Questions and Comments

(from clearly troubled minds)




3 Sep 00  Things have been slow in the laundry room since the beginning of another school year.  About a month or so ago I picked up a pair of Utah magnetic phones, 150 ohm, and a pair of Murdock HS 16 phones, 125 ohm, also magnetic.  The latter were made in '59, and there appears to be a modest supply of them available.  Anyway, both were very sensitive,  and are good dx chasers when used with a matching transformer.

26 July 00  Diodes:  I was playing around with a little kit radio today, the one from Techtronics, just to check it's performance with a short antenna, and I noticed that not only did the crystal earplug not have a 47k resistor in parallel with it, but didn't need it, and that the diode was kind of funny looking.  I reached into my junk box, the part with carefully labeled and sorted parts, and pulled out some 1N60 diodes I had gotten from Dan's recently.  Yep, they were the same, and I had already tested them and found them lacking in sensitivity against my standards, including the ones I use for high level detection.  Okay, time to experiment.  I tried a variety of diodes in place of the 1N60, and guess what?  They all pretty much stunk... until I put in the 47k resistor, and they acted very nicely.  Even most of the 1N34A type diodes benefited from the resistor, but the '60 didn't really care one way or the other.  I suspect it has a fair amount of back leakage so the crystal earplug has a path to discharge between peaks.
    Racheting back:  one nice thing about my project radio is that I always have a tank circuit and an antenna tuner and lots of taps when I am trying something out, such as how my latest "student" radio will play in Peoria.  I finally realized that I had better have a "student" antenna up as well, just to ensure I wasn't overestimating the capabilities of the sets, so I took my alternate antenna, which had outlived its need, and cut it back to 50 feet of wire out the window - 18 feet high(roofline) and the rest over to a tree horizontally.  I even got it pointing north, where most of my big signals come from.  It still tunes normally with the antenna tuner, but it  doesn't grab the daytime stations quite like the older one, a little more than 100 feet long, did.   It suffers at night too, as was expected.  However, it will still pick up bona fide dx.  Next effort will be to add a parallel line or two to the horizontal part in flattop fashion and see how much difference that makes.

23 July 00  The more I try the less I know.  I have been playing around a bit with simple circuits in the last two weeks, and had some interesting revelations on the effect of connecting the antenna to the top of the tank coil.  Let's assume, for the sake of discussion, that a grounded tank circuit consisting of a parallel coil and capacitor, is a quarter wavelength long.  In that case, adding some 9 or 10 feet of wire to the top of the circuit would make it resonant, at the top of the BC band, about 100 kHz lower;  WRONG.   The effect is much more profound than that; it makes it resonant by about 3 times that expected.  Adding as little as 3 feet of wire to the top of the tank can lower the resonant frequency by about 100 kHz, or enough to tune you well off a station you had tuned.  That little capacitor you see with some set designs between the antenna and the top of the tank circuit is a necessity to keep from detuning the circuit.  If you are like me, you probably assumed it just made the antenna "work better".  If you actually know something about this business, you will talk about cancelling the inherent capacitance of a short antenna.  Many communications receivers have an antenna tuning control, which is simply a small variable capacitor to cancel antenna capacitance, to keep from affecting the tuning of the antenna circuit of a superheterodyne.  With crystal sets, the problem is more serious.  In the least respectful manner, one can get away with a very low  Q front end with a superhet ( and many do), since the serious work is done by the local oscillator, the mixer, and the IF filtering.  With crystal sets, you have to pay attention to the details.

11 July 00  Some thoughts on magic bullets and free lunches:  I have spent a lot of time messing around with simple radios this year, the latest being the Audion Receiver.   All work to one degree or another, and can be lots of fun as long as you understand their limitations and obey the laws of nature.  For example, more than one person has built a loop receiver of the "cigar box" variety, and suffered great frustration when it didn't work, or performed poorly.  Any number of xtal sets have been designed and advertised, and faithfully built by the unsuspecting newbie, who doesn't know (perhaps the designer didn't either), that it can turn out to be really neat or a complete bust depending on expectations and  how it is used.  Let's review some basic first order axioms: (1) Sensitive headphones hear more.  (2)A good antenna and ground are essential for optimum performance.  (3)Impedance matching and hi Q do make a difference; a big difference.(4) Failure to observe (3) can negate any benefits of  (1) and (2).  Lets take a look at an example of this.  Where do you connect the antenna to a simple set with a coil and a variable capacitor for the tank circuit?  With a very short antenna, on the order of 10 to 20 feet, you are using it more as a simple signal collector, and should best connect it to the end of the coil opposite the ground end, but will still need to interpose some capacitance between it and the set, on the order of a pF per foot of antenna.  As the antenna  gets longer, you collect more signal, but the antenna becomes a real part of the resonant circuit, and affects tuning even more drastically.  You now have  to start connecting the antenna to a tap on the coil - the longer the antenna, the closer the tap is to the ground end of the coil.  This gives you better impedance matching, and you are now working with what is, in effect, two tuned circuits, sharing the same coil.  You will reach a limit here, no matter how much you try, and should start looking at separate tuning for the antenna.  As an intermediate step, you can try a coupling coil, and your selectivity will increase.  There is a cost in complexity and parts as you get more performance.  Fortunately, it ain't that bad as soon as you learn some basics and dare to experiment.

7 July 00  Got a msg from Tom Rodriguez that Mouser no longer carried the 140/60 pF polyfilm variable capacitor I have finally come to know and love.  Said they no longer had a supplier - this happened a couple of years ago with the dual ganged 220/220 pF polyfilm variable they and others used to carry.  I checked the online catalog, and, sure enough, it no longer comes up.  Also checked Circuit Specialists site, and they don't have it either.  Electronix Express, Dan's, and the Goldmine still list it.  The same variable is used by Transtronics.  Radio Shack uses a single polyfilm capacitor of about 265 pF.  What's on the horizon for availability?  I don't know, but the news from Mouser bodes ill for those who build sets with kids.  Unless the bottom drops out of the low end transistor radio market, and all the little portable radios start using tuning diodes ( it could happen), I am confident that something will be out there.  On the other hand, we could become an even more elite niche hobby, using only vintage or homebrew capacitors.

27 June 00  Loopers revisited:  After playing around with the MK484 radio a bit, and noting how it liked to have a ground for better performance, I went back and reviewed what I had done with simple radios (see my 17 Feb 00 comments).  With one of the simple radios, I got something with just a single connection to a ground.  So, today I did a little more with the loop idea.  Hooking my good sound powered phones and detector to a tank circuit with a 5 inch diameter coil of #24 wire, about 265 uH, I found that even without a ground I could hear a few local stations, albeit weakly.  Then I connected a ground the the base of the tank circuit and the signals improved considerably, and I was able to hear a 50 kw station about 70 miles away (daytime).  Then I replaced the sp phones with a mouser xtal earplug, and still could hear something, but lost some loudness (actually, a lot).  Okay, time to play around.  I hooked up my little headphone amp and the stations came in much better.  Tried it with smaller coils, and the signals went down, down with the coil diameter.  Forget it without the amp on the smaller coils too.  A ferrite core coil was useless.  Guess the next thing to try is a larger coil made of some big wire.  The message is clear, however; a looper will work, particularly if it is grounded.  The headphone amp, just like I have on my crystal convertible makes it work even better, even though it isn't a crystal set anymore.  There ain't no free lunch, and the laws of physics still apply.  There is nothing magic about a loop crystal set, but in some cases you can get one working for you.  I still don't recommend it for a first time project, however, unless you live uncomfortably near several really strong locals, or just want to listen to the peanut whistle just up the road.

22 June 00  I finally got around to building a MK 484 IC radio.  This was a breadboard model, unlike the neat little package Phil Miller Tate sent me, and which you can see on his Yahoo! photo site.  Still, once I got started, it was a snap to build, worked as advertised, and now I can fiddle with it to prototype a "student " model.  See it HERE.(pix sometime, maybe).  See the pix on Phil's photo page now - go to my links.
    Had some fun chasing DX for a very informal Summer Solstice DX Party yesterday.  There is still some nice DX to be had, but not as much of it, and the band is a bit noisier.  Mike Tuggle still managed to get from Hawaii to the west coast, barely.

17 June 00  Getting the most out of a single-tuned rig; some thoughts and experiments.  I had much better success in the hobby class of the latest dx contest than I thought I would have.  My set was pretty basic, sort of; I used a main tuning coil 4 inches in diameter using close wound #20 plastic coated hookup wire.  Inductance was about 250 uH.  In parallel with it was a XSS 365 pF capacitor.  The detectors, connected to the top of the tank coil, were a parallel pair of Hewlett Packard 5082-2835 Schottky diodes at night, and a single 1N34A-type diode from Radio Shack in the daytime.  These led to a matching transformer a'la Ben Tongue, then to a second 4k:600 ohm transformer for my sound powered phones.  The antenna was about 160 feet of wire in the air, between 50 and 30 feet off the ground.  I used this in combo with a "Tuggle" antenna tuner for the Open class.  For the Hobby Class, I couldn't tune the antenna ( per the rules), so did the next best thing; I used a loading coil of about 35 turns of #24 hookup wire on a 5" form, inductively coupled to the tank coil.  The number of turns for the loading coil was arrived at experimentally, and got me down to the bottom of the band - it seemed to work acceptably at the higher frequencies as well.  The reason I mention this is that conventional design wisdom for antenna transfer/ loading coils used in simple sets is that the antenna coil should have about 1/6 as many turns as the tank coil;  following this philosophy, my coil would have had about 10 turns.  I have used tapped antenna coils before without a capacitor, and they worked pretty well; a variable capacitor added to make it an antenna tuner gives you sharper tuning, but a tapped coil works pretty well by itself.  However, my thinking now is that perhaps the number of turns for an untapped loading coil can be better selected than just using x number of turns.  I plan on working with this some more.

29 May 00  High impedance headphones and other myths:  From my earliest experiences with radios, I heard that the gold standard of headphones was 2000 ohm magnetic phones, and higher was even better - this is dc resistance by the way, so it is simple to check.  Getting back into the xtal set hobby after a long hiatus, using ceramic earplugs since that was all I had readily available, I yearned for a set of the 2k phones so I could get some "real" dx.  When I finally purchased a set, new, I even held off using them for a few weeks until I had a set I felt "worthy" of them.  You guessed it; they really stunk.  I had an old telephone (not so old really, it was just left unattended and unattached for too long to resist, so I liberated it), and the earpiece from it rivaled the ceramic plugs, but it only read 165 ohms of resistance.  Then Larry Pizzella tells me how he used top quality walkman type earphones of 16 or so ohms of resistance, feeding them from a small transformer, and the hunt was on.  My best set now is a pair of surplus sound powered phones, fed by a transformer (actually a couple) from the detector.  I also have several nice 2k and 4k phones around.  Here's my take on phones - dc resistance doesn't matter much; it is the sensitivity of the phones that does.  So why the mystique of the 2k phones?  Well,  I suspect it is because of a couple of things.  First is impedance matching between the resonant tank circuit of the set and the diode/phones.  The second is about getting enough signal to the detector.  The detector attached to  a good set of 2k phones can be tapped high enough on the tank coil to give an adequate match and also provide enough signal level to overcome the detector signal threshold, about 300 millivolts for the commonly used galena and germanium detectors.  The optimum tap point seems to be between 1/2 and 1/3 of the coil's length from the ground end of the tank circuit.  You can tap all the way to the top of the tank coil, and many set designs do this, but while you get a even higher signal level to the diode, your selectivity AND sensitivity are really slammed due to an impedance mismatch, and you end up hearing even fewer stations since the big guys dominate the dial.  As Ben Tongue and Al Klase have shown us, the thing to do is to get the most sensitive phones you can find, then use a transformer to match their impedance up to the top of the tank circuit.  Yes, transformers involve a little more complexity and cost.  However, decent ones with little loss can be had for anywhere from a couple of bucks and up (way up if you like).  Go to Ben's pages for what you can use and how to use them.  The balanced armature phones seem to be the most sensitive, and date back to the Baldwin headphones of the early part of the last century.  Sound powered phones use this design, and they can be found in a number of similar applications, including some telephone handsets.  Getting back to the 2k magnetic phones, I suspect that somewhere between 2k and 4k is the max practical limit for good sensitivity at a reasonable cost and which be successfully used without an expensive transformer.  As far as the ceramic (crystal) earplugs go, they still work fine, the price is right, and they are good for kid (both small and large) projects - don't knock 'em.  Oh yeah, my crappiest set of 2k magnetics work nicely with my one transistor amplifier.

16 May 00 So the y2k dx contest is over and the results announced.  I retained the Georgia State Championship, and felt pretty good about it.  However  , Mike Tuggle swung in from Hawaii, knocking off west coast stations at 2500+ miles and the price of poker just went up.  Just goes to show that location matters too, and Mike had a clear shot over the ocean.  But on the BC bands!  Unheard of.  The details of his set will trickle out over the next few months, I am sure, but for now think big basket coils using big Litz.  Another lesson hammered home is that the details do matter.  A dB here, a dB there, and soon it adds up to more stations heard.  In other words, don't sniff at marginal improvements.

    Let me put in a plug for the Xtal Set Society.  The web is great, and so are our on-line forums.  I see something nearly every day that puts new ideas into my kit bag.  However, digging through the archives for some of these nuggets is frustrating and time consuming.  If you want to record it for posterity, or at least past the time when Al Gore uninvents the internet, how about sending something to Rebecca for the newsletter?  A description and drawing of your favorite circuit is always welcome, and can be seen by the 700 subscribers that aren't wired in.  Any hobby needs a common meeting place backed up by some hard copy to survive, such as a magazine or newsletter, and in this niche hobby the XSS is it.  You can renew your subscription with your next submission.  I am working on a couple of things to send in myself, so if you don't want to relive my crappy drawings, beat me to it.  Twenty years from now, no one will remember your favorite URL, but the XSS literature will still be around.

    More thoughts on Q:  A popular myth in crystal sets is that selectivity is gained at the cost of sensitivity.  Let me be one of the first to say it ain't so.  There are several ways commonly used to improve selectivity in a set.  One is to tap the detector down on the tuning coil.  This gives you a better impedance match between the detector/phone circuit and the tank circuit, meaning better energy transfer.  It also keeps the Q of the tank circuit preserved a bit as well.  Lately some sharp minds have realized that transforming the detector circuit impedance up to that of the tank circuit not only keeps the efficiency of the tank circuit high while giving a good impedance match, but also bumps up the signal voltage to get it over the detector barrier.  Yes, tapping down on the coil is a good idea, to a point, and tapping down even lower seems to improve selectivity, but at a cost in overall signal levels.  In other words, there is an optimum tap point for impedance matching, if you have to do that, but tapping down even further isn't really doing much for you, particularly when you are trying to dig out the weak ones.  If you have a tank circuit with an inherently high Q, you also have minimized resistance losses in the circuit, meaning more signal to detect.  Using double tuning, that is, having an antenna tuner feeding the tuning circuit of the set, also improves both selectivity as well as sensitivity - the extra tuned circuit doesn't cost you.  First, your antenna is now most efficient at the frequency of interest, therefore you have more signal to work with.  Second, you are rejecting both man made and natural noise with the second tuned circuit, so your signal to noise ratio goes up.  As further evidence of this, consider how wave traps on strong locals allow you to nudge up closer to them to hear weaker stations.  When your set can separate stations in adjacent channels, you will start to appreciate  Q.   Even a simple coupling coil from an untuned antenna can help your selectivity; by controlling the coupling, you keep the Q of the tank circuit up, and have started the transition from a couple of stations heard to some real performance.  Bottom line:  good selectivity is essential to high sensitivity.

6 May 00  The winter dx season is over, so I broke out my superconvertible the other day to see it it still worked.  Using it as a regenny I picked up some daytime stations that I had never heard before.  It is a tricky little set to operate, being a two handed operation, but can really perform.  I also checked out the simple one transistor audio amp that is featured on the basic convertible, just to see how it stacked up from my earlier enthusiasm - it too gave me pleasant results, and will give a so-so set of headphones new life.
     More simple radios.  If you haven't tried it out yet, get one of the Radio Shack xtal sets and give it a try using my project page on it.  Better still, try out some of your own ideas.  Rick Moss sent me an email or two with some pix of his very modified  set.  Seems like every time I get back to basics I learn another trick or two.
    Cutting out the intruders:  In my neck of the woods, it seems, the hf broadcasters seem to sneak into your crystal set and make it tough sledding.  Traps for these guys seem to work a little, but let your guard down just a little and they come roaring back.  One technique that seems to work fairly well is to use a tuned antenna, but they still like to sneak in, suggesting that something else is at work here.  No, they don't come in without at least an antenna or ground attached to the set, which rules out the need for a Faraday shield around the whole set.  For the contest, I decided to try out Ben Tongue's scheme of impedance matching and using a high level detector tap, that is, tapping the detector to the top of the tank circuit instead of down the coil.  An unexpected benefit of this arrangement was that I heard no intruders.  My educated guess is that by tapping the detector down on the coil, needed when you don't have the matching transformer for Ben's scheme, you actually have two resonant circuits at work, the second being resonant above the BC band, and perhaps due in part to self resonance of the tank coil.  Better made coils with less self capacitance, and a corresponding higher self resonant frequency, may help cut out the intruders as well.  Incidentally, Al Klase has used the high impedance matching for the detector for some time now, and I was inspired to give it a try after he kicked my behind using it in last year's contest.
    I see on the web that there are some more simple crystal sets available on the market, including one that  looks like the old Radio Shack set.  If you buy one of these little sets, how about giving me a short review of it?  Every now and again something pops up that is better or worse, and it's nice to get the word out.  I only link vendors I have tried and liked or haven't tried but have no reason not to like.

17 Feb 00  Simple Radios:  I know I tend to rant about these on occasion, but there must be some reason or reasons they just won't go away.  .  You know the type I am thinking of; parallel tank circuit, usually with antenna and detector connected at one end, and ground at the other.  No taps on the coil, and either the capacitor or the inductor is variable as in the "rocket" radio.  Sometimes they are a "foxhole radio" or just a simple slider, without even the capacitor, but relying on coil capacitance to get resonance.  I had several of these in the distant past, and thought they were pretty neat....then.  I even posted a query to my panel of experts, and got back possible explanations for their being out there ranging from "technical incompetence", "corporate greed", and rip-off to "pinko commie plot".  All of these have some merit, but then I finally figured out what I think is the real reason they keep showing up:  they are not only simple but they work.   I went and made  one up tonight, hooking detector and a lead with an alligator clip to the top of the tank, and started touching metal, not even bothering to use a ground.  Regular antenna - worked.  Switch plate screw on wall - worked.  Antenna on cordless phone (used to use the metal finger stop on dial telephone)-worked.  Metal lamp-worked.  Also tried using just a ground lead - heard something, but when I hooked ground to bottom of tank and put finger on antenna slot, could make something out.  When I hooked up a decent antenna and ground, however, it still worked, but didn't tune as well until I went "back to the taps".  By working, I mean I got two or three stations.  This doesn't seem like much, but I don't live in the shadow of any powerhouse stations, and with the antenna alone got the three night time locals clearly.  Also, when you can imagine the technical ability and mental state of a 10 year old boy, a very simple radio without batteries that usually gets something no matter what is a pretty neat thing to have.  I remember attaching mine to fences, window screens, and anything metal about the house, mentally cataloging what worked, and what didn't.  Simple radios have a whole bunch of limitations, and don't really impress the totally jaded techie, etc. etc., but they work.  So, my possible recommendation for you out there when a 10 y.o. asks for a crystal radio, make up one of these, put an alligator clip on the end of a yard of antenna wire, and hand it over.  You can set the hook when you start to get questions about how to make it work better.  After all, it  works, and that's the first hurdle.

15 Feb 00  Headphones:   It seems to me that the limiting item for newcomers and would-be newcomers to crystal radio is finding good earphones.  The supply of quality high impedance earphones is pretty much limited to antique sets, and the price is going up.  I have a pair of new 2000 ohm magnetic phones from the XSS, but they are not very sensitive, and useless for going after dx in my neighborhood.  High impedance crystal earplugs, which I get from Mouser, are pretty good, and the price is right, but these are fairly delicate, and a little knocking around can put them out of commission.  Larry Pizzella uses high quality Sony walkman earbuds, feeding them with a matching transformer.    Scott Balderston of Scott's Crystal Radios rounds up the older headphones, and evaluates their sensitivity before he puts them up for sale at reasonable prices.  The Navy still uses sound powered phones for reliable communication on ships, and these work very well if you can find them.  The are still being made, and come on the surplus market from time to time.  Scott found a set for me, and they are the most sensitive phones on my wall.   They are only about 160 ohms dc resistance (600 ohms impedance), but a matching transformer works very nicely for them.  I tried a couple of 1000 ohm dynamic earplugs from Mouser, but, as expected, they were not even as good as the new magnetics from the XSS.   A good set of cans will last and last, and are a sound investment, so get the best you can find.  If you want to improve the results you get from the crystal earplugs, wear a set of sound attenuation earmuffs over them - you can get them from hardware stores and gun shops.  The will reduce the background noise level immensely.

5 Feb 00  Radio on a chip:  Sometime while I was "out", Plessey/Ferranti came out with the ZN414 and the ZN416 chips, which have three terminals, just like a transistor, but are crammed with the components for a tuned RF receiver with agc.  All you add is a small handful of parts for biasing, etc, and a tank circuit, battery, and phones, and you have a nice little receiver.  Just recently, I heard of the MK484 chip, which does the same thing ( the ZNs are out of production, I understand, but still available from time to time.  I think Mouser sells an SK replacement version of it , Mouser stock number 599-SK3246A, and would like someone out there to tell me if it is - 7 July 00 someone told me it wasn't).  The MK 484 is available in Australia and the UK.  Philip Miller Tate was kind enough to send me a working model of a radio using the 484, and is it ever hot.  No external antenna needed, and it really pulls them in.  Uses a tank circuit and 4 more parts; two capacitors and two resistors.  So, I went and ordered some of the chips, and will see how I fare in building one of these little sets myself.  Ought to make a nice school project, particularly since it doesn't need an antenna or a whole lot of patience to operate (two items in short supply for the active teen).

10 Feb 00 Science fairs and Xtal sets.  Do not plan on having a working set on display at the fair.  All the judges care about is the experiment and the results of it.  Sure, working displays are nice, but not needed, and if your experiment uses a crystal set, you will probably not be able to get it working satisfactorily at the fair.  Many are held under metal roofs; strikes one and two.  Getting an antenna and ground in place is also usually virtually impossible.  As a science teacher and sometime science fair judge, I am happy with a picture and  a schematic or circuit drawing along with the results.  Don't bother asking your kid's teacher about this; 99+% know nothing about crystal sets, and will blithely chirp how wonderful it would be for the set to be demonstrated at the fair.  Don't be suckered in by spirited ignorance.  Just in case your budding scientist gets interviewed by a judge, make sure they can say "crystal sets used solid state technology before vacuum tubes or transistors were invented" or something similar.

And earlier:  FM Xtal sets:  I have had a couple of queries on these, looking for designs, etc.  Anybody out there have any experience with these?  Send me a line and I'll put it out..  From what I've seen, they should work, perhaps with a a stage of amplification to help them along.  Just make sure you keep your leads short and cut your resonant circuit for the right frequencies.  One drawing I found of a VHF Receiver uses 4 turns of #16 wire on a 3/4 inch diameter form, and spaced to 1/2 inch,  in parallel with a 9 pf variable.  Diode taps off at 2 turns, then feeds a two transistor amplifier. This gives you about 50 MHz of spread starting at about 85 MHz, so tune slowly. Oh, up here, a resonant quarter wave antenna is less than a meter long, so invest in some type of gain antenna if you can, even if it is only a dipole.  Go to my links page and see the two schematics for FM xtal sets that Gollum has.

Here are a couple of links to discussion group comments on xtal sets that I found on the web; thought they were better listed here than on the links page:
 AMFMTVDX List reposts on Crystal Receivers
 GLOWBUGS V1 #17 via AB4EL Web Digests @ SunSITE -scroll down a couple of postings to get to the xtal set ones

HOME    LINKS CONSTRUCTION TIPS     ANTENNAS
RADIO SHACK PROJECT  PROJECT RADIO