Driver to drive?

[Joerg]

On 2017-03-28 00:48, rickman wrote:

On 3/28/2017 3:18 AM, Clifford Heath wrote:
On 28/03/17 14:33, rickman wrote:
On 3/27/2017 7:58 PM, Joerg wrote:
On 2017-03-27 16:16, Clifford Heath wrote:
On 27/03/17 23:25, amdx wrote:
On 3/27/2017 6:57 AM, Tauno Voipio wrote:
Joerg, you seem not understand the innermost sense of the crystal-
radio people. They are closely related to audiophools, and it is
quite impossible to use sensible technical argumentation here.

I don't get that at all, can you back that up with any facts?

~1MHz, Q=500, bandwidth ~ 2KHz. Remind me why you need Q>1500?
More voltage does not create better audio.


That's what I was wondering all the time. In the ranges tested there
are
largely just AM stations, no CW or morse code. Even at 10kHz BW the
audio experience will not be very pleasing, it'll sound more like on a
telephone.

How much bandwidth is available on an AM radio station??? Do you
actually know much about crystal radios?

The carrier, and the sidebands, which spread either side by the
bandwidth of the audio being broadcast - perhaps 5 or 8KHz.
That means as soon as your Q exceeds 200 (@1MHz), you start to
reject some of the transmitted power.

Most of all it muffles higher audio frequencies in the sidebands which
makes the listening experience less than pleasant. This is why good
receivers like the ones in my lab have 6- or 8-pole crystal filters.

Determined by the Q of the resonant circuit in the radio, not the Q of
the coil.


Then there is the tempco. Someone opens a window and whoops the
resonant
frequency goes somewhere else.

Again, you are not speaking from knowledge. There are ham radio
antennas with exactly this sort of high Q and they manage to maintain
tuning during 100 watt transmissions.

As the old Romans said, hic Rhodus, hic salta.

Give us an example of such an antenna with a Q of 500 or more. Just one.
And yes, I have a ham radio license and decades of RF experience. My
clients pay me for that.

Why are you guys trying to discuss a topic you actually know little
about

You really know so little that you can't even conceive of how little you
know.

I can clearly see what you don't understand. Why are you being rude
about it?

<holds up mirror>

See?



--
Regards, Joerg

http://www.analogconsultants.com/

 

[amdx]

On 3/28/2017 5:58 AM, Clifford Heath wrote:

On 28/03/17 19:35, rickman wrote:
On 3/28/2017 4:07 AM, Clifford Heath wrote:
On 28/03/17 18:42, rickman wrote:
On 3/28/2017 3:15 AM, Clifford Heath wrote:
On 28/03/17 14:47, rickman wrote:
On 3/27/2017 12:22 PM, Joerg wrote:
On 2017-03-27 07:59, rickman wrote:
On 3/27/2017 10:46 AM, Joerg wrote:

The Q is over 500. It was claimed in this thread that PVC is not
suitable as coil winding material and this proves that it is quite
suitable. Which I already knew because I've used it in RF power
amps
decades ago.

Again, context. You are living in a different world. PVC is *not*
suitable in the crystal radio world because there are *much* better
materials. If PVC was the only material available it would be a
*great*
coil support. If mud was the only material available it would be a
*great* coil support.


Can you explain why one even want a Q in excess of 500 in the AM
band?

Q is a measure of the losses. Having a high Q in the coil means the
coil has lower losses. Having a higher Q in the tuning capacitor
means
it has lower losses. The lower the losses the more power that
ends up
in the headphones.

You have this quite wrong. A Q of 500 is a loss of 0.2%, and
1500 is a loss of only 0.067%. Increase your coil area by 0.13%
(a fraction of a millimeter of diameter) and you recover the
power you lost by "only" having a Q of 500, by gathering more
power to start with.

Furthermore, because the signal of interest spreads +-5Khz or so,
when you use a super high Q, you *reject* the parts of the signal
that are outside the passband. So yeah, you get more voltage for
for the low audio frequencies, but *less power overall*.

Duh. Doesn't sound like such a good deal now, does it?

You trimmed the part of my post that dealt with the bandwidth issue.
Why
don't you read that again?

Because you said 10KHz, when your bandwidth is nothing like that.
I just couldn't be bothered to call you on all your
lie^H^H^Halternative facts

Why are you here? You aren't interested in discussing anything. You
just want to make drama.

I'm sorry you're so confused by facts, but I'm not responsible for
fixing your innumeracy. I'll just get the popcorn and watch while you
chase unicorns.

I'm afraid you are missing some facts on crystal radio crystal
optimization. Here's research by one of the real gurus of crystal radio
optimization, Ben Tongue of Blonder Tongue fame.
> http://www.bentongue.com/xtalset/xtalset.html
There is a lot here so A quick look says article 22 and 24 may help,
but others have relevant information.
Mikek


---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

 

[amdx]

On 3/28/2017 4:51 AM, upsidedown@downunder.com wrote:

On Mon, 27 Mar 2017 23:47:00 -0400, rickman <gnuarm@gmail.com> wrote:

On 3/27/2017 12:22 PM, Joerg wrote:
On 2017-03-27 07:59, rickman wrote:
On 3/27/2017 10:46 AM, Joerg wrote:

The Q is over 500. It was claimed in this thread that PVC is not
suitable as coil winding material and this proves that it is quite
suitable. Which I already knew because I've used it in RF power amps
decades ago.

Again, context. You are living in a different world. PVC is *not*
suitable in the crystal radio world because there are *much* better
materials. If PVC was the only material available it would be a *great*
coil support. If mud was the only material available it would be a
*great* coil support.


Can you explain why one even want a Q in excess of 500 in the AM band?

That is the _unloaded_ Qu. Still 500 sounds excessive high for
ordinary LC circuits or loop antennas. At higher frequencies a silver
coated cavity resonator would easily ac hive that.

Q is a measure of the losses. Having a high Q in the coil means the
coil has lower losses. Having a higher Q in the tuning capacitor means
it has lower losses. The lower the losses the more power that ends up
in the headphones.

When you connect the load you get the _loaded_ Ql. This Ql determines
the actual receiver bandwidth. For AM reception a 10 kHz bandwidth
would be desirable, so at 1 MHz the loaded Ql=100 would be suitable.

Bell labs reported a bandwidth of 300Hz to 3000Hz is range required
for intelligible speech. When you slip your headphones on, all you need
is to identify the call letters of a distant station.
No one complains of excessive Q, it is so easy to spoil.
I have a coil on my Q meter now, adding a 10M resistor across a coil
with Q=1300 lowers it to Q=1100.

The Qu/Ql ratio determines the signal losses. a Qu/Ql=10:1 gives less
than 1 dB losses, a 2:1 ratio and the loss is a few decibels. A
500/100 ratio gives about 2 dB losses. In receivers with
amplification, this loss is directly added to the noise figure. For
this reason, it is undesirable to have too high loaded Ql in front of
the first amplifier stage. On the other hand, a wide front end can
easily overload the first amplifier or mixer. A high Ql filter between
the first amplifier and mixer helps a lot. This is an issue on VHF and
above.


The Q of the radio won't be the same as the Q of the components because
you are sucking off power to drive the headphones. Besides, the number
you came up with (10 kHz) would be perfect for AM radio if you can get
that, that is the channel spacing. Great selectivity.
It should be noted that BW = f / Q only gives the -3 dB point of the
response curve. There might be a 40-60 dB signal on the adjacent
channel, which will effectively kill the reception of the wanted
signal.

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

 

[Phil Hobbs]

On 03/28/2017 04:01 AM, rickman wrote:

On 3/28/2017 3:22 AM, Clifford Heath wrote:
On 28/03/17 13:18, amdx wrote:
On 3/27/2017 6:16 PM, Clifford Heath wrote:
On 27/03/17 23:25, amdx wrote:
On 3/27/2017 6:57 AM, Tauno Voipio wrote:
Joerg, you seem not understand the innermost sense of the crystal-
radio people. They are closely related to audiophools, and it is
quite impossible to use sensible technical argumentation here.

I don't get that at all, can you back that up with any facts?

~1MHz, Q=500, bandwidth ~ 2KHz. Remind me why you need Q>1500?

You haven't connected an antenna, and tried to drive a headset yet.

Umm, I think that was me in about 1971.

More voltage does not create better audio.

But if it is a very weak signal on the antenna, you don't want to
waste any signal in loss resistances.

You start to *reject* some of the received power as soon as the
Q passes 200ish, and you destroy the audio at the same time.

You can quadruple your received power by doubling your coil
diameter, or fitting a longer wire. That's *far* FAR more
effective than saving 0.2% by using higher Q.

With a crystal ear-piece, you still may have an impedance matching
problem. For that, you should use an audio transformer *after*
the detector.

Sorry to puncture your dogma.

You don't have any idea of what he wrote. Read it again and try to
understand what he is talking about before you dismiss it so (actually
not so) glibly.

It's funny to see so much emotion over somebody else's hobby circuit.
Building the ultimate crystal set is like making an oscillator that will
run on a 5-mV supply. It's not good for much, but it can be a lot of
fun if you have time on your hands and nothing to prove.

I can see the fun of trying to make the highest-Q coil you can, just as
I can see the beauty of shiny silver-plated air variable caps.

Think stamp collecting, not designing for 1e6 units.

Cheers

Phil Hobbs

(Who once built a completely impractical capacitive gauge that was
linear to 0.2% right down to where the plates touched. Nice polished
brass things they were, too.)

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

 

[amdx]

On 3/28/2017 2:22 AM, Clifford Heath wrote:

On 28/03/17 13:18, amdx wrote:
On 3/27/2017 6:16 PM, Clifford Heath wrote:
On 27/03/17 23:25, amdx wrote:
On 3/27/2017 6:57 AM, Tauno Voipio wrote:
Joerg, you seem not understand the innermost sense of the crystal-
radio people. They are closely related to audiophools, and it is
quite impossible to use sensible technical argumentation here.

I don't get that at all, can you back that up with any facts?

~1MHz, Q=500, bandwidth ~ 2KHz. Remind me why you need Q>1500?

You haven't connected an antenna, and tried to drive a headset yet.

Umm, I think that was me in about 1971.


More voltage does not create better audio.

But if it is a very weak signal on the antenna, you don't want to
waste any signal in loss resistances.

You start to *reject* some of the received power as soon as the
Q passes 200ish, and you destroy the audio at the same time.

You can quadruple your received power by doubling your coil
diameter, or fitting a longer wire. That's *far* FAR more
effective than saving 0.2% by using higher Q.

Crystal radios do not use the coil to collect RF energy from the air,
a long wire antenna of 25ft to 150 is attached tothe coil to pickup the
RF signal.
If you want to discuss loop antennas, start a new thread.

With a crystal ear-piece, you still may have an impedance matching
problem. For that, you should use an audio transformer *after*
the detector.

Sorry to puncture your dogma.

Take a breath!
I'll start simple, Coil and tuning cap have a unloaded Q of
1000. To extract maximum power you use a load that matches the
Q times Xl of the coil. Loading the coil with it's matched
impedance lowers the Q to 500. But before you do that match,
you need to insert a diode to demodulate the signal
The diode adds a series resistance into the circuit, that is
before the matching audio transformer.
Mikek


---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

 

[amdx]

On 3/28/2017 10:04 AM, Phil Hobbs wrote:

On 03/28/2017 04:01 AM, rickman wrote:
On 3/28/2017 3:22 AM, Clifford Heath wrote:
On 28/03/17 13:18, amdx wrote:
On 3/27/2017 6:16 PM, Clifford Heath wrote:
On 27/03/17 23:25, amdx wrote:
On 3/27/2017 6:57 AM, Tauno Voipio wrote:
Joerg, you seem not understand the innermost sense of the crystal-
radio people. They are closely related to audiophools, and it is
quite impossible to use sensible technical argumentation here.

I don't get that at all, can you back that up with any facts?

~1MHz, Q=500, bandwidth ~ 2KHz. Remind me why you need Q>1500?

You haven't connected an antenna, and tried to drive a headset yet.

Umm, I think that was me in about 1971.

More voltage does not create better audio.

But if it is a very weak signal on the antenna, you don't want to
waste any signal in loss resistances.

You start to *reject* some of the received power as soon as the
Q passes 200ish, and you destroy the audio at the same time.

You can quadruple your received power by doubling your coil
diameter, or fitting a longer wire. That's *far* FAR more
effective than saving 0.2% by using higher Q.

With a crystal ear-piece, you still may have an impedance matching
problem. For that, you should use an audio transformer *after*
the detector.

Sorry to puncture your dogma.

You don't have any idea of what he wrote. Read it again and try to
understand what he is talking about before you dismiss it so (actually
not so) glibly.


It's funny to see so much emotion over somebody else's hobby circuit.
Building the ultimate crystal set is like making an oscillator that will
run on a 5-mV supply. It's not good for much, but it can be a lot of
fun if you have time on your hands and nothing to prove.

I can see the fun of trying to make the highest-Q coil you can, just as
I can see the beauty of shiny silver-plated air variable caps.

The highest Q's I've seen are just over 2000, using 2 parallel 660/46
litz wires solenoid wound. The site that it was on is defunct and the
Wayback machine server for the link seems to be down.

However, I did find the link to a Ferrite coil with a Q over 2000.

> http://theradioboard.com/rb/viewtopic.php?t=7058

And, isn't it great that we have evolved to a point that food, water
and shelter are not concerns, and we can play at anything we want!

Mikek

Think stamp collecting, not designing for 1e6 units.

Cheers

Phil Hobbs

(Who once built a completely impractical capacitive gauge that was
linear to 0.2% right down to where the plates touched. Nice polished
brass things they were, too.)

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

 

[Joerg]

On 2017-03-28 07:57, amdx wrote:

On 3/28/2017 4:51 AM, upsidedown@downunder.com wrote:
On Mon, 27 Mar 2017 23:47:00 -0400, rickman <gnuarm@gmail.com> wrote:

On 3/27/2017 12:22 PM, Joerg wrote:
On 2017-03-27 07:59, rickman wrote:
On 3/27/2017 10:46 AM, Joerg wrote:

The Q is over 500. It was claimed in this thread that PVC is not
suitable as coil winding material and this proves that it is quite
suitable. Which I already knew because I've used it in RF power amps
decades ago.

Again, context. You are living in a different world. PVC is *not*
suitable in the crystal radio world because there are *much* better
materials. If PVC was the only material available it would be a
*great*
coil support. If mud was the only material available it would be a
*great* coil support.


Can you explain why one even want a Q in excess of 500 in the AM band?

That is the _unloaded_ Qu. Still 500 sounds excessive high for
ordinary LC circuits or loop antennas. At higher frequencies a silver
coated cavity resonator would easily ac hive that.

Q is a measure of the losses. Having a high Q in the coil means the
coil has lower losses. Having a higher Q in the tuning capacitor means
it has lower losses. The lower the losses the more power that ends up
in the headphones.

When you connect the load you get the _loaded_ Ql. This Ql determines
the actual receiver bandwidth. For AM reception a 10 kHz bandwidth
would be desirable, so at 1 MHz the loaded Ql=100 would be suitable.


Bell labs reported a bandwidth of 300Hz to 3000Hz is range required
for intelligible speech. When you slip your headphones on, all you need
is to identify the call letters of a distant station.
No one complains of excessive Q, it is so easy to spoil.

Here is a high-Q sound

https://www.youtube.com/watch?v=q_BU5hR9gXE

[...]

--
Regards, Joerg

http://www.analogconsultants.com/

 

[amdx]

On 3/28/2017 11:05 AM, Joerg wrote:

On 2017-03-28 07:57, amdx wrote:
On 3/28/2017 4:51 AM, upsidedown@downunder.com wrote:
On Mon, 27 Mar 2017 23:47:00 -0400, rickman <gnuarm@gmail.com> wrote:

On 3/27/2017 12:22 PM, Joerg wrote:
On 2017-03-27 07:59, rickman wrote:
On 3/27/2017 10:46 AM, Joerg wrote:

The Q is over 500. It was claimed in this thread that PVC is not
suitable as coil winding material and this proves that it is quite
suitable. Which I already knew because I've used it in RF power amps
decades ago.

Again, context. You are living in a different world. PVC is *not*
suitable in the crystal radio world because there are *much* better
materials. If PVC was the only material available it would be a
*great*
coil support. If mud was the only material available it would be a
*great* coil support.


Can you explain why one even want a Q in excess of 500 in the AM band?

That is the _unloaded_ Qu. Still 500 sounds excessive high for
ordinary LC circuits or loop antennas. At higher frequencies a silver
coated cavity resonator would easily ac hive that.

Q is a measure of the losses. Having a high Q in the coil means the
coil has lower losses. Having a higher Q in the tuning capacitor means
it has lower losses. The lower the losses the more power that ends up
in the headphones.

When you connect the load you get the _loaded_ Ql. This Ql determines
the actual receiver bandwidth. For AM reception a 10 kHz bandwidth
would be desirable, so at 1 MHz the loaded Ql=100 would be suitable.


Bell labs reported a bandwidth of 300Hz to 3000Hz is range required
for intelligible speech. When you slip your headphones on, all you need
is to identify the call letters of a distant station.
No one complains of excessive Q, it is so easy to spoil.


Here is a high-Q sound

https://www.youtube.com/watch?v=q_BU5hR9gXE

[...]

Yes, needs a Q spoiler!
Mikek

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

 

[Clifford Heath]

On 29/03/17 02:12, amdx wrote:

On 3/28/2017 2:22 AM, Clifford Heath wrote:
On 28/03/17 13:18, amdx wrote:
On 3/27/2017 6:16 PM, Clifford Heath wrote:
On 27/03/17 23:25, amdx wrote:
On 3/27/2017 6:57 AM, Tauno Voipio wrote:
Joerg, you seem not understand the innermost sense of the crystal-
radio people. They are closely related to audiophools, and it is
quite impossible to use sensible technical argumentation here.

I don't get that at all, can you back that up with any facts?

~1MHz, Q=500, bandwidth ~ 2KHz. Remind me why you need Q>1500?

You haven't connected an antenna, and tried to drive a headset yet.

Umm, I think that was me in about 1971.


More voltage does not create better audio.

But if it is a very weak signal on the antenna, you don't want to
waste any signal in loss resistances.

You start to *reject* some of the received power as soon as the
Q passes 200ish, and you destroy the audio at the same time.

You can quadruple your received power by doubling your coil
diameter, or fitting a longer wire. That's *far* FAR more
effective than saving 0.2% by using higher Q.

Crystal radios do not use the coil to collect RF energy from the air,
a long wire antenna of 25ft to 150 is attached tothe coil to pickup the
RF signal.
If you want to discuss loop antennas, start a new thread.


With a crystal ear-piece, you still may have an impedance matching
problem. For that, you should use an audio transformer *after*
the detector.

Sorry to puncture your dogma.

Take a breath!
I'll start simple, Coil and tuning cap have a unloaded Q of
1000. To extract maximum power you use a load that matches the
Q times Xl of the coil. Loading the coil with it's matched
impedance lowers the Q to 500.

You're making slightly more sense than Rickman, who hasn't
responded to *any* of the numbers I suggested, but one thing
still puzzles me.

Why worry about increasing Q above 500, when you can get more
than the extra 0.2% (lost energy at Q=500) by adding 3 inches
more antenna wire?

 

[Clifford Heath]

On 30/03/17 11:12, amdx wrote:

On 3/28/2017 6:18 PM, Clifford Heath wrote:
On 29/03/17 02:12, amdx wrote:
On 3/28/2017 2:22 AM, Clifford Heath wrote:
On 28/03/17 13:18, amdx wrote:
On 3/27/2017 6:16 PM, Clifford Heath wrote:
On 27/03/17 23:25, amdx wrote:
On 3/27/2017 6:57 AM, Tauno Voipio wrote:
Joerg, you seem not understand the innermost sense of the crystal-
radio people. They are closely related to audiophools, and it is
quite impossible to use sensible technical argumentation here.

I don't get that at all, can you back that up with any facts?

~1MHz, Q=500, bandwidth ~ 2KHz. Remind me why you need Q>1500?

You haven't connected an antenna, and tried to drive a headset yet.

Umm, I think that was me in about 1971.


More voltage does not create better audio.

But if it is a very weak signal on the antenna, you don't want to
waste any signal in loss resistances.

You start to *reject* some of the received power as soon as the
Q passes 200ish, and you destroy the audio at the same time.

You can quadruple your received power by doubling your coil
diameter, or fitting a longer wire. That's *far* FAR more
effective than saving 0.2% by using higher Q.

Crystal radios do not use the coil to collect RF energy from the air,
a long wire antenna of 25ft to 150 is attached tothe coil to pickup the
RF signal.
If you want to discuss loop antennas, start a new thread.


With a crystal ear-piece, you still may have an impedance matching
problem. For that, you should use an audio transformer *after*
the detector.

Sorry to puncture your dogma.

Take a breath!
I'll start simple, Coil and tuning cap have a unloaded Q of
1000. To extract maximum power you use a load that matches the
Q times Xl of the coil. Loading the coil with it's matched
impedance lowers the Q to 500.

You're making slightly more sense than Rickman, who hasn't
responded to *any* of the numbers I suggested, but one thing
still puzzles me.

Why worry about increasing Q above 500, when you can get more
than the extra 0.2% (lost energy at Q=500) by adding 3 inches
more antenna wire?

I don't know the math but logically I don't see that adding 3 inches
to a 50ft antenna is going to have the same effect as doubling Q

Doubling the Q doubles the voltage and halves the current, four
times the impedance. It creates NO EXTRA POWER, except the tiny
1/Q resistive losses that are eliminated.

For a Q of 500 the power loss (hence the maximum possible saving)
is only 0.2%. If you have a 40 foot long wire antenna, you can
gain 0.2% by lengthening it by ONE INCH.

Furthermore, raising Q to 1500 (say) merely restricts the amount
of side-band energy you gather in the first place. The amount of
loss depends on how much energy exists in the high-audio part of
the program material - but the point is that you're discarding
that power in the hope of getting *more* power, but all you're
doing is destroying your audio fidelity. Ben Tongue understands
all this, but it seems that you don't understand him.

If you needed a high impedance to match your phones, and if you can
tolerate the loss of high frequencies in the audio, you could use a
higher Q. But you'd be better to limit the RF so you get the full
audio bandwidth, then use an audio transformer to match your phones
instead.

Does that make my points clearer?

Clifford Heath.

 

[Clifford Heath]

On 30/03/17 12:21, Jeff Liebermann wrote:

On Wed, 29 Mar 2017 11:49:43 +1100, Clifford Heath
no.spam@please.net> wrote:

Doubling the Q doubles the voltage and halves the current, four
times the impedance. It creates NO EXTRA POWER, except the tiny
1/Q resistive losses that are eliminated.

Umm... I beg to differ.
Q = Xl / R = Inductive_reactance / Resistance_and_losses

I know that. I said that. It creates NO EXTRA POWER, and only
(from starting Q of 500) makes a TINY reduction in the net power
after losses - because the losses were so small to start with.

Now go back and read what I wrote, before you respond again,
because this response wasn't even relevant to what I actually wrote.

Clifford Heath.

Assuming the inductance remains the same, changing the style of the
inductor, such as a ferrite core, loop antenna, Litz wire, scramble
wound solenoid, or honeycomb basketweave, will not change the
inductive reactance. Only the resistive and other losses part
dissipates power. Any change in Q will involve the resistive part of
the puzzle. For example, switching from magnet wire to Litz wire
offers an increase in surface area. Since RF conduction occurs on the
wire surface (skin effect), and Litz wire has a larger surface area,
the Q will be higher. A receiver design that doesn't over-load the
coil, where the losses from the load across the coil is substantially
greater than the dissipative losses from the coil resistance and
surface area, should be able to benefit from the increase Q by
recovering more power. Whether this appears as an increase in voltage
or current depends on how the power is "tapped" from the coil.

https://en.wikipedia.org/wiki/Litz_wire
"The ratio of distributed inductance to distributed resistance
is increased, relative to a solid conductor, resulting in a higher
Q factor at these frequencies."

I guess I should mention that Litz wire works well up to about 1MHz.
At higher frequencies, the eddy currents in the inside windings form,
creating additional losses which negate the benefits of using Litz
wire.

 

[Clifford Heath]

On 30/03/17 12:01, Jeff Liebermann wrote:

On Wed, 29 Mar 2017 10:18:33 +1100, Clifford Heath
no.spam@please.net> wrote:

Why worry about increasing Q above 500, when you can get more
than the extra 0.2% (lost energy at Q=500) by adding 3 inches
more antenna wire?

Because the added 3 inches will add equal amounts of signal and noise,
resulting in no net improvement. It's always fun to drag things to
their logical extreme. Instead of 3 inches, add 3 miles of wire. The
recovered energy will be huge and possibly sufficient to draw an arc.
But, so will the received atmospheric noise be huge by the same
amount. Besides less loss, a higher Q also means more selectivity, so
the receiver won't act the like the first crystal set I built in
middle skool, which received and demodulated most of the AM broadcast
band simultaneously. There's an upper limit to Q as set by the
modulation bandwidth. The AM broadcast band runs about 10KHz per
channel, so the maximum usable loaded Q would be:
Q = 1,000 KHz / 10 KHz = 100
Any Q more than 100 at BCB frequencies will begin dropping the higher
audio frequencies from the modulation. Too narrow band, and all you
can hear will be CW (Morse Code). There's also the problem of tuning
stability, where anything that causes the high Q coil to drift in
frequency, will produce even harmonic distortion of the audio. Copper
wire is not particularly temperature stable and will require some form
of frequency or tuning stabilization.

Ok, back to taxes...

If you aren't going to read the thread for context, then just stick with
your taxes.

We're talking about increasing Q above where it starts to exclude
some of the signal frequencies. Of course there is also noise
excluded in that process, but that's unlikely to improve SNR.

 

[Clifford Heath]

On 30/03/17 12:40, rickman wrote:

On 3/29/2017 5:47 PM, upsidedown@downunder.com wrote:
On Wed, 29 Mar 2017 13:47:15 -0400, rickman <gnuarm@gmail.com> wrote:

On 3/28/2017 11:12 AM, amdx wrote:
On 3/28/2017 2:22 AM, Clifford Heath wrote:
On 28/03/17 13:18, amdx wrote:
On 3/27/2017 6:16 PM, Clifford Heath wrote:
On 27/03/17 23:25, amdx wrote:
On 3/27/2017 6:57 AM, Tauno Voipio wrote:
Joerg, you seem not understand the innermost sense of the crystal-
radio people. They are closely related to audiophools, and it is
quite impossible to use sensible technical argumentation here.

I don't get that at all, can you back that up with any facts?

~1MHz, Q=500, bandwidth ~ 2KHz. Remind me why you need Q>1500?

You haven't connected an antenna, and tried to drive a headset yet.

Umm, I think that was me in about 1971.


More voltage does not create better audio.

But if it is a very weak signal on the antenna, you don't want to
waste any signal in loss resistances.

You start to *reject* some of the received power as soon as the
Q passes 200ish, and you destroy the audio at the same time.

You can quadruple your received power by doubling your coil
diameter, or fitting a longer wire. That's *far* FAR more
effective than saving 0.2% by using higher Q.

Crystal radios do not use the coil to collect RF energy from the air,
a long wire antenna of 25ft to 150 is attached tothe coil to pickup the
RF signal.
If you want to discuss loop antennas, start a new thread.


With a crystal ear-piece, you still may have an impedance matching
problem. For that, you should use an audio transformer *after*
the detector.

Sorry to puncture your dogma.

Take a breath!
I'll start simple, Coil and tuning cap have a unloaded Q of
1000. To extract maximum power you use a load that matches the
Q times Xl of the coil. Loading the coil with it's matched
impedance lowers the Q to 500. But before you do that match,
you need to insert a diode to demodulate the signal
The diode adds a series resistance into the circuit, that is
before the matching audio transformer.

Somewhere in the process of being in this discussion I cam across a web
site (perhaps Kleijer's) that deeply analyzed the losses in the detector
diode. I we pretty impressed.

Save a few dB in the coil, a few dB in the tuning capacitor, a few dB in
the diode and use a very sensitive pair of headphones and you have a
*much* improved radio!


Why do you want to keep the receiver completely passive ?

A small solar cell o the headphone will charge a small battery, that
can be used to drive a small RF/IF/AF amplifier over the night.

How about a two circuit MF AM receiver, with one resonant circuit
tuned to your local strong MF (AM) station and the rectified DC power
is used to run the "DX" receiver ?

Why do you care what I do?

Because we don't have any convenient audiophools to ridicule,
but we can enjoy ridiculing radiophools.

 

[Tauno Voipio]

On 29.3.17 03:10, tabbypurr@gmail.com wrote:

I've not done an inductorless class D before. This will be under 1w of output, run off expensive battery power sometimes, and will be budget critical. The question is what frequency to pick.

What's the lowest frequency one could be run at? What percentage of random small transistors could operate adequately at that frequency?

I'm looking primarily for minimum cost & maximum transistor compatibility, and secondarily for low current consumption. A degree of sound quality compromise is ok, these will be running fairly small speakers.


NT

Like to invite EMC trouble?

--

-TV

 

On Wednesday, 29 March 2017 08:14:26 UTC+1, Tauno Voipio wrote:

On 29.3.17 03:10, tabbypurr wrote:
I've not done an inductorless class D before. This will be under 1w of output, run off expensive battery power sometimes, and will be budget critical. The question is what frequency to pick.

What's the lowest frequency one could be run at? What percentage of random small transistors could operate adequately at that frequency?

I'm looking primarily for minimum cost & maximum transistor compatibility, and secondarily for low current consumption. A degree of sound quality compromise is ok, these will be running fairly small speakers.

Like to invite EMC trouble?

It'll be in a tin can


NT

 

On Wednesday, 29 March 2017 02:58:59 UTC+1, bitrex wrote:

On 03/28/2017 09:15 PM, tabbypurr wrote:
On Wednesday, 29 March 2017 01:17:39 UTC+1, bitrex wrote:
On 03/28/2017 08:10 PM, tabbypurr wrote:

I've not done an inductorless class D before. This will be under 1w of output, run off expensive battery power sometimes, and will be budget critical. The question is what frequency to pick.

What's the lowest frequency one could be run at? What percentage of random small transistors could operate adequately at that frequency?

I'm looking primarily for minimum cost & maximum transistor compatibility, and secondarily for low current consumption. A degree of sound quality compromise is ok, these will be running fairly small speakers.

Former regular Jan did this several years ago:

https://groups.google.com/forum/#!topic/sci.electronics.design/jGykD0-rW9U

The ftp site seems to be locked down so you can't get at the source but
you could probably email him for it, looks like he GPL-ed it

Thanks. That type of approach isn't doable in this app though.


NT


It doesn't get much more dirt cheap than a 555 timer:

https://www.youtube.com/watch?v=HJhOOJ4qTOc

It doesn't even sound half-bad at the end of the video.

You could use a CMOS version with a driver transistor on the output to
reduce quiescent power consumption. Pulling "RESET" low should shut down
the output.

7400 series can be cheaper.


NT

 

[bitrex]

On 03/29/2017 07:50 AM, tabbypurr@gmail.com wrote:

It doesn't get much more dirt cheap than a 555 timer:

https://www.youtube.com/watch?v=HJhOOJ4qTOc

It doesn't even sound half-bad at the end of the video.

You could use a CMOS version with a driver transistor on the output to
reduce quiescent power consumption. Pulling "RESET" low should shut down
the output.

7400 series can be cheaper.


NT

Sorry, I don't follow. To make a class D amp (unless it's
self-oscillating, which doesn't really seem appropriate in this case)
you need at the very least a triangle/saw generator and voltage
comparator - how do you intend to do that with a single 7400-series IC?

The 555 is an analog IC.

 

On Wednesday, 29 March 2017 14:31:35 UTC+1, bitrex wrote:

On 03/29/2017 07:50 AM, tabbypurr wrote:

It doesn't get much more dirt cheap than a 555 timer:

https://www.youtube.com/watch?v=HJhOOJ4qTOc

It doesn't even sound half-bad at the end of the video.

You could use a CMOS version with a driver transistor on the output to
reduce quiescent power consumption. Pulling "RESET" low should shut down
the output.

7400 series can be cheaper.

Sorry, I don't follow. To make a class D amp (unless it's
self-oscillating, which doesn't really seem appropriate in this case)
you need at the very least a triangle/saw generator and voltage
comparator - how do you intend to do that with a single 7400-series IC?

The 555 is an analog IC.

Stage 1 produces a square wave, stage 2 turns it into a triangle wave. Stage 3 preamplifies the analogue input, and its output plus the triangle wave are fed into stage 4. Add output transistors & choke. A range of 74 series gates can be used, as long as they can operate as a quad or hex invertor without schmitt triggering.


NT

 

[rickman]

On 3/28/2017 10:23 AM, Joerg wrote:

On 2017-03-28 02:51, upsidedown@downunder.com wrote:
On Mon, 27 Mar 2017 23:47:00 -0400, rickman <gnuarm@gmail.com> wrote:

On 3/27/2017 12:22 PM, Joerg wrote:
On 2017-03-27 07:59, rickman wrote:
On 3/27/2017 10:46 AM, Joerg wrote:

The Q is over 500. It was claimed in this thread that PVC is not
suitable as coil winding material and this proves that it is quite
suitable. Which I already knew because I've used it in RF power amps
decades ago.

Again, context. You are living in a different world. PVC is *not*
suitable in the crystal radio world because there are *much* better
materials. If PVC was the only material available it would be a
*great*
coil support. If mud was the only material available it would be a
*great* coil support.


Can you explain why one even want a Q in excess of 500 in the AM band?

That is the _unloaded_ Qu. Still 500 sounds excessive high for
ordinary LC circuits or loop antennas. At higher frequencies a silver
coated cavity resonator would easily ac hive that.

Q is a measure of the losses. Having a high Q in the coil means the
coil has lower losses. Having a higher Q in the tuning capacitor means
it has lower losses. The lower the losses the more power that ends up
in the headphones.

When you connect the load you get the _loaded_ Ql. This Ql determines
the actual receiver bandwidth. For AM reception a 10 kHz bandwidth
would be desirable, so at 1 MHz the loaded Ql=100 would be suitable.


Exactly, not 500 or 1000. Still, even with a somewhat matching loaded Q
the audio quality is mediocre because a simple tuned circuit has a poor
shape factor. This results in higher audio frequencies being muffled,
speech to be less easy to understand and as you said further below a
strong station on the next AM channel will spoil the whole experience by
swamping the signal.


The Qu/Ql ratio determines the signal losses. a Qu/Ql=10:1 gives less
than 1 dB losses, a 2:1 ratio and the loss is a few decibels. A
500/100 ratio gives about 2 dB losses.


Which is close to impossible to even hear.


... In receivers with
amplification, this loss is directly added to the noise figure. For
this reason, it is undesirable to have too high loaded Ql in front of
the first amplifier stage. On the other hand, a wide front end can
easily overload the first amplifier or mixer. A high Ql filter between
the first amplifier and mixer helps a lot. This is an issue on VHF and
above.


The Q of the radio won't be the same as the Q of the components because
you are sucking off power to drive the headphones. Besides, the number
you came up with (10 kHz) would be perfect for AM radio if you can get
that, that is the channel spacing. Great selectivity.
It should be noted that BW = f / Q only gives the -3 dB point of the
response curve. There might be a 40-60 dB signal on the adjacent
channel, which will effectively kill the reception of the wanted
signal.


There is probably some sort of sports ambition where the guy with the
highest Q wins a trophy or at least a free beer

It *is* a hobby. No one builds a crystal radio set because he wants to
listen to talk radio.

You comments indicate you neither understand the hobby nor do you really
understand the technology, but I believe the latter is a choice on your
part. You seem to be deliberately refusing to try to understand what is
being said.

Your comment above about not being able to hear a few dB is a good
example of your mindset. Rather than trying to understand why anyone
would care if tweaking the core material for a coil adds a few dB, you
make a comment showing your disdain. Open your mind. Look at the
bigger picture. You tell me why anyone cares about the few dB from
optimizing this one feature? Someone could just as easily point to the
many things you are involved in an nitpick any one action. But that
would in no way diminish the activity as a whole.

--

Rick C

 

[rickman]

On 3/28/2017 12:05 PM, Joerg wrote:

On 2017-03-28 07:57, amdx wrote:
On 3/28/2017 4:51 AM, upsidedown@downunder.com wrote:
On Mon, 27 Mar 2017 23:47:00 -0400, rickman <gnuarm@gmail.com> wrote:

On 3/27/2017 12:22 PM, Joerg wrote:
On 2017-03-27 07:59, rickman wrote:
On 3/27/2017 10:46 AM, Joerg wrote:

The Q is over 500. It was claimed in this thread that PVC is not
suitable as coil winding material and this proves that it is quite
suitable. Which I already knew because I've used it in RF power amps
decades ago.

Again, context. You are living in a different world. PVC is *not*
suitable in the crystal radio world because there are *much* better
materials. If PVC was the only material available it would be a
*great*
coil support. If mud was the only material available it would be a
*great* coil support.


Can you explain why one even want a Q in excess of 500 in the AM band?

That is the _unloaded_ Qu. Still 500 sounds excessive high for
ordinary LC circuits or loop antennas. At higher frequencies a silver
coated cavity resonator would easily ac hive that.

Q is a measure of the losses. Having a high Q in the coil means the
coil has lower losses. Having a higher Q in the tuning capacitor means
it has lower losses. The lower the losses the more power that ends up
in the headphones.

When you connect the load you get the _loaded_ Ql. This Ql determines
the actual receiver bandwidth. For AM reception a 10 kHz bandwidth
would be desirable, so at 1 MHz the loaded Ql=100 would be suitable.


Bell labs reported a bandwidth of 300Hz to 3000Hz is range required
for intelligible speech. When you slip your headphones on, all you need
is to identify the call letters of a distant station.
No one complains of excessive Q, it is so easy to spoil.


Here is a high-Q sound

https://www.youtube.com/watch?v=q_BU5hR9gXE

You have eyes, yet you can not see.

--

Rick C