The crystal radio is a time-honored build that sadly doesn’t get much traction anymore. Once a rite of passage for electronics hobbyists, the classic coil-on-an-oatmeal-carton and cat’s whisker design just isn’t that easy to pull off anymore, mainly because the BOM isn’t really something that you can just whistle up from DigiKey or Mouser.
Or is it? To push the crystal radio into the future a bit, [tsbrownie] tried to design a receiver around standard surface-mount inductors, and spoiler alert — it didn’t go so well. His starting point was a design using a hand-wound air-core coil, a germanium diode for a detector, and a variable capacitor that was probably scrapped from an old radio. The coil had three sections, so [tsbrownie] first estimated the inductance of each section and sourced some surface-mount inductors that were as close as possible to their values. This required putting standard value inductors in series and soldering taps into the correct places, but at best the SMD coil was only an approximation of the original air-core coil. Plugging the replacement coil into the crystal radio circuit was unsatisfying, to say the least. Only one AM station was heard, and then only barely. A few tweaks to the SMD coil improved the sensitivity of the receiver a bit, but still only brought in one very local station.
[tsbrownie] chalked up the failure to the lower efficiency of SMD inductors, but we’re not so sure about that. If memory serves, the windings in an SMD inductor are usually wrapped around a core that sits perpendicular to the PCB. If that’s true, then perhaps stacking the inductors rather than connecting them end-to-end would have worked better. We’d try that now if only we had one of those nice old variable caps. Still, hats off to [tsbrownie] for at least giving it a go.
Note: Right after we wrote this, a follow-up video popped up in our feed where [tsbrownie] tried exactly the modification we suggested, and it certainly improves performance, but in a weird way. The video is included below if you want to see the details.
Long wave lenght -> use big parts, coils with a big Q, wires not being too short, components not too close to each others.
Short wave lenght -> miniaturization welcomed, use small parts, short wires, but make sure components are being insulated by metal chambers if possible.
As a rule of thumb. :)
I was thinking the Q on SMD inductors must be abysmal for this application.
It is pretty typical that a crystal radio without amplifier would receive only local stations, even when constructed the old-fashioned way.
What exactly do you mean by local?
I played with crystal radios as a kid. I don’t go back so far as Oatmeal tubes and actual crystals but rather the Science Fair variety with ferrites, germanium diodes and spring clips.
Anyway.. I’m pretty sure I remember getting 2 or 3 dozen stations from my rural MI home. AM was already in decline then though and no doubt has declined a lot more since so I wouldn’t assume to know how many stations to expect today.
Of course having a good ground made a world of difference. You wouldn’t get a lot without that. Cold water pipes and/or the stopper on a rotary dial telephone were the places to find ground.
That is not the same circuit with SMT coils, a typical crystal radio has a single coil with multiple taps, which is not the same as multiple coils in series as they do not share the magnetic field.
Spider web coils are also an cool alternative to miniaturization fans.
They can be built easily from a beer mat or dummy plastic CD of a CD/DVD spindle.
And they’re flat, can have a higher Q than poorly made air coils etc.
They’re the 2D version of honey comb coils, so to say.
https://pe2bz.philpem.me.uk/Comm/-%20Receivers/-%20Crystal/Prj-010-/spider.htm
https://www.instructables.com/Spider-Web-Coil/
https://pe2bz.philpem.me.uk/Comm/-%20Receivers/-%20Crystal/-%20CrystalDesign/Coils/HoneyComb/honeyc.htm
Could these be made from FR4?
I would like to know this too! I did a quick search, but didn’t show up anyone who has tried it. There was a nice article about PCB coils, complete with a calculator https://coil32.net/pcb-coil.html but all the examples were for small spiral inductors. I remember many small FM “bug” transmitters using such PCB inductors when I was building kits back in the 1980s/90s
“not the same as multiple coils in series as they do not share the magnetic field…”
The inductors he used are not shielded, and in the end the way he stacked them ensured that they were in close proximity on a common axis. So the inductors DO share the same magnetic field. The assembly he ended up with was pretty much one coil with multiple taps – with one caveat…
It’s hard to be sure, but in the one he broke open it looked as though the coil might have been in two layers. If each coil does have two layers, then all bets are off regarding how much the final assembly had in common with a single tapped coil of the same physical dimensions and wire gauge.
Those SMD inductors probably have some core material to get any kind of inductance into a small space. Core material properties vary significantly depending on their composition. I.e., are those SMD parts any good at the frequency of interest?
these smd inductors are shielded types to minimize EMC as typically are used in switching regulators. You can try with RF wirewound SMD inductors, easily found on Mouser, Digikey, etc…
These inductors are not shielded and can perform better for this project…
The best inductors I have used for crystal radios are air loop antennas, about four to six loops, approx two feet in diameter, can be square instead of circular, on a non conductive form. Windings can either be cylindrical or spiral, the choice of which affects directionality. If you think of a spirally wound loop antenna as a pancake laying down flat, the highest sensitivity would point up and down from it. Cylindrically wound has best sensitivity radially. Space windings by a quarter of an inch to minimize internal capacitance to obtain a high Q.
The loop antenna is much more sensitive than a long wire for am radio reception and doesn’t pick up as much electrical noise. It picks up magnetic fields preferentially instead of electrical fields.
That air coil isn’t just three coils, it is a transformer. It depends on coupling between the coils as well as their inductance. As mentioned SMD coils are shielded, use cores which keep the magnetic field wrapped close around the windings, and are in general crap for this purpose. It would probably be possible to build a SMD scale transformer that would work, although it would still have poor Q, but at least would do the feedback functions correctly. This would require all windings to be on a single ferrite core. This would essentially be an even smaller version of the loop antennas which were universal in portable transistor radios of the late 1970’s / early 1980’s.
Exactly this. Somebody who doesn’t even understand how the circuit works swaps out some big parts with some smaller (and inferior) ones in a way that doesn’t work and then seems surprised. Only on HaD is a multiple fail a hack.
I thought that the title of this article mentioned something about “Fail of the week”… Let me check… Yep. “Fail of the week”, it says so. Not “Hack of the week”. Do you tend to say “bah, humbug” a lot? :)
I learned two things today.
The first is what the ‘Q factor’ of an inductor is. I did have to look it up.
The second is that I, just like [tsbrownie] didn’t realise that the inductors in a classic radio are magnetically coupled through the one shared ferrite core. And that is probably the reason why the initial configuration failed, and the second configuration was better (but still bad). The second configuration probably couples the magnetic fields better, but there’s still an air gap.
So, it was quite informative to me. If ever I wanted to make a radio, I would remember this.
The HaD article itself was a fail … it didn’t have any idea why the experiment failed. It just assumed that the Q of the coils wasn’t high enough. And you didn’t learn what you learned from the article … you learned it from the critiques of the article.
At this scale, uneven solder joints would impair the signal enough to prevent functionality (as shown in both videos). If orientation of the capacitors are enough to screw up the coil, then maybe the lack of quality control in the SMD components themselves may also provided wild fluctuations in use. Coils are just long wires wrapped into… a coil. There isn’t quality control to consider when it’s just solid core copper wire (you aren’t going to have a chunk of nickel randomly changing up the resistance) whereas those SMD caps are ceramic and while they may have tolerances, the tolerances will provide enough error to swing the entire wave into unreadable territory. I’m very surprised that he didn’t address any of this and is still wondering why it won’t work.
I think the Oscilloscope is a Zeeweii DSO3D12 https://www.elektormagazine.com/review/zeeweii-dso3d12-review, based on an image search.