High Q smd capacitors

I'm far from an an expert, I just enjoy reading the crystal radio forum, (since 2010). >

There are plenty of double tuned circuits, Ithink this is fairly high end, >

Here's a URL with 78 well done crystal radios.

Mikek

There are two ways of broadening the pass band width while still maintaining good skirts in a double tuned circuit.

One is stagger tuning, i.e. the resonant circuits are tuned at slightly different frequencies.

The other is varying the coupling between to circuits tuned at the same frequency. With a very light coupling, the response becomes peaky. When increasing the coupling, the pass band flattens. With very tight coupling between the resonance circuits, you finally get a double hump response.

Here's video with a little info about the ring down counter. I suspect you would figure it out but I'll get you started, it counts oscillations after drive is removed, down to a certain percentage of the original amplitude. From the little I see, I think the input impedance (loading) during ring down it to low for a high Q measurement to be accurate. But the schematic is hard to read.

Don't watch more that 4 minutes.

Mikke

Oh. I was thinking hybrid ring, ring oscillator, Token Ring, tinnitus, Saturn's rings, etc.

The Q values shown on the LCD display were rather low. Q=550 was the highest I saw (at the end). He was using the same ferrite core, which would not have yield a very high Q. Air cores would have been more interesting.

You keep switching back and forth between unloaded Q and loaded Q. Unloaded Q is just the L-C part of the crystal radio. Loaded Q include the antenna, diode detector, and whatever else is connected to the L-C circuit. Unloaded Q is always larger than loaded Q. If you're measuring the unloaded Q with various instruments and techniques, you're not getting the operational parameters. For example, my estimate of the unloaded Q as being narrower than the modulation bandwidth is obviously not correct because it is possible to hear AM stations with reasonable fidelity. That's because the antenna, diode, etc are adding losses across the L-C circuit and reducing its Q. The lower Q results in wider receiver bandwidth which not passes the higher audio frequencies. I suggest that when you mention the Q of a circuit, you specify whether you're talking about unloaded or loaded Q.

Incidentally, unloaded Q is usually measured with test equipment, usually an RLC meter/bridge, or the aforementioned ring down detector. Loaded Q is measured with an RF signal/sweep generator and an RF voltmeter or spectrum analyzer looking at the -3dB points and specified as the RF/IF bandwidth.

I skimmed through the various crystal radio project pages at: There were far too many examples for me to skim them all. However, I noticed the absence of conventional AM SINAD sensitivity, IMD, dynamic range, modulation acceptance bandwidth, etc measurements. (1KHz tone,

30% modulation, 12dB SINAD, at >50% of rated audio output). These tests appear for every manner of tube, valve, and transistor receiver, by not crystal radios. I can see that there might be a problem matching the antenna input to the signal generator 50 ohm output, but that can always be done with a very lossy resistive pad. I found plenty of articles mumbling something about high sensitivity crystal radio, but no measurements. Am I missing something obvious here?

Thanks. I fell asleep a little after 6 minutes. It eventually became clear that he was testing various variable capacitors to see which offered the best Q. However, no numbers, results, or clues as to what he was testing. It might have been interesting if he measured all the variable caps at the same frequency and recorded the measured Q.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

There is very high Q ferrite that the enthusiasts are using and getting Qs in the 1200 to 1400 range. (BCB) And because I know who the author is, I'm pretty sure he is using that ferrite.

Well not really, I'm really only interested in unloaded Q. I'm talking about a tool to measure Q of a coil and maybe cap. If I were building it I wouldn't put a 500k resistor across my LC and then try to measure the Q of the LC.

Absolutely, but it is an important part and usually where the build starts. You won't have a high end crystal radio if you don't start with a high Q coil and cap.

I'm talking about measuring the Q of a coil, I kinda have to use either my experience or someone else's to inform me if an air cap is a good air cap or not. More difficult to measure but you can compare caps if you have a good coil and an instrument with a high input impedance. Looking at the Boonton 260A manual page 14, it shows about 15Mohms of input resistance at 1MHz.

Yes, that can be done, I don't see a lot of discussion about measuring the Q of a completed crystal radio system.

No, I don't think you are missing it, as I say I don't think much of that type work is done. Most people make a good LC, try to find a diode that matches the high impedance and then use an audio transformer to match a good set of sound powered phones (often old deck talker headphones). Then all goes to hell depending on signal strength. So then they switch in different diodes and have multiple taps on the audio transformer. Also variable coupling if using a double tuned system. And every thing is designed to minimize loss, note in the Ringdown video, he gets distance from lossy items by a layer of styrofoam. It seems we have been in this discussion before, but a real guru in this was Ben Tongue of Blonder-Tongue fame. He wrote many pages with lots of research into crystal radio design.

He has died and the look of the page was changed, but I think it's all there. Here's the original from the Wayback machine.

I like the old presentation better. Be forewarned, there is a month of information here.

I almost bitched to him about that same thing, but instead ask for the value of the input resistors. Mikek

From one of Ben Tongue's articles that you referenced: Table 2 shows that the ratio of loaded to unloaded Q (QI/Qo) varies from 0.1 to 0.6 depending on matching. Since the unloaded Q is always higher than the loaded Q, I would deduce from Table 2 that improvements in unloaded Q are not going to make much of an overall improvement. Methinks that experimentation with better antennas, diodes and impedance matching might yield better results.

Fine. So use a tapped spiral wound air core loop inductor as a combination antenna and L-C inductor and be done with the L-C part of the design. It would be difficult to do better than that for the L-C part. Increasing the unloaded Q, while seemingly ignoring the loading components, is not going to make any major improvements. Worse, if you actually succeed in making a major improvement in reducing the reduction in Q caused by the diode, you run the risk of ending up with a receive with too little bandwidth (as previously described).

The Boonton 260A maxes out at Q=625. The HP4342A was my Q meter of choice during the 1970's with a maximum Q=1000. I also worked with a Heathkit QM-1 Q meter which had a maximum Q=250. Today, we have the HP-4191 Impedance Analyzer. Maximum (calculated) Q=1000. I've found it difficult here to get perfectly repeatable measurements even using expensive laboratory equipment. For example I use a HP-4191A Impedance Analyzer. I still cross check inductors by placing them in a large copper-lined box and tune them to resonance with vacuum variables (Q>50,000). By measuring inductor current and voltage across the vacuum variable, or measuring the series or parallel RF resistance, I can determine Q or ESR. The highest Q I have found is around 1000 or just over 1000 when determining Q by using X / ESR = Q.

Sweep the receiver and measure the output voltage below and above the center frequency peak. When the voltage drops to 0.707 times the peak voltage, you have the -3dB power point. The difference in the two frequencies is the bandwidth. Divide the center frequency by the bandwidth and you have the receiver loaded Q.

I agree. The old web pages are better and easier to read.

Drivel: I find it rather amusing that some web designers try to emulate the look of newspaper columns. Such columns were originally invented because movable type was difficult to align in a straight line across the page. By cutting the width of the page into columns, the wandering lines of type were less visible. Like the typewriter keyboard, these problems were solved long ago, but we're stuck with the band-aids to this day. Sigh.

Perhaps I'm overly paranoid, but when I see important numbers and measurements missing, I tend to suspect that they're hiding something. Also, it's really difficult to determine if you're making progress without such measurements (or comparison to a known reference). How do you know that a receiver is "high sensitivity" if nobody bothers to make the measurements? I can't even find graphs for the audio frequency response or dynamic range (which I suspect sucks).

"Testing Crystal Radio Reception" The strength could be assessed with a sensitive voltmeter. Most stations will generate only millivolts of signal. You might try hooking a cheap digital voltmeter across the headphone connection.

Sigh...

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

Load it with 0.7 and you will get the same loaded Q and hence bandwidth.

The losses at 0.35 is about 5 dB, but at 0.7 maybe 9 dB, so the recovered audio would be down by 4 dB. Is this too much, it is up to you to decide.

A decent antenna is very important with passive receivers. A short vertical has nearly the same capture area as a full sized 1/4 wave vertical antenna. The nasty thing with electrically short antennas is that they have strong capacitive reactance and a radiation resistance of perhaps only a few ohms (insteead of nominal 37 ohms of a 1/4 sized antenna). The problem is getting all the captured power into yor LC circuit.

There are several tricks to do this. Adding a top hat dcapacitanse to your short vertical will help a lot. A T-antenna or Z-top hat capacitance are often used. A series loading coil can be used to tune out the antenna capacitive reactance. Now you have a resistive impedance of a few ohms. Feed the antenna to a tap very close to ground of your LC coil or use a few turn primary to perform the impedance matching. .

The other way of looking at short verticals is to talk about effective heights, which is close to 1/4 wavelength even for a short antenna. If the field strength [V/m] is known, then you can calculate the signal voltage into a full sized vertical. To calculate what the short antenna vill produce, you must include the impedance ratios.

And less audio signal. Audio is low enough, no need to start with a poor LC.

Oh, looks like you knew about less audio. :-)

First I've heard about verticals discussed with crystal radios. Can't add anything. Here's what Ben Tongue had to say about long wire characteristics.

Mikek

This is some kind of noise bridge. Usually a sensitive receiver in AM mode is used as null indicator. Broadband noise is sent to the antenna and R and C in the other leg. The R and C are adjusted until the bridge is in balance and the resistance and capacitive reactance Xc is read from the R and C settings. It is very handy e.g. when trimming a wire antenna for resonance at a specific frequency. Start with a slightly too long wire and measure resonance below allocated frequency, reduce length until you are at desired frequency. No need to transmit on frequencies allocated to other services during trimming of the antenna.

The antenna system measured in this link seems quite strange and the measured capacitance and resistance behaves also strangely.

For a short vertical, one would expect that the capacitance remains nearly constant regardless of frequency and hence the capacitive reactance is inversely proportional to frequency. The radiation resistance should increase with the square of frequency (until 1/4 wavelength) but a lossy grounding resistance in series will swamp some of this change.

When designing a transmitter system, one tries to get the expensive RF power from the transmitter and radiated into the space around the antenna. To get maximum power transfer, the source and load impedance should be the same. If they aren't you have to include some low loss matching tools between the transmitter and antenna (e.g. T or Pi-match).

For a passive receiver, it is as desirable to transfer maximum available power as in transmitting applications.

The maximum available power can be estimated if the field strength [V/m] is known for a specific frequency. The maximum antenna capture power is proportional to the capture area which inversely proportional to the square of frequency and hence relative to the square of wavelength.

f WL A P

0.5 600 43000 +16 1.0 300 10800 +10 1.5 200 4800 +7

For a small size dipole, the figures are a few dB less.

If the field strength is only one tenth (0.2 ,V/m) subtract 20 dB from those power levels calculated.

That is the theoretical maximum power available from a single antenna. Depending on various losses, only a fraction will reach the headphones.

With active amplifiers you can use ineffective antennas and badly mismatched antennas on LF, MF and lower HF due to the large band noise, so a significant signal loss is acceptable. On VHF, UHF and up you really have to power match to get all available power into the preamplifier.

By the way, is it cheating using a separate resonance circuit tuned to a local strong station, rectifying the signal and using the DC to power an emiter/source follower between the DX receiver resonant circuit and detector ? After all, it is still a passive receive.

First it's not a vertical, it''s a long wire and from what I gather noy all that long.

I don't think it's cheating. There are circuits online showing how people have done that. The next stop is putting big cap on the circuit and charging it all day, before using the energy for night time listening. That seem like a grey area, a very dark grey area. :-)

Mikek

From the discussion in this thread, I assumed that the idea was doing weak signal DXing.

A horizontal line installed at a low hight will have a radiation pattern pointing upwards due to the ground reflection. For this reason it is effective only for NVIES communication.up to few hundred kilometers. For longer distances, either the horizontal antenna must be much higher (in the order of a wavelength) or vertical polarization could be used.

Many think that in an inverted L- or T-antenna that the horizontal wire is the actual antenna, while the vertical part is just a feedline. In reality, the vertical "feedline" is responsible for the long distance vertical polarization, while the horizontal part just form a top hat capacitance (capacitive loading) making the vertical antenna less (capacitively) reactive.

On 7/30/2019 11:30 AM, snipped-for-privacy@downunder.com wrote:

I can make a few points and ask questions. Historically a long wire was used for crystal radios. And yes your right that a long wire is an inverted L antenna and the "feedline" is much of the antenna if not all and the long wire part is a top hat. To the best of my understanding at this point a short vertical is going to have a much high capacitive reactance than a long wire (short vertical with a long wire capacitive hat) Making it harder to match. When I think of verticals I think of something tuned to resonance and having close to 50 ohms (37 +) and no reactance. I read Jerry Sevicks book on short verticals. OK, I read parts of it. I never wanted to think this much! The T antenna often was built with multiple parallel wires to increase capacitance to ground to help with tuning. Reading more on Ben's page it seems a somewhat short and low, long wire antenna including ground resistance, and this is really only one antenna, looks like 300pf and 11 ohms at 1MHz. 53ft long 14ft high. I don't know how a vertical would compare and I assume the problem would be tuning losses. I lost my long wire when all the skyhooks were blown over by hurricane Michael. Also lost my BOG when the neighbor cleared his lot of everything. I'll get to play again after I finish building a new screened porch, also need a new roof on my shed, it's now 23 years old, metal roof. I realize I need to read for comprehension rather than just thinking I know what I want to tell you. You were right about the long wire being a vertical, but it really should have a top hat. Mikek

That's not hard to disprove. I had a long wire antenna with horizontal feed wire. It worked very well, thus the horizontal section does act as an antenna, not just top capacitance.

NT

Please do, I have read several times the feed is part of the antenna. Just how much?

I had a long wire antenna with horizontal feed wire. It worked very well, thus the horizontal section does act as an antenna, not just top capacitance.

I had one too, and it worked well, but how well compared to just a vertical. I don't know.

How does the pattern change over a vertical? Let's discuss a real antenna. Radio 4ft from the ground, antenna goes from 4ft up vertical 20ft then slopes up over 130ft to a height of 50ft. Ground resistance, I don't know, pick something realistic but has been nurtured to help lower it. Sandy loam soil.

Mikek

OH, I did a video when I had a 130ft long wire aiming West a

250ft BOG aiming North and a Mini Whip that is supposedly omni. I'm in the Florida panhandle and the station 1530 is in Cincinnati Ohio. I loved the BOG, that will be my first rebuild after all three were destroyed by Hurricane Michael, when I finally get to it. >

In this video I did more of a band scan comparing the antennas. I find it interesting to review after a year, you might too!

If we forget real long wire antennas such as Beverages that are wavelengths long and assume that we are talking about horizontal wires with one end hung high up in a tree and the receiver terminal inside the house, the antenna is actually a sloper with a height difference between wire end points. This absorbs the vertically polarized part of

any vertical component.

As far as I understand all LW/MW broadcast stations intended for local ground wave reception has always been vertically polarized. Typically a single mast with an elaborate grounding, such as 98 buried radials.

Especially on LW broadcast stations often two towers were used with one or multiple horizontal wires installed between them forming a top loading capacitance. A vertical radiator was then connected from the transmitter to the center of the horizontal wire(s) forming a T-antenna. Thanks to the top loading not so much loading coils and less grounding networks were required.

In Poland they even build a 600 b high center feed half wave LW dipole, which of course did not need a grounding network. Unfortunately it crashed during maintenance several years ago.

interesting comparison. To compare the upright versus horizontal sections of a non-beverage a start would be to receive using each section alone as well as combined: I don't have the setup to do that now.

NT

My best coil has Q=1250 at 500kHz, it peaks at Q=1500 at 800kHz, and drops to Q=850 at 1700kHz. Everyone seems very concerned about Q being to high. A builder of a high end crystal radio would be proud to have that problem, it is so easily fixed, where building to a spec of having that problem is difficult.

Mikek

I posted on a crystal radio grou[, here is a response I received.

"The conductive ground shorts out (and absorbs) any Horizontally polarised component close to the ground. So unless the antenna is fairly high (in wavelengths) there is little horizontal signal for it to respond to.

Plus the local AM transmitters will all be transmitting a Vertically polarised signal.

So a short "L" shaped antenna (close to the ground) will only pick up significant signals on its vertical section.

Adding the horizontal section adds top capacitance. Because the current distribution must start at zero at the end of the wire, top loading moves the current distribution upwards (in the vertical section). This means that the Radiation Resistance of a top loaded antenna will be higher, plus the extra Capacitance means that less Inductance will be needed to bring it to resonance. All this results in considerably greater efficiency, assuming a good ground.

The radiation pattern actually changes very little. The effect of the top section is to tilt the lobe slightly away from the horizontal section.

All this only applies when the antenna is relatively small compared with the wavelength. As the frequency rises, the radiation pattern changes dramatically.

The exception to this is the "Non-Resonant Travelling-Wave Antenna" (eg Beverage). A very long wire close to the the ground responds to Vertical signals traveling along its length, because the signals induce a current into the ground and into the antenna. But this pickup is negligible in the horizontal section of a short "L" antenna, as it is too short to respond to the traveling wave. However it is this "wave pickup" which tilts the lobe slightly.

P.S. People most definitely do use short vertical Antennas with Crystal sets. They work very well, but do require a very efficient Earth (or ground plane) and a very efficient loading coil. This usually means a remote antenna tuner and coax feed. The big advantage of an efficient Vertical Antenna is greater sensitivity to very low-angle incoming signals. eg international DX signals."

Here's an analysis of an L antenna. It's over my head but may work for you.

Mikek