, but only the vacuum tube, and neither BJTs nor FETs with their control elements at common for ac, can provide at least some degree of true electrostatic I/O shielding in this role.

Some EMRFD correspondents may recall the BJT Hartley _crystal_ oscillator circuit used by Hayward and Lawson as a basis for the front-end-conversion and beat-frequency oscillators in their Progressive Communications Receiver of 1981. In that oscillator topology, a crystal in series with the emitter-to-resonator-tap path stabilized the oscillator frequency in the manner reminiscent of locked oscillators. In October 1989 _QST_'s "The QRP Three-Bander," Zack Lau, KH6CP (now W1VT), of the ARRL Lab developed for triband use an _untuned-coil_ version of the Hayward-Lawson crystal-oscillator circuit that used a tapped broadband transformer instead of a tuned circuit. In attempting to use the samne approach at 7 and 10 MHz, however, I found that the [seeming] untunedness of the "tank," in conjunction with the inverse gain v frequency characteristic of BJTs, to make the circuit prone to parasitic oscillation out of control of the crystal. Zack's implementation operated at 18 through 24 MHz, in which region the gain of the BJTs he used was lower, making the behavior of the circuit more predictable.

> So you have equal part Hartley and equal part Colpitts. A Colpartley or Hartpitts ;-/

Now you're channeling By Goodman, W1DX, dba as ham-radio satirist extraordinaire Larson E. Rapp, WIOU, in April 1956 _QST_'s "A Radical Approach to VFO Design." Based on his straw-man "finding" that Hartleys keyed for Morse chirp up in frequency and keyed Colpitts oscillators chirp down, Rapp proceeds to propose a combination topology--the Coltley or Harpitts--in which the opposing chirp polarities should cancel for astounding stability. He goes on to find that unfortunately that as a practical result of this cancellation effect the Coltley/Harpitts strongly resists changing its frequency when tuned, and then responds only after a time delay--a sort of "electronic backlash" that also results in keying waveshaping circuits having no effect on the circuit at all!

As I wrote in celebrating Rapp's opus in "75 Years of _QST_" (December 1990), "Every amateur owes something to Larson E. Rapp; few amateurs are sure of what."

Best regards,

Dave
amateur radio W9BRD