Resonance
|
Hello Jim,
Thanks for your amusing comments. Since you have already apologized once
for doubting the 554 uh inductance value in the PL-380 7.5" external loopstick
project (as pasted below), I find it amusing that you are now rehashing the same
arguments for which you apologized :-)
Since you are obviously contradicting yourself, I'll let everyone decide
whether to take you seriously, or not. I have already made my decision, by the
way :-)
73, Gary
--- In ultralightdx@
wrote: > If my measurements (based on varactor > readings) and calculations are correct, the stock configuration > can resonate up to about 550uH at 1710 MHz (sound familiar? :). > With the clipped diode, about 700 uH may resonate. I stand corrected, and apologize to Gary for doubting him. But yesterday, he got a peak response at around 330 uH on an internal loopstick. I'm trying to understand this. His external antenna has to have more stray C, due to the longer leads. So either the proximity of the internal antenna to the circuit board is detuning it, or the Tecsun ferrite has peak Q at a frequency far from -61's. 73, Jim, KR1S http://qrp.kearman. In a message dated 1/23/2010 7:44:06 A.M. Pacific Standard Time,
jkearman@... writes:
|
|
|
|
jim_kr1s <jkearman@...>
Roy Dyball's notes on winding an antenna coil mentioned the need to re-tune the radio after each change in the number of turns on the antenna. As Roy mentioned, confirmed by my experiments, the Tecsuns (I have a G8 and a PL-380) don't like inductances higher than about 350 uH. What's that all about?
The antenna forms half of a tuned circuit, the other half being on the integrated circuit. Is resonance important? You betcha! It's been shown that inductances greater than 350 uH give improved signal strengths at frequencies below the broadcast band. The integrated circuit has limited tuning range, so what you gain at the bottom you necessarily give up at the top. The problem with removing turns until you just see an increase at the top is two-fold. First, you may stop before you actually achieve resonance. It isn't either-or. The integrated circuit has lots of amplification, and you may see stronger signals, but they could be even stronger if you keep going! The other problem comes when you don't have local stations at the high end of the band, or they are far enough away to vary in signal strength. That's why I built a shielded enclosure, into which I can inject signals of repeatable strength. Starting at about 500 uH and removing turns, I do see X-band signal strengths starting to improve at around 430 uH, but they improve further at lower inductances. Signal strengths at LW are reduced with lower inductance, so you have to decide what's more important. Tecsun made their PL-380 coils about 330 uH, which does a good job from 530-1710 kHz. Again: You can't have good LW and good high-band MW performance with one coil. Looking around for a lay-person's explanation of resonance and its effects, I found an old Rider publication from the early days of radio. It's in DJU format; you can get the free reader here: http://djvu.org/resources/ The file is here: http://www.tuberadio.it/download/rrider.djvu Like Roy, I'm not seeing evidence that using a spacer under the coil improves reception; test-bench measurements by Ben Tongue, a well-known engineer (co-founder of Blonder-Tongue) and crystal-set experimenter make me think it's not worth doing. See Table 2 here: http://www.bentongue.com/xtalset/29MxQFL/29MxQFL.html His best results across the band were with 1/16-inch polyethylene sheet, but not greatly different than with no spacer. Steve Ratzlaff, who has an expensive Hewlett Packard "Q" meter has also found no advantage to using a spacer. My own Q-measurement setup (from http://w7zoi.net/coilq.pdf -- Wes Hayward, W7ZOI was co-author of "Solid State Design" and "Experimental Methods in RF Design," and author of "Introduction to Radio Frequency Design," as well as many articles on related subjects) is not as accurate as Steve's meter, but nothing I've seen contradicts Tongue's and Steve's results. If you can get some 1/16-inch polyethylene, it will smooth out the Q (see the Rider book) across the band, and perk it up as much as 10-percent at the high end. What is critical is using an inductance the integrated circuit can resonate at the frequencies of interest. The tuning varactor has a design minimum capacitance, and there is stray capacitance in the circuit between the board and antenna (the connection between the rf board and the display board is a separate circuit). If the inductance is too large the varactor will be adjusted to its minimum capacitance, like a mechanical variable capacitor turned fully clockwise. You might still hear signals, just as you can still hear signals when a tunable loop antenna is mistuned. I can still hear Radio Enciclopedia on 530 kHz when my Terk is tuned to 1600 kHz, but the ladies are much stronger when the Terk is tuned to 530 kHz. Here's an article about varactors: http://en.wikipedia.org/wiki/Varactor While I cite others' work in this post, I speak only for myself. My results, which I feel are corroborated by others, are at variance with some other results reported here. You are free to make up your own mind. It is suggested that, if you have enough Litz wire and time, it might be worth experimenting with the coil size on your antenna modifications. When you can remove one or two turns without seeing any difference in high-band signal strength, you've hit the mark. This is hard to do accurately with off-air signals. Anyone who has a stable, accurate signal generator with an output meter, a big cardboard box, and some aluminum screen material can duplicate my tests. If you'd like to discuss the tests, email me at jkearman at att dot net. Thanks to Roy, Steve and other correspondents for sharing data! 73, Jim, KR1S http://qrp.kearman.com/ |
|
|
|
Hi Roy, Thanks for your comments, which are appreciated. I think everybody is in agreement that the loopstick coil inductance providing maximum AM sensitivity in the PL-300WT/ G8 models is typically from 300 uh- 350 uh, and always within the 180- 450 limits. Steve Ratzlaff and I did many AM alignments on these models for various DXers, although our procedure was to loosen up the loopstick coil form and slide the coil while listening for an audio signal peak on a weak station, instead of using the RSSI readings to determine a peak. Trying to find a signal strength peak by using RSSI readings (and/or S/N readings) alone is frustrating, since as Scott has confirmed there is a delay in the display response to sudden increases (or decreases) in signal strength. The audio signal peaking method used by Steve and me was successful in about 15 alignments (and is described at http://www.mediafire.com/?jyjg5hmkyqm ), since the coil position could be optimized for maximum AM sensitivity during actual weak signal audio reception. We always recorded the final coil inductances after the alignments, which usually were from mabout 300-350 uh (although there were a few oddballs outside of this range). When Tecsun and Kchibo switched to loopstick coils without movable forms (as in the D96L, PL-310 and PL-380 models), it was no longer possible to use this sliding-coil method to find the optimal coil position providing maximum AM sensitivity. Rather than accept all these DSP models as already having the maximum level of AM sensitivity thoughtfully provided by the Chinese factories (a ludicrous thought, for someone who has performed as many AM alignments as me), I decided to set up an actual dynamic testing method that would perform the exact same function as the sliding-coil tests in the PL-300WT/ G8 models, allowing the use of a sliding a coil to peak a weak AM audio signal under actual reception conditions. I turned to one of our ULR group innovations, the Slider coil, which provides a variable inductance by means of a special 40/44 Litz wire coil that can be easily shifted across a 7.5" Amidon ferrite bar. When transplanted into these newer DSP models in place of the stock loopstick, this Slider coil provides the exact same AM-sensitivity testing function as the stock sliding coils in the PL-300WT/ G8 models (that Steve and I aligned). When this 81-turn Slider coil is shifted along the ferrite bar after being transplanted into a PL-310 or PL-380 model, it can always quickly determine the actual coil inductance providing the maximum AM sensitivity in the model, by showing a sharp audio peak whenever the coil is close to the optimal position during reception of a weak AM station. This diagnostic test immediately corrects for any circuitry changes (from previous models) in the Si4734 radio, or in the DSP chip itself, since it is being performed under actual DXing conditions, with the best weak-signal AM reception being the bottom line. As such, it essentially creates a test where the radio tells you exactly which loopstick coil inductance it wants, to provide the best possible AM weak-signal reception. The sharp audio peak never lies, and is not confusing, as are the delayed RSSI and S/N readings. This was the diagnostic method by which the 554 uh optimal inductance was found for best AM sensitivity, in the PL-380 7.5" external loopstick project. The Slider coils can be somewhat intimidating to construct without experience (the process is described at http://www.mediafire.com/?nqggfm2jymc ), but they are routinely made here, and I would be happy to send one to you (complete with an Amidon type 61 ferrite bar) free of charge, if you wish to try out this method in your PL-380. Thanks for listening, and of course I respect any differences of opinion. Thanks also for your detailed descriptions of the Si4734 chip functions, which are greatly appreciated. 73 and Good DX, Gary DeBock (in Puyallup, WA USA)
When I first started experimenting with the original G8 loopstick moths ago I found it did not make much difference where I put the stock coil on the rod because as soon as I retuned the station the signal strength readings came back to about where they were before I moved it. I could not just simply slide the coil and look for a peak. Not then knowing how this little radio worked it made me want to find out what was happening. After obtaining a programming manual and seeing the range of commands available I was hooked and then spent hundreds of hours leaning how to talk to the Si4734 chip by building a USB interface and writing software to use it. In a message dated 1/23/2010 2:06:52 P.M. Pacific Standard Time,
roy.dyball@... writes:
|
|
|
|
Roy <roy.dyball@...>
Hi Jim Thank you for the links to the Rider article, I have just had a skim through but can see it is well written and I will enjoy digesting it. I first learnt about resonance 45 years ago during a 6 year traineeship in electronic telecommunications which I started straight out of school. Resonance fascinated me as much then as it does today (funny how the formulas haven't changed). I have experienced very similar results with loopstick coils with my Tecsun radios as you have noted. Yesterday I wound a large inductance sliding coil to check the upper limit of the loopstick coil in the G8 radio by using the Si4734 AM_TUNE_STATUS function to return the value of varactor capacitance being used to tune the loopstick coil at different inductances. The lowest value of capacitance the Si4734 can tune to (just like turning a mechanical variable capacitor to one end) is when the varactor is set to (decimal) 1 this represents a value of 0.095Pf, however SiLabs says we must add 7Pf to this value giving us 7.095Pf. The upper limit of the varactor in decimal is 6143 and that equals 590.5Pf. The chip in my G8 reported it was at 7.095Pf (rock bottom) with a value of 430µH adjusted on the loopstick. At around 350µH the varactor was reading around 12Pf which is where I found my best results. When I first started experimenting with the original G8 loopstick moths ago I found it did not make much difference where I put the stock coil on the rod because as soon as I retuned the station the signal strength readings came back to about where they were before I moved it. I could not just simply slide the coil and look for a peak. Not then knowing how this little radio worked it made me want to find out what was happening. After obtaining a programming manual and seeing the range of commands available I was hooked and then spent hundreds of hours leaning how to talk to the Si4734 chip by building a USB interface and writing software to use it. Now that we are all in love with our PL-380 PL-310 I recommend you put your old G8 to good use and convert it to a USB interface. You can do this in an afternoon just by removing three SMD resistors and attaching three wires plus earth. The USB interface PCB already assembled and only cost a little over twenty dollars. When it is finished you can use it as a tool for winding loopsticks and see for yourself what the radio is telling you (no guessing). The G8 380 310 have similar front ends in that the loopstick goes straight to SI4734 AM input via a coupling capacitor. The results you get on your G8 will be valid for the 310 and the 380. The project is on this site and it is free Chreers Roy. |
|
|
|
jim_kr1s <jkearman@...>
--- In ultralightdx@..., D1028Gary@... wrote: > > Hello Jim, > > Thanks for your amusing comments. Since you have already apologized once > for doubting the 554 uh inductance value in the PL-380 7.5" external > loopstick project (as pasted below), I find it amusing that you are now rehashing > the same arguments for which you apologized :-) > > Since you are obviously contradicting yourself, I'll let everyone decide > whether to take you seriously, or not. I have already made my decision, by > the way :-) > > 73, Gary Gary, I'm sorry you're making this about you. I hope that others besides you have the right to talk about ULR experiments, even when their findings differ from yours. I have gone out of my way to explain my equipment and methods, even providing online references where possible. I won't respond to your condescending sarcasm, but will continue to stick to provable theory and experimental results. For those who just tuned in, Gary's ire stems from my repeated attempts to convince him that the antenna circuit is tuned, which is critical to understand if you're winding antenna coils. I don't know if Gary accepts that, so let's get it out of the way right now. We heard it from Scott Willingham, an engineer at Silicon Labs, the chip manufacturer: "First, the AM front-end of the Si4734 is a tuned-tank circuit with a fairly conventional LNA and quadrature mixer as shown in the block diagram.... "As Roy has pointed out, the loopstick (or air-loop) inductance is resonated with an on-chip varactor, which tunes in small discrete steps. This tuning is done each time the frequency is changed." http://groups.yahoo.com/group/ultralightdx/message/6928 Now, on to tonight's episode: After Scott Willingham of Silicon Labs made his initial comment about the chip possibly tuning a 550 uH at 1700 kHz, which prompted my apology, he backed away from it. (I guess I should have then backed away from my apology, but I was trying to maintain the peace.) I quote (emphasis mine): "I will slightly amend what I said about the stock inductance range topping out at about 550 uH. My assumption there is that parasitic capacitance stays constant. It occurs to me now that as the inductance is increased, the loopstick's self-capacitance will increase as well, limiting the ultimate achievable inductance. On the other hand, at 10% too high inductance, you will only give up at the top 5% of the band. And if the Q at the top of the band is only 20..." http://groups.yahoo.com/group/ultralightdx/message/7345 The "top 5% of the band" is 1624-1710 kHz. I count 51 active stations in the U.S. and Canada from 1630-1700 kHz. I'm not willing to give them up. YMMV. In a private email to me, Steve Ratzlaff reported these Q values for a 562-uH coil wound on the center of a 7.5" -61 rod with foam-core spacer: 1300 kHz, Q 400 900 kHz Q 442 700 kHz Q 460 530 kHz Q 449 530 kHz Q 442 400 kHz Q 420 300 kHz Q 361 200 kHz Q 303 Steve was unable to measure above 1300 kHz, but you can discern the trend. Q is starting to decline at 900 kHz. Q declined about 10-percent from 900-1300 kHz. Extending the curve, we might expect Q=357 at 1700 kHz. I think Steve may have measured the Q of a Tecsun antenna and posted the results to this forum, but I don't feel like looking for them. Even if the rewound Tecsun ferrite's Q is half of Amidon's, and even if the varactor is low Q (the similar, discrete MVAM108 varactor has a specified minimum Q of 150. Data sheet: http://kearman.com/images/MVAM108.pdf), circuit Q probably exceeds 20 at 1710 kHz, especially when better-quality Litz wire replaces the original stuff. Therefore, my statement, "The problem with removing turns until you just see an increase at the top is two-fold. First, you may stop before you actually achieve resonance. It isn't either-or. The integrated circuit has lots of amplification, and you may see stronger signals, but they could be even stronger if you keep going!" is valid. If you find it amusing, perhaps you don't understand resonant circuits and Q. That Rider booklet may help. http://www.tuberadio.it/download/rrider.djvu Or dust off that ARRL Handbook I know you have on the shelf. NB, though, I never once mentioned "554 uH" in this thread. Since you brought it up, however, let me quote now from an email from Roy Dyball, from three days ago (January 20): "I wound that 81 [-turn] coil on the 7.5" rod and it needed 18 turns off it* before the varactor came off the bottom." * Removing 18 turns works out to about 338 uH. In other words, the chip was unable to resonate the 81-turn coil at the high end of the MW BC band. Roy's isn't an empirical, subjective observation; he's seeing varactor (tuning) capacitance reported by the chip. I, too, could not get an 81-turn coil on a 7.5-inch -61 rod to resonate. (Hint: There's too much stray capacitance (Scott Willingham calls it "parasitic capacitance," but we are referring to the same thing) in the radio, the wires going up to the rod, and in the coil winding.) You have previously stated that you have no nearby X-band stations to use for references, so given your experimental methods the error is understandable. At the time you didn't understand how the chip tuned the antenna, either, though I tried more than once to explain it. You may not take me seriously, but I took you seriously enough to try to help you understand something. According to John Bryant, that's the great thing about this forum. Until somebody disagrees with a founder. That I disagree with your methods and results is no reflection on you as a person, though I'm afraid that's how you perceive it. Last weekend you told me to repeat your experiments, and I did. Then I reported what I found. To ensure my observations were not skewed by fading signals I made a Faraday box, and used a stable signal source in lieu of off-air signals. Based on my results, some of which have been independently verified by Roy Dyball, I suggested people try removing more turns from the internal re-wound loopstick, and the consequences of boosting LW reception: a reduction in high-band sensitivity. Where's the harm in that suggestion? (BTW, your observation of increased LW sensitivity provided a clue, which you overlooked. The chip cannot tune the same inductance from 1710 kHz far down into the LW band. If it's perking up on LW, it has to be sucking wind higher up. As I mentioned at the beginning of this thread, just because you're hearing stations doesn't mean the antenna is resonated.) It is hard to discuss technical subjects without using formal language, lest we fall into the tar pits of ambiguity and imprecision. My regrets to those who find my writing too instructive or professorial. This isn't rocket science, but it is science, and I take it seriously. If you want to discuss radios, let's do it. If you want to attack me, let's take it off-line. 73, Jim Kearman, KR1S http://qrp.kearman.com/ |
|
|
|
jim_kr1s <jkearman@...>
--- In ultralightdx@..., "Roy" wrote: > > > Hi Jim > > Thank you for the links to the Rider article, I have just had a skim > through but can see it is well written and I will enjoy digesting it. Thanks, Roy! After the up-front theory, the old-time circuits are interesting, too. Radio hasn't changed all that much, though the devices we use sure have. I started out messing with radios similar to the circuits in that book, before I got my ham license. And I have the scars to prove it! > I first learnt about resonance 45 years ago during a 6 year traineeship > in electronic telecommunications which I started straight out of school. > Resonance fascinated me as much then as it does today (funny how the > formulas haven't changed). Got my ham license in 1962, FCC commercial license in 1964, and graduated from Hutchinson Technical HS, an engineering-college prep school in Buffalo, NY in 1965. Joined the USAF, where I was a ground-radio tech, but I bypassed tech school by taking a "bypass specialist's" test in Basic training. The ham training really paid off, especially for the AF, which didn't have to pay to train me! I worked on some interesting gear in some interesting places, and came to appreciate mil-spec quality. After the service I went for a liberal arts education, but always worked in tech environments, including 5 years for ARRL, the U.S. ham-radio organization, as a technical editor and author. I like writing, and I like the precision of science. I find it impossible to write about things I don't understand, but the basics always come through for me. I'm contemplating hacking the G8. Lately it's serving to look for parallel SW transmissions, but the soft mute's driving me nuts. The ULRs were a diversion from my home-made radios. Now that we've put the antenna business to rest, I plan to get back to them. 73, Jim, KR1S http://qrp.kearman.com/ |
|
|
|
Hi Scott,
Thanks for the link to the US Patent reference, and I'll certainly read it
over.
If I was a businessman I'd certainly be very thankful for this
"debate"-- demand for the 7.5" loopstick PL-380's seems to be going
through the roof :-)
73, Gary
In a message dated 1/23/2010 11:01:34 P.M. Pacific Standard Time,
sdwillingham@... writes:
|
|
|
|
John H. Bryant <bjohnorcas@...>
Friends,
toggle quoted message
Show quoted text
I think that Gary has hit on a wonderful idea..... Gary, as I understand it, Roy's set up is almost an exact match for the 380, as well, with the 310 being a bit of a strange duck, having longer leads and some other differences that the 300 and the 380 don't exhibit. ANYWAY, sending Roy a 554 Slider plus a Amidon bar would let him watch it operate with his software as well as try it in his 380. He should have a bare Amidon bar, already, so he can emulate your Slider if he wants and monkey around with the new one, leaving your original un-monkeyed with as the "control." If he doesn't have a spare Amidon bar, could you please send him two... and I insist on paying for the second one. Other than that, I'm making out a whole lot of Awards and getting ready to announce a new award category and graphic.... the graphic has been dubbed "Our Lady of the Globes" by Rob Ross and me.... she is barely dressed, but we are dismayed to report that the model was likely born about 1905 and the graphic first published as the cover of Citizen's Radio Callbook in their Spring 1928 issue. Gary, I really do want to pay for that second rod! (And all this assumes that Roy will play Ball.... erh Dyball.... ) John Bryant Stillwater, OK WinRadio G313e + Ultralights Wellbrook Phased Array The Slider coils can be somewhat intimidating to construct without experience (the process is described at http://www.mediafire.com/?nqggfm2jymc ), but they are routinely made here, and I would be happy to send one to you (complete with an Amidon type 61 ferrite bar) free of charge, if you wish to try out this method in PL-380. |
|
|
|
sdwillingham
--- In ultralightdx@..., "jim_kr1s" quotes me: > "First, the AM front-end of the Si4734 is a tuned-tank circuit with a > fairly conventional LNA and quadrature mixer as shown in the block > diagram.... > > "As Roy has pointed out, the loopstick (or air-loop) inductance is > resonated with an on-chip varactor, which tunes in small discrete steps. > This tuning is done each time the frequency is changed." Gentlemen, Radio reception and antennas involve complex, multidimensional tradeoffs, so I'm not taking sides (yet, anyway) in this debate. I do agree with Jim that a necessary first step to resolving the disagreement is to carefully understand loopstick resonance and how the Silabs' tuning method differs from that implemented in all other radios. For the techies in the group who are interested, I think this can answer most of your questions: US Patent 7561865 . Cheers, Scott |
|
|
|
Roy <roy.dyball@...>
Hi Gary Thank you Gary and John for the offer of the Slider. I think I have a better idea. As I already have a 7.5" Amidon rod and have wound and tested the coil in my PL-380 and G8 it would make more sense if you sent me one of your spare G8s. I will gladly modify it fitting the USB interface and testing it. I will return it to you with all the necessary software so it will work when unpacked. I would like to do this for you and John and the group as a free gift. You can then use this G8 as a tool when building your loopsticks. Why not just let the radio tell you the answers. I know that the thought of working on a radio with such small parts is daunting and nobody that I know off except Scott (with his own interface) has tried this. Perhaps by having one at hand in the USA more people may get to see that the handy man can work on this type of radio. I will email you with my address via Yahoo. All you have to do is pack the radio up and send it off. Cheers Roy. --- In ultralightdx@..., "John H. Bryant" wrote:
> > Friends, > > I think that Gary has hit on a wonderful idea..... > > Gary, as I understand it, Roy's set up is almost > an exact match for the 380, as well, with the 310 > being a bit of a strange duck, having longer > leads and some other differences that the 300 and > the 380 don't exhibit. ANYWAY, sending Roy a 554 > Slider plus a Amidon bar would let him watch it > operate with his software as well as try it in > his 380. He should have a bare Amidon bar, > already, so he can emulate your Slider if he > wants and monkey around with the new one, leaving > your original un-monkeyed with as the > "control." If he doesn't have a spare Amidon > bar, could you please send him two... and I > insist on paying for the second one. > > Other than that, I'm making out a whole lot of > Awards and getting ready to announce a new award > category and graphic.... the graphic has been > dubbed "Our Lady of the Globes" by Rob Ross and > me.... she is barely dressed, but we are dismayed > to report that the model was likely born about > 1905 and the graphic first published as the cover > of Citizen's Radio Callbook in their Spring 1928 issue. > > Gary, I really do want to pay for that second > rod! (And all this assumes that Roy will play Ball.... erh Dyball.... ) > > John Bryant > Stillwater, OK > WinRadio G313e + Ultralights > Wellbrook Phased Array > > > > >The Slider coils can be somewhat intimidating to > >construct without experience (the process is > >described > >at > >> >), but they are routinely made here, and I would > >be happy to send one to you (complete with an > >Amidon type 61 ferrite bar) free of charge, if > >you wish to try out this method in PL-380. > > > >Thanks for listening, and of course I respect > >any differences of opinion. Thanks also for your > >detailed descriptions of the Si4734 chip > >functions, which are greatly appreciated. > > > >73 and Good DX, > > > >Gary DeBock (in Puyallup, WA USA) > > > > > > > > > > > > > > > > > > > >When I first started experimenting with the > >original G8 loopstick moths ago I found it did > >not make much difference where I put the stock > >coil on the rod because as soon as I retuned the > >station the signal strength readings came back > >to about where they were before I moved it. I > >could not just simply slide the coil and look > >for a peak. Not then knowing how this little > >radio worked it made me want to find out what > >was happening. After obtaining a programming > >manual and seeing the range of commands > >available I was hooked and then spent hundreds > >of hours leaning how to talk to the Si4734 chip > >by building a USB interface and writing software to use it. > > > >In a message dated 1/23/2010 2:06:52 P.M. > >Pacific Standard Time, roy.dyball@... writes: > > > > > >Hi Jim > > > >Thank you for the links to the Rider article, I > >have just had a skim through but can see it is > >well written and I will enjoy digesting it. > > > >I first learnt about resonance 45 years ago > >during a 6 year traineeship in electronic > >telecommunications which I started straight out > >of school. Resonance fascinated me as much then > >as it does today (funny how the formulas haven't changed). > > > >I have experienced very similar results with > >loopstick coils with my Tecsun radios as you > >have noted. Yesterday I wound a large inductance > >sliding coil to check the upper limit of the > >loopstick coil in the G8 radio by using the > >Si4734 AM_TUNE_STATUS function to return the > >value of varactor capacitance being used to tune > >the loopstick coil at different inductances. > > > >The lowest value of capacitance the Si4734 can > >tune to (just like turning a mechanical variable > >capacitor to one end) is when the varactor is > >set to (decimal) 1 this represents a value of > >0.095Pf, however SiLabs says we must add 7Pf to this value giving us 7.095Pf. > > > >The upper limit of the varactor in decimal is 6143 and that equals 590.5Pf. > > > >The chip in my G8 reported it was at 7.095Pf > >(rock bottom) with a value of 430µH adjusted on > >the loopstick. At around 350µH the varactor was > >reading around 12Pf which is where I found my best results. > > > >When I first started experimenting with the > >original G8 loopstick moths ago I found it did > >not make much difference where I put the stock > >coil on the rod because as soon as I retuned the > >station the signal strength readings came back > >to about where they were before I moved it. I > >could not just simply slide the coil and look > >for a peak. Not then knowing how this little > >radio worked it made me want to find out what > >was happening. After obtaining a programming > >manual and seeing the range of commands > >available I was hooked and then spent hundreds > >of hours leaning how to talk to the Si4734 chip > >by building a USB interface and writing software to use it. > > > >Now that we are all in love with our PL-380 > >PL-310 I recommend you put your old G8 to good > >use and convert it to a USB interface. You can > >do this in an afternoon just by removing three > >SMD resistors and attaching three wires plus > >earth. The USB interface PCB already assembled > >and only cost a little over twenty dollars. When > >it is finished you can use it as a tool for > >winding loopsticks and see for yourself what the > >radio is telling you (no guessing). The G8 380 > >310 have similar front ends in that the > >loopstick goes straight to SI4734 AM input via a > >coupling capacitor. The results you get on your > >G8 will be valid for the 310 and the 380. The > >project is on this site and it is free > > > >Chreers Roy. > >--- In ultralightdx@..., "jim_kr1s" jkearman@ wrote: > > > > > > Roy Dyball's notes on winding an antenna coil mentioned the need to > > > re-tune the radio after each change in the number of turns on the > > > antenna. As Roy mentioned, confirmed by my experiments, the Tecsuns (I > > > have a G8 and a PL-380) don't like inductances higher than about 350 uH. > > > What's that all about? > > > > > > The antenna forms half of a tuned circuit, the other half being on the > > > integrated circuit. Is resonance important? You betcha! It's been shown > > > that inductances greater than 350 uH give improved signal strengths at > > > frequencies below the broadcast band. The integrated circuit has limited > > > tuning range, so what you gain at the bottom you necessarily give up at > > > the top. The problem with removing turns until you just see an increase > > > at the top is two-fold. First, you may stop before you actually achieve > > > resonance. It isn't either-or. The integrated circuit has lots of > > > amplification, and you may see stronger signals, but they could be even > > > stronger if you keep going! > > > > > > The other problem comes when you don't have local stations at the high > > > end of the band, or they are far enough away to vary in signal strength. > > > That's why I built a shielded enclosure, into which I can inject signals > > > of repeatable strength. Starting at about 500 uH and removing turns, I > > > do see X-band signal strengths starting to improve at around 430 uH, but > > > they improve further at lower inductances. Signal strengths at LW are > > > reduced with lower inductance, so you have to decide what's more > > > important. Tecsun made their PL-380 coils about 330 uH, which does a > > > good job from 530-1710 kHz. Again: You can't have good LW and good > > > high-band MW performance with one coil. > > > > > > Looking around for a lay-person's explanation of resonance and its > > > effects, I found an old Rider publication from the early days of radio. > > > It's in DJU format; you can get the free reader here: > > > http://djvu.org/resources/ > > > > > > The file is here: http://www.tuberadio.it/download/rrider.djvu > > > > > > Like Roy, I'm not seeing evidence that using a spacer under the coil > > > improves reception; test-bench measurements by Ben Tongue, a well-known > > > engineer (co-founder of Blonder-Tongue) and crystal-set experimenter > > > make me think it's not worth doing. See Table 2 here: > > > http://www.bentongue.com/xtalset/29MxQFL/29MxQFL.html His best results > > > across the band were with 1/16-inch polyethylene sheet, but not greatly > > > different than with no spacer. Steve Ratzlaff, who has an expensive > > > Hewlett Packard "Q" meter has also found no advantage to using a spacer. > > > My own Q-measurement setup (from http://w7zoi.net/coilq.pdf -- Wes > > > Hayward, W7ZOI was co-author of "Solid State Design" and "Experimental > > > Methods in RF Design," and author of "Introduction to Radio Frequency > > > Design," as well as many articles on related subjects) is not as > > > accurate as Steve's meter, but nothing I've seen contradicts Tongue's > > > and Steve's results. If you can get some 1/16-inch polyethylene, it will > > > smooth out the Q (see the Rider book) across the band, and perk it up as > > > much as 10-percent at the high end. > > > > > > What is critical is using an inductance the integrated circuit can > > > resonate at the frequencies of interest. The tuning varactor has a > > > design minimum capacitance, and there is stray capacitance in the > > > circuit between the board and antenna (the connection between the rf > > > board and the display board is a separate circuit). If the inductance is > > > too large the varactor will be adjusted to its minimum capacitance, like > > > a mechanical variable capacitor turned fully clockwise. You might still > > > hear signals, just as you can still hear signals when a tunable loop > > > antenna is mistuned. I can still hear Radio Enciclopedia on 530 kHz when > > > my Terk is tuned to 1600 kHz, but the ladies are much stronger when the > > > Terk is tuned to 530 kHz. Here's an article about varactors: > > > http://en.wikipedia.org/wiki/Varactor > > > > > > While I cite others' work in this post, I speak only for myself. My > > > results, which I feel are corroborated by others, are at variance with > > > some other results reported here. You are free to make up your own mind. > > > It is suggested that, if you have enough Litz wire and time, it might be > > > worth experimenting with the coil size on your antenna modifications. > > > When you can remove one or two turns without seeing any difference in > > > high-band signal strength, you've hit the mark. This is hard to do > > > accurately with off-air signals. Anyone who has a stable, accurate > > > signal generator with an output meter, a big cardboard box, and some > > > aluminum screen material can duplicate my tests. If you'd like to > > > discuss the tests, email me at jkearman at att dot net. Thanks to Roy, > > > Steve and other correspondents for sharing data! > > > > > > 73, > > > > > > Jim, KR1S > > > http://qrp.kearman.com/ > > > > > > > > |
|
|
|
Roy <roy.dyball@...>
Hi Scott Thank you also for the link to the patent. I realise you must be a busy man and the answers are all contained in the patent. It will take me a time to read and work through it. A lot of the group members will not make sense of it. Would you be able to provide a brief overview to help us all understand the process. Roy. --- In ultralightdx@..., D1028Gary@... wrote: > > Hi Scott, > > Thanks for the link to the US Patent reference, and I'll certainly read it > over. > > If I was a businessman I'd certainly be very thankful for this "debate"-- > demand for the 7.5" loopstick PL-380's seems to be going through the roof > :-) > > 73, Gary > > > In a message dated 1/23/2010 11:01:34 P.M. Pacific Standard Time, > sdwillingham@... writes: > > > > > --- In ultralightdx@ In ultra--- In ultralightdx@jkearman@ quotes > me: > > "First, the AM front-end of the Si4734 is a tuned-tank circuit with a > > fairly conventional LNA and quadrature mixer as shown in the block > > diagram.... > > > > "As Roy has pointed out, the loopstick (or air-loop) inductance is > > resonated with an on-chip varactor, which tunes in small discrete steps. > > This tuning is done each time the frequency is changed." > > > > > Gentlemen, > > > Radio reception and antennas involve complex, multidimensional tradeoffs, > so I'm not taking sides (yet, anyway) in this debate. I do agree with > Jim that a necessary first step to resolving the disagreement is to > carefully understand loopstick resonance and how the Silabs' tuning > method differs fro m that implemented in all other radios. > > > For the techies in the group who are interested, I think this can > answer most of your questions: _US Patent 7561865_ > (http://www.google.com/patents/about?id=zCLIAAAAEBAJ&dq=7,561,865) . > > > Cheers, > Scott > |
|
|
|
just_rtfm
layman's explanation (preliminary)....
http://home.comcast.net/~phils_radio_designs/index2.htm regards, phil :) |
|
|
|
jim_kr1s <jkearman@...>
--- In ultralightdx@..., "sdwillingham" wrote: > > For the techies in the group who are interested, I think this cananswer > most of your questions: US Patent 7561865 > . Thanks, Scott! I see in Fig. 3A that you're measuring the energy in the tuned circuit to determine resonance. In my Faraday-box experiments, I'm using the radio to measure the field strength from the probe antenna. When I have the right number of turns on the core (I turn off the radio between changes, and rock the knob one click off, then back on frequency before taking measurements just to be sure), I always see the highest signal level on the radio's display. (I can see from the display that sliding my hand under the screen and up to the radio doesn't affect the measurements, so I'm not inducing a hand-capacity artifact.) The test equipment requirement is minimal, as the radio does all the heavy lifting. It's a great device. I hope SiLabs can find markets to enlarge the feature set, to include SSB for one thing. Another useful feature for larger radios would be the ability to select either LW or MW inductances, so compromise inductances aren't required. I think you'll agree that you can't get best performance on both bands with a single coil. Next time you meet someone from Tecsun, could you ask them to add a button that completely disables soft mute? It has its place, but sometimes we wish you'd left out that feature. :) Regards, Jim Kearman, KR1S http://qrp.kearman.com/ |
|
|
|
sdwillingham
Phil,
Thanks for writing up your guide to the Si4734 receiver chip! I'm personally thrilled to see it on your site. Your "Phil's Portable Guide" and other articles were very influential to me when researching AM receivers while we were designing the Si4734. (History lesson: The Si472x and Si473x are part of our 2nd generation of radio chips following the Si4700, which is an FM-only receiver.) The description of AM antenna tuning, however, is not quite right. The method described (injecting a test signal and adjusting to maximize RSSI) is workable, but tends to be slow and have some practical difficulties. For instance, if the tuning is in error by a large amount, the RSSI gives little information for correction. That method can also be fooled by large received signals. So here's a layman's description of the tuning. The ferrite bar inductance and the varactor capacitance form a resonant circuit with reasonably large Q. An analogy to this resonator is the tuning fork, a mechanical resonator. When one "whacks" a tuning fork with a mallet, it rings for a while with a damped tone. The Si4734 whacks on its LC-tank circuit with a switch that forces current in the inductor. The tank rings for a short while -- at its resonant frequency -- and the chip measures that frequency using a counting method. Using iteration with changing capacitance values, the chip tunes the resonance to its target. Some other small comments: 1) On page 2, you describe our clock method as "radiometric". The term is actually "ratiometric", as in all the frequencies are related by integer ratios. 2) It would be helpful to your readers if you added citations for the documents you have used in compiling your guide. Regards, Scott |
|
|
|
just_rtfm
howdy Scott,
you're welcome and THANK YOU for the kind words! i appreciate your help: i'll read that patent and create a version 2. FWIW i used a patent sited for the Si4730, is that a documentation error? IMO Silicon Labs is making double-conversion/ceramic filters look as old-school as regens. low-IF with DSP is an elegant solution. are you still developing the chips? are you open to suggestions? btw, i am writing a companion PDF. do you know the approximate "Q" of the internal capacitance? regards, phil :) PS radiometric, that's pretty funny LOL. PPS guys/gals AFA the OT, and despite my error, we still have a tank. |
|
|