Discussion:
Circuit Analysis with LTSpice: 12-V Lamp Flasher
(too old to reply)
Fred Bloggs
2024-02-09 12:02:26 UTC
Permalink
This is the new circuit analysis: exploring operation in Spice and looking at graphs.

https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
John Larkin
2024-02-09 15:24:43 UTC
Permalink
On Fri, 9 Feb 2024 04:02:26 -0800 (PST), Fred Bloggs
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Cursitor Doom
2024-02-09 16:25:31 UTC
Permalink
Post by John Larkin
On Fri, 9 Feb 2024 04:02:26 -0800 (PST), Fred Bloggs
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
John Larkin
2024-02-09 16:54:12 UTC
Permalink
Post by Cursitor Doom
Post by John Larkin
On Fri, 9 Feb 2024 04:02:26 -0800 (PST), Fred Bloggs
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
Jan Panteltje
2024-02-09 17:16:54 UTC
Permalink
On a sunny day (Fri, 09 Feb 2024 08:54:12 -0800) it happened John Larkin
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
On Fri, 9 Feb 2024 04:02:26 -0800 (PST), Fred Bloggs
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?

And no guarantee that anything in LTspice works -in- or even resembles reality!

Real hands on testing is needed to get a clue.
john larkin
2024-02-09 18:17:15 UTC
Permalink
Post by Jan Panteltje
On a sunny day (Fri, 09 Feb 2024 08:54:12 -0800) it happened John Larkin
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
On Fri, 9 Feb 2024 04:02:26 -0800 (PST), Fred Bloggs
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Almost all our products were simulated in LT Spice, and that steered
or verified the designs.
Post by Jan Panteltje
Real hands on testing is needed to get a clue.
We don't prototype many products; they are too complex. It's more
efficient to design, simulate, review, and manufacture. If the
engineering is done right, you can sell rev A. I nkow of one giant
organization that formally plans for six PCB spins before production.
Why bother to be careful on the first few?

We do breadboard little subcircuits if we're not sure how they will
behave. Data sheets can be vague. Dremeling is fun.

Hers's a product sim.

Loading Image...

I did this all analog, in LT Spice, to get the dynamics right. Most of
this will actually be done in an FPGA.
Jan Panteltje
2024-02-10 06:44:06 UTC
Permalink
On a sunny day (Fri, 09 Feb 2024 10:17:15 -0800) it happened john larkin
Post by john larkin
Post by Jan Panteltje
On a sunny day (Fri, 09 Feb 2024 08:54:12 -0800) it happened John Larkin
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
On Fri, 9 Feb 2024 04:02:26 -0800 (PST), Fred Bloggs
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Almost all our products were simulated in LT Spice, and that steered
or verified the designs.
Post by Jan Panteltje
Real hands on testing is needed to get a clue.
We don't prototype many products; they are too complex. It's more
efficient to design, simulate, review, and manufacture. If the
engineering is done right, you can sell rev A. I nkow of one giant
organization that formally plans for six PCB spins before production.
Why bother to be careful on the first few?
We do breadboard little subcircuits if we're not sure how they will
behave. Data sheets can be vague. Dremeling is fun.
Hers's a product sim.
https://www.dropbox.com/scl/fi/gpjalf4g0c5m6qlg2j6ce/PMA_Sim_4.jpg?rlkey=8bh9cjpl5e3ltqm683hss1ufm&raw=1
I did this all analog, in LT Spice, to get the dynamics right. Most of
this will actually be done in an FPGA.
Last time I uses LTspice was for this filter:
Loading Image...
Loading Image...
That circuit shifts audio 90 degrees from few hundred Hz to about 3.4 kHz,
is used to generate a SSB signal using quadrature modulation.
As I did not have the right caps (circuit was not designed by me is maybe from the fifies of last century but now with opamps)
I wanted to know what the phase shifts were at different frequenies.
Loading Image...
Simpler to use a crystal filter (I had one back then, what was it XF-9B ?).
https://www.radiomuseum.org/r/kristallvn_quarzfilter_xf9_b.html
But curiosity :-)
John Larkin
2024-02-10 13:42:01 UTC
Permalink
Post by Jan Panteltje
On a sunny day (Fri, 09 Feb 2024 10:17:15 -0800) it happened john larkin
Post by john larkin
Post by Jan Panteltje
On a sunny day (Fri, 09 Feb 2024 08:54:12 -0800) it happened John Larkin
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
On Fri, 9 Feb 2024 04:02:26 -0800 (PST), Fred Bloggs
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Almost all our products were simulated in LT Spice, and that steered
or verified the designs.
Post by Jan Panteltje
Real hands on testing is needed to get a clue.
We don't prototype many products; they are too complex. It's more
efficient to design, simulate, review, and manufacture. If the
engineering is done right, you can sell rev A. I nkow of one giant
organization that formally plans for six PCB spins before production.
Why bother to be careful on the first few?
We do breadboard little subcircuits if we're not sure how they will
behave. Data sheets can be vague. Dremeling is fun.
Hers's a product sim.
https://www.dropbox.com/scl/fi/gpjalf4g0c5m6qlg2j6ce/PMA_Sim_4.jpg?rlkey=8bh9cjpl5e3ltqm683hss1ufm&raw=1
I did this all analog, in LT Spice, to get the dynamics right. Most of
this will actually be done in an FPGA.
https://panteltje.nl/pub/audio_90_degrees_phase_shifter_top_IMG_6903.JPG
https://panteltje.nl/pub/audio_90_degrees_phase_shifter_circuit_diagram_IMG_6958.JPG
That circuit shifts audio 90 degrees from few hundred Hz to about 3.4 kHz,
is used to generate a SSB signal using quadrature modulation.
As I did not have the right caps (circuit was not designed by me is maybe from the fifies of last century but now with opamps)
I wanted to know what the phase shifts were at different frequenies.
https://panteltje.nl/pub/audio_90_degrees_phase_shifter_PCB_bottom_view_IMG_6960.JPG
Simpler to use a crystal filter (I had one back then, what was it XF-9B ?).
https://www.radiomuseum.org/r/kristallvn_quarzfilter_xf9_b.html
But curiosity :-)
If you Spiced that filter, you could quickly plot gain and phase vs
frequency, and twiddle part values and tolerances to see the effects.
And not need trimpots. Measuring gain and phase vs frequency on a
breadboard sounds tedious.

I know some engineers who seem reluctant to learn LT Spice. In fact,
it's really easy. And free.
Jan Panteltje
2024-02-10 15:15:39 UTC
Permalink
On a sunny day (Sat, 10 Feb 2024 05:42:01 -0800) it happened John Larkin
Post by John Larkin
Post by Jan Panteltje
On a sunny day (Fri, 09 Feb 2024 10:17:15 -0800) it happened john larkin
Post by john larkin
Post by Jan Panteltje
On a sunny day (Fri, 09 Feb 2024 08:54:12 -0800) it happened John Larkin
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
On Fri, 9 Feb 2024 04:02:26 -0800 (PST), Fred Bloggs
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Almost all our products were simulated in LT Spice, and that steered
or verified the designs.
Post by Jan Panteltje
Real hands on testing is needed to get a clue.
We don't prototype many products; they are too complex. It's more
efficient to design, simulate, review, and manufacture. If the
engineering is done right, you can sell rev A. I nkow of one giant
organization that formally plans for six PCB spins before production.
Why bother to be careful on the first few?
We do breadboard little subcircuits if we're not sure how they will
behave. Data sheets can be vague. Dremeling is fun.
Hers's a product sim.
https://www.dropbox.com/scl/fi/gpjalf4g0c5m6qlg2j6ce/PMA_Sim_4.jpg?rlkey=8bh9cjpl5e3ltqm683hss1ufm&raw=1
I did this all analog, in LT Spice, to get the dynamics right. Most of
this will actually be done in an FPGA.
https://panteltje.nl/pub/audio_90_degrees_phase_shifter_top_IMG_6903.JPG
https://panteltje.nl/pub/audio_90_degrees_phase_shifter_circuit_diagram_IMG_6958.JPG
That circuit shifts audio 90 degrees from few hundred Hz to about 3.4 kHz,
is used to generate a SSB signal using quadrature modulation.
As I did not have the right caps (circuit was not designed by me is maybe from the fifies of last century but now with opamps)
I wanted to know what the phase shifts were at different frequenies.
https://panteltje.nl/pub/audio_90_degrees_phase_shifter_PCB_bottom_view_IMG_6960.JPG
Simpler to use a crystal filter (I had one back then, what was it XF-9B ?).
https://www.radiomuseum.org/r/kristallvn_quarzfilter_xf9_b.html
But curiosity :-)
If you Spiced that filter, you could quickly plot gain and phase vs
frequency, and twiddle part values and tolerances to see the effects.
And not need trimpots. Measuring gain and phase vs frequency on a
breadboard sounds tedious.
I know some engineers who seem reluctant to learn LT Spice. In fact,
it's really easy. And free.
Yes its easy, I used it
For the 90 degrees:
signal generator on input, x y scope input and output
sweep, see if it stays a circle.
Who needs spice.
Then they can learn soldering an see what a real component looks like.
So they do not have to go dremmeling ;-)
KJW93
2024-02-10 19:14:22 UTC
Permalink
...
Post by john larkin
We don't prototype many products; they are too complex. It's more
efficient to design, simulate, review, and manufacture. If the
engineering is done right, you can sell rev A. I nkow of one giant
organization that formally plans for six PCB spins before production.
Why bother to be careful on the first few?
...

I used to be a developer at such a giant organization and we did usually
have many spins of the PCB.

However it was rarely because there were any faults with the designs,
each revision usually worked as intended, designers were very careful
with each revision with a full suite of simulations of the design
including AC and DC validation of the PCB layout automatically ran at
each stage.

The various revisions were planned so that the other parts of the
organization would have hardware to use for development. It was not
unusual for 5,000 prototype units to be manufactured for such work.

For example the final form factor was not known until the external
product design was completed which usually didn't happen until fairly
late in development.

ASIC design, human factors engineering, software development,
manufacturing, quality assurance, regulatory and other groups all had
their needs and input into the final design but they needed something to
work with in the meanwhile.

The first revisions would be in a development form factor where each
major function had its own PCB that mounted on a large mother board.

Occasionally unusual interactions would occur when the final form factor
designs were created. I was involved in one critical aspect where
acoustic noise from a ceramic capacitor was getting into a microphone as
they ended up only being about 3mm apart; this was not an issue on the
development board where there were many centimeters between them. I
solved it by using a combination of tantalum and COG ceramic capacitors;
Tantalum capacitors in 0402 form factor were becoming available just in
time for production.

kw
Don
2024-02-09 23:29:34 UTC
Permalink
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
With a simple radio set:

<Loading Image...>
<Loading Image...>

No worries all's not nyet
Spice fiddling's a safe bet...

Danke,
--
Don, KB7RPU, https://www.qsl.net/kb7rpu
There was a young lady named Bright Whose speed was far faster than light;
She set out one day In a relative way And returned on the previous night.
Jan Panteltje
2024-02-10 06:23:53 UTC
Permalink
On a sunny day (Fri, 9 Feb 2024 23:29:34 -0000 (UTC)) it happened "Don"
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
<https://crcomp.net/crystalrx/b.png>
That is very nice, crystal earphopne,
I have some from ebay.
In the fifties I was secretly listening to AM radio in my bed at night
using the metal matras of the bed as radio antenna and just such an earphone and home wound
coil crystal radio :-)
Later in the sixties I did read these (Dutch):
https://archive.org/details/radio-blan/Radio_Blan_01_juli_1960/
Jeroen Belleman
2024-02-10 09:41:31 UTC
Permalink
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
[...]
Why not?

Jeroen Belleman
Don
2024-02-10 17:27:25 UTC
Permalink
Post by Jeroen Belleman
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
[...]
Why not?
Good question. And, with all due respect, you ought to post some Spice to
show me how you do it.

Danke,
--
Don, KB7RPU, https://www.qsl.net/kb7rpu
There was a young lady named Bright Whose speed was far faster than light;
She set out one day In a relative way And returned on the previous night.
Jeroen Belleman
2024-02-10 21:33:15 UTC
Permalink
Post by Don
Post by Jeroen Belleman
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
[...]
Why not?
Good question. And, with all due respect, you ought to post some Spice to
show me how you do it.
Danke,
Something akin to this.

Jeroen Belleman

================================================
Version 4
SHEET 1 880 680
WIRE -144 96 -176 96
WIRE -32 96 -64 96
WIRE 272 96 176 96
WIRE 352 96 272 96
WIRE 464 96 416 96
WIRE 512 96 464 96
WIRE 544 96 512 96
WIRE -176 128 -176 96
WIRE -32 128 -32 96
WIRE 176 128 176 96
WIRE 272 128 272 96
WIRE 464 128 464 96
WIRE 544 144 544 96
WIRE -176 240 -176 208
WIRE -32 240 -32 208
WIRE 176 240 176 208
WIRE 272 240 272 192
WIRE 464 240 464 208
WIRE 544 240 544 208
FLAG 176 240 0
FLAG 272 240 0
FLAG 464 240 0
FLAG 544 240 0
FLAG -32 240 0
FLAG -176 240 0
FLAG 512 96 out
SYMBOL ind2 160 112 R0
SYMATTR InstName L1
SYMATTR Value 16µ
SYMATTR Type ind
SYMBOL cap 256 128 R0
SYMATTR InstName C1
SYMATTR Value 6.28n
SYMBOL diode 352 112 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D1
SYMATTR Value AA112
SYMBOL res 448 112 R0
SYMATTR InstName R1
SYMATTR Value 1meg
SYMBOL cap 528 144 R0
SYMATTR InstName C2
SYMATTR Value 10n
SYMBOL ind2 -48 112 R0
SYMATTR InstName L2
SYMATTR Value 16µ
SYMATTR Type ind
SYMBOL bv -176 112 R0
WINDOW 3 -186 181 Left 2
SYMATTR InstName B1
SYMATTR Value V=10k*(1+sin(time*1k*2*pi))*sin(time*1meg*2*pi)
SYMBOL res -48 80 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R2
SYMATTR Value 1
TEXT 72 0 Left 2 !.tran 5m
TEXT 0 112 Left 2 !K12 L1 L2 1u
TEXT -152 0 Left 2 ;Transmitter
TEXT 264 0 Left 2 ;Crystal set receiver
Don
2024-02-10 23:23:14 UTC
Permalink
Post by Jeroen Belleman
Post by Don
Post by Jeroen Belleman
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
[...]
Why not?
Good question. And, with all due respect, you ought to post some Spice to
show me how you do it.
Something akin to this.
<snip>

Brilliant - you're a gentleman, a scholar, and a genius!

While my mind mulled how to capture the crystal circuit in Spice, it
never occurred to me inject its input with a transformer.

Your sim will be put to good use. Thank you.

# # #

Readers who overlooked the final two lines of my poem (the stanzas
someone snipped along the way) will note how they say something
similar to "let he who doesn't Spice fiddle cast the first stone."

ROTFLMAO.

Danke,
--
Don, KB7RPU, https://www.qsl.net/kb7rpu
There was a young lady named Bright Whose speed was far faster than light;
She set out one day In a relative way And returned on the previous night.
John Larkin
2024-02-11 01:22:11 UTC
Permalink
On Sat, 10 Feb 2024 22:33:15 +0100, Jeroen Belleman
Post by Jeroen Belleman
Post by Don
Post by Jeroen Belleman
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
[...]
Why not?
Good question. And, with all due respect, you ought to post some Spice to
show me how you do it.
Danke,
Something akin to this.
Jeroen Belleman
================================================
Version 4
SHEET 1 880 680
WIRE -144 96 -176 96
WIRE -32 96 -64 96
WIRE 272 96 176 96
WIRE 352 96 272 96
WIRE 464 96 416 96
WIRE 512 96 464 96
WIRE 544 96 512 96
WIRE -176 128 -176 96
WIRE -32 128 -32 96
WIRE 176 128 176 96
WIRE 272 128 272 96
WIRE 464 128 464 96
WIRE 544 144 544 96
WIRE -176 240 -176 208
WIRE -32 240 -32 208
WIRE 176 240 176 208
WIRE 272 240 272 192
WIRE 464 240 464 208
WIRE 544 240 544 208
FLAG 176 240 0
FLAG 272 240 0
FLAG 464 240 0
FLAG 544 240 0
FLAG -32 240 0
FLAG -176 240 0
FLAG 512 96 out
SYMBOL ind2 160 112 R0
SYMATTR InstName L1
SYMATTR Value 16µ
SYMATTR Type ind
SYMBOL cap 256 128 R0
SYMATTR InstName C1
SYMATTR Value 6.28n
SYMBOL diode 352 112 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D1
SYMATTR Value AA112
SYMBOL res 448 112 R0
SYMATTR InstName R1
SYMATTR Value 1meg
SYMBOL cap 528 144 R0
SYMATTR InstName C2
SYMATTR Value 10n
SYMBOL ind2 -48 112 R0
SYMATTR InstName L2
SYMATTR Value 16µ
SYMATTR Type ind
SYMBOL bv -176 112 R0
WINDOW 3 -186 181 Left 2
SYMATTR InstName B1
SYMATTR Value V=10k*(1+sin(time*1k*2*pi))*sin(time*1meg*2*pi)
SYMBOL res -48 80 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R2
SYMATTR Value 1
TEXT 72 0 Left 2 !.tran 5m
TEXT 0 112 Left 2 !K12 L1 L2 1u
TEXT -152 0 Left 2 ;Transmitter
TEXT 264 0 Left 2 ;Crystal set receiver
Suggestions:

Reduce the modulation a bit. It distorts at 100%.

Make R1 2K, like an old dynamic headphone.

Change the coupling to 50u. A good antenna makes a lot of signal.

You can use a BAT54 schottky.
Jeroen Belleman
2024-02-11 09:39:49 UTC
Permalink
Post by John Larkin
On Sat, 10 Feb 2024 22:33:15 +0100, Jeroen Belleman
Post by Jeroen Belleman
Post by Don
Post by Jeroen Belleman
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
[...]
Why not?
Good question. And, with all due respect, you ought to post some Spice to
show me how you do it.
Danke,
Something akin to this.
Jeroen Belleman
================================================
Version 4
SHEET 1 880 680
WIRE -144 96 -176 96
WIRE -32 96 -64 96
WIRE 272 96 176 96
WIRE 352 96 272 96
WIRE 464 96 416 96
WIRE 512 96 464 96
WIRE 544 96 512 96
WIRE -176 128 -176 96
WIRE -32 128 -32 96
WIRE 176 128 176 96
WIRE 272 128 272 96
WIRE 464 128 464 96
WIRE 544 144 544 96
WIRE -176 240 -176 208
WIRE -32 240 -32 208
WIRE 176 240 176 208
WIRE 272 240 272 192
WIRE 464 240 464 208
WIRE 544 240 544 208
FLAG 176 240 0
FLAG 272 240 0
FLAG 464 240 0
FLAG 544 240 0
FLAG -32 240 0
FLAG -176 240 0
FLAG 512 96 out
SYMBOL ind2 160 112 R0
SYMATTR InstName L1
SYMATTR Value 16µ
SYMATTR Type ind
SYMBOL cap 256 128 R0
SYMATTR InstName C1
SYMATTR Value 6.28n
SYMBOL diode 352 112 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D1
SYMATTR Value AA112
SYMBOL res 448 112 R0
SYMATTR InstName R1
SYMATTR Value 1meg
SYMBOL cap 528 144 R0
SYMATTR InstName C2
SYMATTR Value 10n
SYMBOL ind2 -48 112 R0
SYMATTR InstName L2
SYMATTR Value 16µ
SYMATTR Type ind
SYMBOL bv -176 112 R0
WINDOW 3 -186 181 Left 2
SYMATTR InstName B1
SYMATTR Value V=10k*(1+sin(time*1k*2*pi))*sin(time*1meg*2*pi)
SYMBOL res -48 80 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R2
SYMATTR Value 1
TEXT 72 0 Left 2 !.tran 5m
TEXT 0 112 Left 2 !K12 L1 L2 1u
TEXT -152 0 Left 2 ;Transmitter
TEXT 264 0 Left 2 ;Crystal set receiver
Reduce the modulation a bit. It distorts at 100%.
Make R1 2K, like an old dynamic headphone.
Change the coupling to 50u. A good antenna makes a lot of signal.
You can use a BAT54 schottky.
You don't expect high fidelity from a crystal set, but OK, fair enough.
C2 was meant to model a piezo earbud. I should actually measure one...

I got the values of C! and L1 wrong. C1=300pF is more reasonable,
and then L1=84.43uH. In practice, of course, C1 would be tuned and L1
would be fixed at 80uH or so, but either way works.

The voltage of B1 and the tranformer coupling factor are just guesses,
based on the knwoledge that crystal sets actually do work.

Jeroen Belleman
John Larkin
2024-02-11 15:28:39 UTC
Permalink
On Sun, 11 Feb 2024 10:39:49 +0100, Jeroen Belleman
Post by Jeroen Belleman
Post by John Larkin
On Sat, 10 Feb 2024 22:33:15 +0100, Jeroen Belleman
Post by Jeroen Belleman
Post by Don
Post by Jeroen Belleman
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
[...]
Why not?
Good question. And, with all due respect, you ought to post some Spice to
show me how you do it.
Danke,
Something akin to this.
Jeroen Belleman
================================================
Version 4
SHEET 1 880 680
WIRE -144 96 -176 96
WIRE -32 96 -64 96
WIRE 272 96 176 96
WIRE 352 96 272 96
WIRE 464 96 416 96
WIRE 512 96 464 96
WIRE 544 96 512 96
WIRE -176 128 -176 96
WIRE -32 128 -32 96
WIRE 176 128 176 96
WIRE 272 128 272 96
WIRE 464 128 464 96
WIRE 544 144 544 96
WIRE -176 240 -176 208
WIRE -32 240 -32 208
WIRE 176 240 176 208
WIRE 272 240 272 192
WIRE 464 240 464 208
WIRE 544 240 544 208
FLAG 176 240 0
FLAG 272 240 0
FLAG 464 240 0
FLAG 544 240 0
FLAG -32 240 0
FLAG -176 240 0
FLAG 512 96 out
SYMBOL ind2 160 112 R0
SYMATTR InstName L1
SYMATTR Value 16µ
SYMATTR Type ind
SYMBOL cap 256 128 R0
SYMATTR InstName C1
SYMATTR Value 6.28n
SYMBOL diode 352 112 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D1
SYMATTR Value AA112
SYMBOL res 448 112 R0
SYMATTR InstName R1
SYMATTR Value 1meg
SYMBOL cap 528 144 R0
SYMATTR InstName C2
SYMATTR Value 10n
SYMBOL ind2 -48 112 R0
SYMATTR InstName L2
SYMATTR Value 16µ
SYMATTR Type ind
SYMBOL bv -176 112 R0
WINDOW 3 -186 181 Left 2
SYMATTR InstName B1
SYMATTR Value V=10k*(1+sin(time*1k*2*pi))*sin(time*1meg*2*pi)
SYMBOL res -48 80 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R2
SYMATTR Value 1
TEXT 72 0 Left 2 !.tran 5m
TEXT 0 112 Left 2 !K12 L1 L2 1u
TEXT -152 0 Left 2 ;Transmitter
TEXT 264 0 Left 2 ;Crystal set receiver
Reduce the modulation a bit. It distorts at 100%.
Make R1 2K, like an old dynamic headphone.
Change the coupling to 50u. A good antenna makes a lot of signal.
You can use a BAT54 schottky.
You don't expect high fidelity from a crystal set, but OK, fair enough.
C2 was meant to model a piezo earbud. I should actually measure one...
I got the values of C! and L1 wrong. C1=300pF is more reasonable,
and then L1=84.43uH. In practice, of course, C1 would be tuned and L1
would be fixed at 80uH or so, but either way works.
The voltage of B1 and the tranformer coupling factor are just guesses,
based on the knwoledge that crystal sets actually do work.
Jeroen Belleman
I think that most xtal sets do some sort of impedance matching from
the antenna up into the main coil, taps on the coil or an extra
winding. A really longwire antanna will have a lot of capacitance and
an equivalent lowish impedance.

The people who have made xtal sets over the last century or so don't
seem to have been very quantitative.

I've wondered if I can slurp enought power out of the air to light an
LED. My office is in a wooden building about 3 miles from Sutro Tower.
Castro valley is a bowl that I suspect focusses the RF onto us.
Cursitor Doom
2024-02-11 17:40:10 UTC
Permalink
Post by John Larkin
On Sun, 11 Feb 2024 10:39:49 +0100, Jeroen Belleman
Post by Jeroen Belleman
Post by John Larkin
On Sat, 10 Feb 2024 22:33:15 +0100, Jeroen Belleman
Post by Jeroen Belleman
Post by Don
Post by Jeroen Belleman
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
[...]
Why not?
Good question. And, with all due respect, you ought to post some Spice to
show me how you do it.
Danke,
Something akin to this.
Jeroen Belleman
================================================
Version 4
SHEET 1 880 680
WIRE -144 96 -176 96
WIRE -32 96 -64 96
WIRE 272 96 176 96
WIRE 352 96 272 96
WIRE 464 96 416 96
WIRE 512 96 464 96
WIRE 544 96 512 96
WIRE -176 128 -176 96
WIRE -32 128 -32 96
WIRE 176 128 176 96
WIRE 272 128 272 96
WIRE 464 128 464 96
WIRE 544 144 544 96
WIRE -176 240 -176 208
WIRE -32 240 -32 208
WIRE 176 240 176 208
WIRE 272 240 272 192
WIRE 464 240 464 208
WIRE 544 240 544 208
FLAG 176 240 0
FLAG 272 240 0
FLAG 464 240 0
FLAG 544 240 0
FLAG -32 240 0
FLAG -176 240 0
FLAG 512 96 out
SYMBOL ind2 160 112 R0
SYMATTR InstName L1
SYMATTR Value 16µ
SYMATTR Type ind
SYMBOL cap 256 128 R0
SYMATTR InstName C1
SYMATTR Value 6.28n
SYMBOL diode 352 112 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D1
SYMATTR Value AA112
SYMBOL res 448 112 R0
SYMATTR InstName R1
SYMATTR Value 1meg
SYMBOL cap 528 144 R0
SYMATTR InstName C2
SYMATTR Value 10n
SYMBOL ind2 -48 112 R0
SYMATTR InstName L2
SYMATTR Value 16µ
SYMATTR Type ind
SYMBOL bv -176 112 R0
WINDOW 3 -186 181 Left 2
SYMATTR InstName B1
SYMATTR Value V=10k*(1+sin(time*1k*2*pi))*sin(time*1meg*2*pi)
SYMBOL res -48 80 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R2
SYMATTR Value 1
TEXT 72 0 Left 2 !.tran 5m
TEXT 0 112 Left 2 !K12 L1 L2 1u
TEXT -152 0 Left 2 ;Transmitter
TEXT 264 0 Left 2 ;Crystal set receiver
Reduce the modulation a bit. It distorts at 100%.
Make R1 2K, like an old dynamic headphone.
Change the coupling to 50u. A good antenna makes a lot of signal.
You can use a BAT54 schottky.
You don't expect high fidelity from a crystal set, but OK, fair enough.
C2 was meant to model a piezo earbud. I should actually measure one...
I got the values of C! and L1 wrong. C1=300pF is more reasonable,
and then L1=84.43uH. In practice, of course, C1 would be tuned and L1
would be fixed at 80uH or so, but either way works.
The voltage of B1 and the tranformer coupling factor are just guesses,
based on the knwoledge that crystal sets actually do work.
Jeroen Belleman
I think that most xtal sets do some sort of impedance matching from
the antenna up into the main coil, taps on the coil or an extra
winding. A really longwire antanna will have a lot of capacitance and
an equivalent lowish impedance.
The people who have made xtal sets over the last century or so don't
seem to have been very quantitative.
I've wondered if I can slurp enought power out of the air to light an
LED. My office is in a wooden building about 3 miles from Sutro Tower.
Castro valley is a bowl that I suspect focusses the RF onto us.
LEDs require relatively large amounts of current, so I don't see how
that would be possible, John. There's only *one* bloke I can think of
who could conceivably pull that off and he's long dead, sadly.
Edison's nemesis:

https://en.wikipedia.org/wiki/Nikola_Tesla
John Larkin
2024-02-11 18:32:32 UTC
Permalink
Post by Cursitor Doom
Post by John Larkin
On Sun, 11 Feb 2024 10:39:49 +0100, Jeroen Belleman
Post by Jeroen Belleman
Post by John Larkin
On Sat, 10 Feb 2024 22:33:15 +0100, Jeroen Belleman
Post by Jeroen Belleman
Post by Don
Post by Jeroen Belleman
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
[...]
Why not?
Good question. And, with all due respect, you ought to post some Spice to
show me how you do it.
Danke,
Something akin to this.
Jeroen Belleman
================================================
Version 4
SHEET 1 880 680
WIRE -144 96 -176 96
WIRE -32 96 -64 96
WIRE 272 96 176 96
WIRE 352 96 272 96
WIRE 464 96 416 96
WIRE 512 96 464 96
WIRE 544 96 512 96
WIRE -176 128 -176 96
WIRE -32 128 -32 96
WIRE 176 128 176 96
WIRE 272 128 272 96
WIRE 464 128 464 96
WIRE 544 144 544 96
WIRE -176 240 -176 208
WIRE -32 240 -32 208
WIRE 176 240 176 208
WIRE 272 240 272 192
WIRE 464 240 464 208
WIRE 544 240 544 208
FLAG 176 240 0
FLAG 272 240 0
FLAG 464 240 0
FLAG 544 240 0
FLAG -32 240 0
FLAG -176 240 0
FLAG 512 96 out
SYMBOL ind2 160 112 R0
SYMATTR InstName L1
SYMATTR Value 16µ
SYMATTR Type ind
SYMBOL cap 256 128 R0
SYMATTR InstName C1
SYMATTR Value 6.28n
SYMBOL diode 352 112 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D1
SYMATTR Value AA112
SYMBOL res 448 112 R0
SYMATTR InstName R1
SYMATTR Value 1meg
SYMBOL cap 528 144 R0
SYMATTR InstName C2
SYMATTR Value 10n
SYMBOL ind2 -48 112 R0
SYMATTR InstName L2
SYMATTR Value 16µ
SYMATTR Type ind
SYMBOL bv -176 112 R0
WINDOW 3 -186 181 Left 2
SYMATTR InstName B1
SYMATTR Value V=10k*(1+sin(time*1k*2*pi))*sin(time*1meg*2*pi)
SYMBOL res -48 80 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R2
SYMATTR Value 1
TEXT 72 0 Left 2 !.tran 5m
TEXT 0 112 Left 2 !K12 L1 L2 1u
TEXT -152 0 Left 2 ;Transmitter
TEXT 264 0 Left 2 ;Crystal set receiver
Reduce the modulation a bit. It distorts at 100%.
Make R1 2K, like an old dynamic headphone.
Change the coupling to 50u. A good antenna makes a lot of signal.
You can use a BAT54 schottky.
You don't expect high fidelity from a crystal set, but OK, fair enough.
C2 was meant to model a piezo earbud. I should actually measure one...
I got the values of C! and L1 wrong. C1=300pF is more reasonable,
and then L1=84.43uH. In practice, of course, C1 would be tuned and L1
would be fixed at 80uH or so, but either way works.
The voltage of B1 and the tranformer coupling factor are just guesses,
based on the knwoledge that crystal sets actually do work.
Jeroen Belleman
I think that most xtal sets do some sort of impedance matching from
the antenna up into the main coil, taps on the coil or an extra
winding. A really longwire antanna will have a lot of capacitance and
an equivalent lowish impedance.
The people who have made xtal sets over the last century or so don't
seem to have been very quantitative.
I've wondered if I can slurp enought power out of the air to light an
LED. My office is in a wooden building about 3 miles from Sutro Tower.
Castro valley is a bowl that I suspect focusses the RF onto us.
LEDs require relatively large amounts of current, so I don't see how
that would be possible, John. There's only *one* bloke I can think of
who could conceivably pull that off and he's long dead, sadly.
https://en.wikipedia.org/wiki/Nikola_Tesla
A good LED is visible in typical room light at 1 uA, which would be
about 2 uW. I'd imagine that a modest antenna and a decent detector
could do much better than that.

The difference between a crystal set and lighting an LED is that the
LED case should use the entire spectrum, AM TV FM, which is densely
packed where I am. That would take some thinking.
Cursitor Doom
2024-02-11 19:34:53 UTC
Permalink
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
On Sun, 11 Feb 2024 10:39:49 +0100, Jeroen Belleman
Post by Jeroen Belleman
Post by John Larkin
On Sat, 10 Feb 2024 22:33:15 +0100, Jeroen Belleman
Post by Jeroen Belleman
Post by Don
Post by Jeroen Belleman
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
[...]
Why not?
Good question. And, with all due respect, you ought to post some Spice to
show me how you do it.
Danke,
Something akin to this.
Jeroen Belleman
================================================
Version 4
SHEET 1 880 680
WIRE -144 96 -176 96
WIRE -32 96 -64 96
WIRE 272 96 176 96
WIRE 352 96 272 96
WIRE 464 96 416 96
WIRE 512 96 464 96
WIRE 544 96 512 96
WIRE -176 128 -176 96
WIRE -32 128 -32 96
WIRE 176 128 176 96
WIRE 272 128 272 96
WIRE 464 128 464 96
WIRE 544 144 544 96
WIRE -176 240 -176 208
WIRE -32 240 -32 208
WIRE 176 240 176 208
WIRE 272 240 272 192
WIRE 464 240 464 208
WIRE 544 240 544 208
FLAG 176 240 0
FLAG 272 240 0
FLAG 464 240 0
FLAG 544 240 0
FLAG -32 240 0
FLAG -176 240 0
FLAG 512 96 out
SYMBOL ind2 160 112 R0
SYMATTR InstName L1
SYMATTR Value 16µ
SYMATTR Type ind
SYMBOL cap 256 128 R0
SYMATTR InstName C1
SYMATTR Value 6.28n
SYMBOL diode 352 112 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D1
SYMATTR Value AA112
SYMBOL res 448 112 R0
SYMATTR InstName R1
SYMATTR Value 1meg
SYMBOL cap 528 144 R0
SYMATTR InstName C2
SYMATTR Value 10n
SYMBOL ind2 -48 112 R0
SYMATTR InstName L2
SYMATTR Value 16µ
SYMATTR Type ind
SYMBOL bv -176 112 R0
WINDOW 3 -186 181 Left 2
SYMATTR InstName B1
SYMATTR Value V=10k*(1+sin(time*1k*2*pi))*sin(time*1meg*2*pi)
SYMBOL res -48 80 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R2
SYMATTR Value 1
TEXT 72 0 Left 2 !.tran 5m
TEXT 0 112 Left 2 !K12 L1 L2 1u
TEXT -152 0 Left 2 ;Transmitter
TEXT 264 0 Left 2 ;Crystal set receiver
Reduce the modulation a bit. It distorts at 100%.
Make R1 2K, like an old dynamic headphone.
Change the coupling to 50u. A good antenna makes a lot of signal.
You can use a BAT54 schottky.
You don't expect high fidelity from a crystal set, but OK, fair enough.
C2 was meant to model a piezo earbud. I should actually measure one...
I got the values of C! and L1 wrong. C1=300pF is more reasonable,
and then L1=84.43uH. In practice, of course, C1 would be tuned and L1
would be fixed at 80uH or so, but either way works.
The voltage of B1 and the tranformer coupling factor are just guesses,
based on the knwoledge that crystal sets actually do work.
Jeroen Belleman
I think that most xtal sets do some sort of impedance matching from
the antenna up into the main coil, taps on the coil or an extra
winding. A really longwire antanna will have a lot of capacitance and
an equivalent lowish impedance.
The people who have made xtal sets over the last century or so don't
seem to have been very quantitative.
I've wondered if I can slurp enought power out of the air to light an
LED. My office is in a wooden building about 3 miles from Sutro Tower.
Castro valley is a bowl that I suspect focusses the RF onto us.
LEDs require relatively large amounts of current, so I don't see how
that would be possible, John. There's only *one* bloke I can think of
who could conceivably pull that off and he's long dead, sadly.
https://en.wikipedia.org/wiki/Nikola_Tesla
A good LED is visible in typical room light at 1 uA, which would be
about 2 uW. I'd imagine that a modest antenna and a decent detector
could do much better than that.
The difference between a crystal set and lighting an LED is that the
LED case should use the entire spectrum, AM TV FM, which is densely
packed where I am. That would take some thinking.
Neon or similar would be far more easily achievable I reckon.
whit3rd
2024-02-11 20:35:14 UTC
Permalink
Post by Cursitor Doom
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
I've wondered if I can slurp enought power out of the air to light an
LED. My office is in a wooden building about 3 miles from Sutro Tower.
Castro valley is a bowl that I suspect focusses the RF onto us.
LEDs require relatively large amounts of current, so I don't see how
that would be possible, John. There's only *one* bloke I can think of
who could conceivably pull that off and he's long dead, sadly.
https://en.wikipedia.org/wiki/Nikola_Tesla
A good LED is visible in typical room light at 1 uA, which would be
about 2 uW. I'd imagine that a modest antenna and a decent detector
could do much better than that.
The difference between a crystal set and lighting an LED is that the
LED case should use the entire spectrum, AM TV FM, which is densely
packed where I am. That would take some thinking.
Neon or similar would be far more easily achievable I reckon.
A neon lamp has negative resistance; it'd blink if you could get the voltage
high enough against rectifier leakage.

Easiest way to get a photon or five would be with antenna-and-tuning,
so would be filtering to a single input carrier frequency; TV station
would be the old-school answer, but they're in the UHF (tricky,
because the tuning requires you to wind a UHF inductor) so
the next-best is AM broadcast receiver.

Absolute easiest, is if a neighbor has a SW station; one of mine had a nice
tower antenna (and I had TV reception for nearly 100 miles because I
could aim a Yagi at his tower and get a bounce from over-the-horizon).
OK, maybe inductive coupling to 60 Hz in power-on-poles would
be easier, but also... slightly illegal.

Instead of creating a dedicated antenna, one could inductively clamp on
a few places, maybe your floor lamp unswitched (neutral) wire is worth
untangling for test purposes.
Cursitor Doom
2024-02-12 00:04:44 UTC
Permalink
Post by whit3rd
Post by Cursitor Doom
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
I've wondered if I can slurp enought power out of the air to light an
LED. My office is in a wooden building about 3 miles from Sutro Tower.
Castro valley is a bowl that I suspect focusses the RF onto us.
LEDs require relatively large amounts of current, so I don't see how
that would be possible, John. There's only *one* bloke I can think of
who could conceivably pull that off and he's long dead, sadly.
https://en.wikipedia.org/wiki/Nikola_Tesla
A good LED is visible in typical room light at 1 uA, which would be
about 2 uW. I'd imagine that a modest antenna and a decent detector
could do much better than that.
The difference between a crystal set and lighting an LED is that the
LED case should use the entire spectrum, AM TV FM, which is densely
packed where I am. That would take some thinking.
Neon or similar would be far more easily achievable I reckon.
A neon lamp has negative resistance; it'd blink if you could get the voltage
high enough against rectifier leakage.
Easiest way to get a photon or five would be with antenna-and-tuning,
so would be filtering to a single input carrier frequency; TV station
would be the old-school answer, but they're in the UHF (tricky,
because the tuning requires you to wind a UHF inductor) so
the next-best is AM broadcast receiver.
Somewhere I still have an old timey wavemeter that works off this
principle. When tuned to the same frequency as a nearby transmitter,
the bulb lights. Amazing how far we've come in 100 years.
John Larkin
2024-02-11 23:14:39 UTC
Permalink
Post by Cursitor Doom
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
On Sun, 11 Feb 2024 10:39:49 +0100, Jeroen Belleman
Post by Jeroen Belleman
Post by John Larkin
On Sat, 10 Feb 2024 22:33:15 +0100, Jeroen Belleman
Post by Jeroen Belleman
Post by Don
Post by Jeroen Belleman
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
[...]
Why not?
Good question. And, with all due respect, you ought to post some Spice to
show me how you do it.
Danke,
Something akin to this.
Jeroen Belleman
================================================
Version 4
SHEET 1 880 680
WIRE -144 96 -176 96
WIRE -32 96 -64 96
WIRE 272 96 176 96
WIRE 352 96 272 96
WIRE 464 96 416 96
WIRE 512 96 464 96
WIRE 544 96 512 96
WIRE -176 128 -176 96
WIRE -32 128 -32 96
WIRE 176 128 176 96
WIRE 272 128 272 96
WIRE 464 128 464 96
WIRE 544 144 544 96
WIRE -176 240 -176 208
WIRE -32 240 -32 208
WIRE 176 240 176 208
WIRE 272 240 272 192
WIRE 464 240 464 208
WIRE 544 240 544 208
FLAG 176 240 0
FLAG 272 240 0
FLAG 464 240 0
FLAG 544 240 0
FLAG -32 240 0
FLAG -176 240 0
FLAG 512 96 out
SYMBOL ind2 160 112 R0
SYMATTR InstName L1
SYMATTR Value 16µ
SYMATTR Type ind
SYMBOL cap 256 128 R0
SYMATTR InstName C1
SYMATTR Value 6.28n
SYMBOL diode 352 112 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D1
SYMATTR Value AA112
SYMBOL res 448 112 R0
SYMATTR InstName R1
SYMATTR Value 1meg
SYMBOL cap 528 144 R0
SYMATTR InstName C2
SYMATTR Value 10n
SYMBOL ind2 -48 112 R0
SYMATTR InstName L2
SYMATTR Value 16µ
SYMATTR Type ind
SYMBOL bv -176 112 R0
WINDOW 3 -186 181 Left 2
SYMATTR InstName B1
SYMATTR Value V=10k*(1+sin(time*1k*2*pi))*sin(time*1meg*2*pi)
SYMBOL res -48 80 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R2
SYMATTR Value 1
TEXT 72 0 Left 2 !.tran 5m
TEXT 0 112 Left 2 !K12 L1 L2 1u
TEXT -152 0 Left 2 ;Transmitter
TEXT 264 0 Left 2 ;Crystal set receiver
Reduce the modulation a bit. It distorts at 100%.
Make R1 2K, like an old dynamic headphone.
Change the coupling to 50u. A good antenna makes a lot of signal.
You can use a BAT54 schottky.
You don't expect high fidelity from a crystal set, but OK, fair enough.
C2 was meant to model a piezo earbud. I should actually measure one...
I got the values of C! and L1 wrong. C1=300pF is more reasonable,
and then L1=84.43uH. In practice, of course, C1 would be tuned and L1
would be fixed at 80uH or so, but either way works.
The voltage of B1 and the tranformer coupling factor are just guesses,
based on the knwoledge that crystal sets actually do work.
Jeroen Belleman
I think that most xtal sets do some sort of impedance matching from
the antenna up into the main coil, taps on the coil or an extra
winding. A really longwire antanna will have a lot of capacitance and
an equivalent lowish impedance.
The people who have made xtal sets over the last century or so don't
seem to have been very quantitative.
I've wondered if I can slurp enought power out of the air to light an
LED. My office is in a wooden building about 3 miles from Sutro Tower.
Castro valley is a bowl that I suspect focusses the RF onto us.
LEDs require relatively large amounts of current, so I don't see how
that would be possible, John. There's only *one* bloke I can think of
who could conceivably pull that off and he's long dead, sadly.
https://en.wikipedia.org/wiki/Nikola_Tesla
A good LED is visible in typical room light at 1 uA, which would be
about 2 uW. I'd imagine that a modest antenna and a decent detector
could do much better than that.
The difference between a crystal set and lighting an LED is that the
LED case should use the entire spectrum, AM TV FM, which is densely
packed where I am. That would take some thinking.
Neon or similar would be far more easily achievable I reckon.
That would need to make about 100 volts from millivolts of RF. Tough.

I'd also expect an LED to be much more efficient than a neon.
Jan Panteltje
2024-02-12 05:45:31 UTC
Permalink
On a sunny day (Sun, 11 Feb 2024 19:34:53 +0000) it happened Cursitor Doom
Post by Cursitor Doom
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
On Sun, 11 Feb 2024 10:39:49 +0100, Jeroen Belleman
Post by Jeroen Belleman
Post by John Larkin
On Sat, 10 Feb 2024 22:33:15 +0100, Jeroen Belleman
Post by Jeroen Belleman
Post by Don
Post by Jeroen Belleman
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
[...]
Why not?
Good question. And, with all due respect, you ought to post some Spice to
show me how you do it.
Danke,
Something akin to this.
Jeroen Belleman
================================================
Version 4
SHEET 1 880 680
WIRE -144 96 -176 96
WIRE -32 96 -64 96
WIRE 272 96 176 96
WIRE 352 96 272 96
WIRE 464 96 416 96
WIRE 512 96 464 96
WIRE 544 96 512 96
WIRE -176 128 -176 96
WIRE -32 128 -32 96
WIRE 176 128 176 96
WIRE 272 128 272 96
WIRE 464 128 464 96
WIRE 544 144 544 96
WIRE -176 240 -176 208
WIRE -32 240 -32 208
WIRE 176 240 176 208
WIRE 272 240 272 192
WIRE 464 240 464 208
WIRE 544 240 544 208
FLAG 176 240 0
FLAG 272 240 0
FLAG 464 240 0
FLAG 544 240 0
FLAG -32 240 0
FLAG -176 240 0
FLAG 512 96 out
SYMBOL ind2 160 112 R0
SYMATTR InstName L1
SYMATTR Value 16µ
SYMATTR Type ind
SYMBOL cap 256 128 R0
SYMATTR InstName C1
SYMATTR Value 6.28n
SYMBOL diode 352 112 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D1
SYMATTR Value AA112
SYMBOL res 448 112 R0
SYMATTR InstName R1
SYMATTR Value 1meg
SYMBOL cap 528 144 R0
SYMATTR InstName C2
SYMATTR Value 10n
SYMBOL ind2 -48 112 R0
SYMATTR InstName L2
SYMATTR Value 16µ
SYMATTR Type ind
SYMBOL bv -176 112 R0
WINDOW 3 -186 181 Left 2
SYMATTR InstName B1
SYMATTR Value V=10k*(1+sin(time*1k*2*pi))*sin(time*1meg*2*pi)
SYMBOL res -48 80 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R2
SYMATTR Value 1
TEXT 72 0 Left 2 !.tran 5m
TEXT 0 112 Left 2 !K12 L1 L2 1u
TEXT -152 0 Left 2 ;Transmitter
TEXT 264 0 Left 2 ;Crystal set receiver
Reduce the modulation a bit. It distorts at 100%.
Make R1 2K, like an old dynamic headphone.
Change the coupling to 50u. A good antenna makes a lot of signal.
You can use a BAT54 schottky.
You don't expect high fidelity from a crystal set, but OK, fair enough.
C2 was meant to model a piezo earbud. I should actually measure one...
I got the values of C! and L1 wrong. C1=300pF is more reasonable,
and then L1=84.43uH. In practice, of course, C1 would be tuned and L1
would be fixed at 80uH or so, but either way works.
The voltage of B1 and the tranformer coupling factor are just guesses,
based on the knwoledge that crystal sets actually do work.
Jeroen Belleman
I think that most xtal sets do some sort of impedance matching from
the antenna up into the main coil, taps on the coil or an extra
winding. A really longwire antanna will have a lot of capacitance and
an equivalent lowish impedance.
The people who have made xtal sets over the last century or so don't
seem to have been very quantitative.
I've wondered if I can slurp enought power out of the air to light an
LED. My office is in a wooden building about 3 miles from Sutro Tower.
Castro valley is a bowl that I suspect focusses the RF onto us.
LEDs require relatively large amounts of current, so I don't see how
that would be possible, John. There's only *one* bloke I can think of
who could conceivably pull that off and he's long dead, sadly.
https://en.wikipedia.org/wiki/Nikola_Tesla
A good LED is visible in typical room light at 1 uA, which would be
about 2 uW. I'd imagine that a modest antenna and a decent detector
could do much better than that.
The difference between a crystal set and lighting an LED is that the
LED case should use the entire spectrum, AM TV FM, which is densely
packed where I am. That would take some thinking.
Neon or similar would be far more easily achievable I reckon.
Yes a neon light would light up next to antenna when I transmitted
on shortwave.
Jan Panteltje
2024-02-12 05:43:27 UTC
Permalink
On a sunny day (Sun, 11 Feb 2024 17:40:10 +0000) it happened Cursitor Doom
Post by Cursitor Doom
Post by John Larkin
I've wondered if I can slurp enought power out of the air to light an
LED. My office is in a wooden building about 3 miles from Sutro Tower.
Castro valley is a bowl that I suspect focusses the RF onto us.
LEDs require relatively large amounts of current, so I don't see how
that would be possible, John. There's only *one* bloke I can think of
who could conceivably pull that off and he's long dead, sadly.
https://en.wikipedia.org/wiki/Nikola_Tesla
Many years ago I tried a LED on my 27 MHz GPA antenna,
It did light up when my cross the street neighbor transmitted with his set.
I asked him how much power he used,
did not get an answer, 4 W was the legal limit back then IIRC,
some Italians on that band used hundreds of watts.
Anthony William Sloman
2024-02-14 03:01:10 UTC
Permalink
Post by Cursitor Doom
Post by John Larkin
I've wondered if I can slurp enought power out of the air to light an
LED. My office is in a wooden building about 3 miles from Sutro Tower.
Castro valley is a bowl that I suspect focusses the RF onto us.
LEDs require relatively large amounts of current, so I don't see how
that would be possible, John. There's only *one* bloke I can think of
who could conceivably pull that off and he's long dead, sadly.
https://en.wikipedia.org/wiki/Nikola_Tesla
That's pure ignorance talking. If you are close to a transmitter you can collect a lot of power.

People who lived close to the Droitwich transmitter in the UK

https://en.wikipedia.org/wiki/Droitwich_Transmitting_Station

were legally obliged to shield their fluorescent lamps with coarse wire mesh (chicken wire) shields.

Without the shield the lamps would run continuously, soaking up up the expensively transmitted power and providing free lighting.
--
Bill Sloman, Sydney
John Larkin
2024-02-10 13:44:11 UTC
Permalink
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
<https://crcomp.net/crystalrx/b.png>
No worries all's not nyet
Spice fiddling's a safe bet...
Danke,
Why does a crystal set need a power supply?

I don't think the circuit will work with a crystal headphone.
Cursitor Doom
2024-02-10 14:32:10 UTC
Permalink
Post by John Larkin
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
<https://crcomp.net/crystalrx/b.png>
No worries all's not nyet
Spice fiddling's a safe bet...
Danke,
Why does a crystal set need a power supply?
I don't think the circuit will work with a crystal headphone.
It wouldn't work without one. Xtal sets need hi-z earphones for
obvious reasons and crystal earpieces are ideal.
Arie de Muijnck
2024-02-10 14:39:14 UTC
Permalink
Post by Cursitor Doom
Post by John Larkin
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
<https://crcomp.net/crystalrx/b.png>
No worries all's not nyet
Spice fiddling's a safe bet...
Danke,
Why does a crystal set need a power supply?
I don't think the circuit will work with a crystal headphone.
It wouldn't work without one. Xtal sets need hi-z earphones for
obvious reasons and crystal earpieces are ideal.
Yes, but they need a leakage resistor over the earpiece so the diode kan stay doing its work. A pencil stripe will do.
Just as the inductive earpiece needs a parallel capacitor for optimal work (often the wiring functions as such).

Arie
Don
2024-02-13 03:46:20 UTC
Permalink
Post by Arie de Muijnck
Post by Cursitor Doom
Post by John Larkin
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
<https://crcomp.net/crystalrx/b.png>
No worries all's not nyet
Spice fiddling's a safe bet...
Why does a crystal set need a power supply?
I don't think the circuit will work with a crystal headphone.
It wouldn't work without one. Xtal sets need hi-z earphones for
obvious reasons and crystal earpieces are ideal.
Yes, but they need a leakage resistor over the earpiece so the diode kan
stay doing its work. A pencil stripe will do.
Just as the inductive earpiece needs a parallel capacitor for optimal work
(often the wiring functions as such).
The crystal cult concurs with your circumspect constructive criticism:

<http://techlib.com/electronics/crystal.html>

It turns out a 0.001 uF cap and an 82 k resistor in parallel with the
earphone increases amplitude perceptibly. Thank you for sharing.

Danke,
--
Don, KB7RPU, https://www.qsl.net/kb7rpu
There was a young lady named Bright Whose speed was far faster than light;
She set out one day In a relative way And returned on the previous night.
John Larkin
2024-02-10 15:58:18 UTC
Permalink
Post by Cursitor Doom
Post by John Larkin
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
<https://crcomp.net/crystalrx/b.png>
No worries all's not nyet
Spice fiddling's a safe bet...
Danke,
Why does a crystal set need a power supply?
I don't think the circuit will work with a crystal headphone.
It wouldn't work without one. Xtal sets need hi-z earphones for
obvious reasons and crystal earpieces are ideal.
In the circuit shown, the headphone capacitance will charge up and the
diode will back-bias.

Unless you use a very leaky diode.

What's the power supply for?
Don
2024-02-10 17:27:48 UTC
Permalink
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
<https://crcomp.net/crystalrx/b.png>
No worries all's not nyet
Spice fiddling's a safe bet...
Why does a crystal set need a power supply?
I don't think the circuit will work with a crystal headphone.
It wouldn't work without one. Xtal sets need hi-z earphones for
obvious reasons and crystal earpieces are ideal.
In the circuit shown, the headphone capacitance will charge up and the
diode will back-bias.
Unless you use a very leaky diode.
What's the power supply for?
Good question. Although it primary provides power for future, yet-to-be
developed protos, it also fulfills a fundamental function for a crystal
radio set.

Think about it.

Hint: the power cord conveniently connects three cables to the case, but
not all are used by this crystal circuit.

# # #

My simplest schematic shown uses a hi-Z earphone (although its schematic
symbol sucks IMHO). The demodulating diode depicted works without a
parallel capacitor.

Danke,
--
Don, KB7RPU, https://www.qsl.net/kb7rpu
There was a young lady named Bright Whose speed was far faster than light;
She set out one day In a relative way And returned on the previous night.
John Larkin
2024-02-10 18:02:02 UTC
Permalink
Post by Don
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
<https://crcomp.net/crystalrx/b.png>
No worries all's not nyet
Spice fiddling's a safe bet...
Why does a crystal set need a power supply?
I don't think the circuit will work with a crystal headphone.
It wouldn't work without one. Xtal sets need hi-z earphones for
obvious reasons and crystal earpieces are ideal.
In the circuit shown, the headphone capacitance will charge up and the
diode will back-bias.
Unless you use a very leaky diode.
What's the power supply for?
Good question. Although it primary provides power for future, yet-to-be
developed protos, it also fulfills a fundamental function for a crystal
radio set.
Think about it.
I can't imagine why.

I had a crystal radio when I was a kid. My uncle Sheldon and I built
it, hand-wound coil, classic 365pF variable cap, built on a piece of
wood. It used a hi-z pair of dynamic headphones. It was LOUD.

Hiz dynamic phones are rare these days, but one could use a small
audio transformer with 4 ohm headphones.
Don
2024-02-10 19:01:54 UTC
Permalink
Post by John Larkin
Post by Don
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
<https://crcomp.net/crystalrx/b.png>
No worries all's not nyet
Spice fiddling's a safe bet...
Why does a crystal set need a power supply?
I don't think the circuit will work with a crystal headphone.
It wouldn't work without one. Xtal sets need hi-z earphones for
obvious reasons and crystal earpieces are ideal.
In the circuit shown, the headphone capacitance will charge up and the
diode will back-bias.
Unless you use a very leaky diode.
What's the power supply for?
Good question. Although it primary provides power for future, yet-to-be
developed protos, it also fulfills a fundamental function for a crystal
radio set.
Think about it.
I can't imagine why.
I had a crystal radio when I was a kid. My uncle Sheldon and I built
it, hand-wound coil, classic 365pF variable cap, built on a piece of
wood. It used a hi-z pair of dynamic headphones. It was LOUD.
Hiz dynamic phones are rare these days, but one could use a small
audio transformer with 4 ohm headphones.
Along with assorted audio convertors, there's at least one old Soviet
military headset here:

<https://www.ebay.com/sch/i.html?_nkw=high+impedance+earphone>
<https://www.ebay.com/itm/314895978798>

(A pariah perhaps to patriots stirred by sanction spirit.)

Danke,
--
Don, KB7RPU, https://www.qsl.net/kb7rpu
There was a young lady named Bright Whose speed was far faster than light;
She set out one day In a relative way And returned on the previous night.
Don
2024-02-14 00:43:48 UTC
Permalink
Post by John Larkin
Post by Don
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Don
Post by Jan Panteltje
Post by John Larkin
Post by Cursitor Doom
Post by John Larkin
Post by Fred Bloggs
This is the new circuit analysis: exploring operation in Spice and looking at graphs.
https://www.elektormagazine.com/articles/circuit-analysis-with-ltspice-12v-lamp-flasher
Too much talking for too little circuit.
Mr. Bloggs had to use LTspice for a 12V lamp flasher?? Jeez!
Using Spice is reasonable, easier than computing exponentials by hand.
I use Spice for voltage dividers, filters, rectifiers, RLC delays, all
sorts of stuff.
I mostly use pepper, Chili and salt as spices, some sugar etc.
If this spice stuff continues people will need it before throwing a ball ...
Why not just ask AI to design the circuit for you,?
And no guarantee that anything in LTspice works -in- or even resembles reality!
Real hands on testing is needed to get a clue.
Tis true the Spice won't help you
<https://crcomp.net/crystalrx/a.png>
<https://crcomp.net/crystalrx/b.png>
No worries all's not nyet
Spice fiddling's a safe bet...
Why does a crystal set need a power supply?
I don't think the circuit will work with a crystal headphone.
It wouldn't work without one. Xtal sets need hi-z earphones for
obvious reasons and crystal earpieces are ideal.
In the circuit shown, the headphone capacitance will charge up and the
diode will back-bias.
Unless you use a very leaky diode.
What's the power supply for?
Good question. Although it primary provides power for future, yet-to-be
developed protos, it also fulfills a fundamental function for a crystal
radio set.
Think about it.
I can't imagine why.
Post by Don
Hint: the power cord conveniently connects three cables to the case, but
not all are used by this crystal circuit.
A good crystal set needs to squeeze as much precious power as possible
out of a radio wave. A solid connection to ground fulfills a fundamental
function - it firmly anchors the signal. The NEMA ground cable provides
a handy ground connection, even though the power supply remains unused.

Anyhow, now's the time to put the previously passive power supply to
good use. It's possible to eliminate the earphone with an electronic
precision voltage reference:

<Loading Image...>

# # #

World Radio Day – February 13, 2024
<https://nationaltoday.com/world-radio-day/>

Radio: A century informing, entertaining and educating
<https://www.unesco.org/en/days/world-radio/radio-next-century>

World Radio Day 2024: How Kenya’s Community Radios are Strengthening Democracy
<https://www.unesco.org/en/articles/world-radio-day-2024-how-kenyas-community-radios-are-strengthening-democracy>

World Radio Day 2024: Theme, History, Significance & Everything Else You Must Know
<https://www.freepressjournal.in/lifestyle/world-radio-day-2024-theme-history-significance-everything-you-must-know-about>

Danke,
--
Don, KB7RPU, https://www.qsl.net/kb7rpu
There was a young lady named Bright Whose speed was far faster than light;
She set out one day In a relative way And returned on the previous night.
Jan Panteltje
2024-02-14 05:52:43 UTC
Permalink
On a sunny day (Wed, 14 Feb 2024 00:43:48 -0000 (UTC)) it happened "Don"
Post by Don
A good crystal set needs to squeeze as much precious power as possible
out of a radio wave. A solid connection to ground fulfills a fundamental
function - it firmly anchors the signal. The NEMA ground cable provides
a handy ground connection, even though the power supply remains unused.
I had crystal portable radio back then
it used a moving ferrite rod in a coil for tuning and antenna,
took it with me on long walks.

Most AM long wave to medium wave radios use a ferrite rod, but with a tuning cap.
Amazing how much signal such a -magnetic- antenna delivers,
sensitive to direction though.
Don
2024-02-14 15:58:22 UTC
Permalink
Post by Jan Panteltje
Post by Don
A good crystal set needs to squeeze as much precious power as possible
out of a radio wave. A solid connection to ground fulfills a fundamental
function - it firmly anchors the signal. The NEMA ground cable provides
a handy ground connection, even though the power supply remains unused.
I had crystal portable radio back then
it used a moving ferrite rod in a coil for tuning and antenna,
took it with me on long walks.
Most AM long wave to medium wave radios use a ferrite rod, but with a tuning cap.
Amazing how much signal such a -magnetic- antenna delivers,
sensitive to direction though.
Potato guns aside, the last rocket launched by me happened a long time
ago. If another small rocket is ever again launched by me it may carry
<https://crystalradio.net/misc/rocket/index.shtml>
Government inflation drove up a rocket radio's price 7,200% since the 1950s:

<https://www.ebay.com/itm/276326468516>

Danke,
--
Don, KB7RPU, https://www.qsl.net/kb7rpu
There was a young lady named Bright Whose speed was far faster than light;
She set out one day In a relative way And returned on the previous night.
Jan Panteltje
2024-02-15 06:18:32 UTC
Permalink
On a sunny day (Wed, 14 Feb 2024 15:40:16 -0000 (UTC)) it happened "Don"
Post by Jan Panteltje
Post by Don
A good crystal set needs to squeeze as much precious power as possible
out of a radio wave. A solid connection to ground fulfills a fundamental
function - it firmly anchors the signal. The NEMA ground cable provides
a handy ground connection, even though the power supply remains unused.
I had crystal portable radio back then
it used a moving ferrite rod in a coil for tuning and antenna,
took it with me on long walks.
Most AM long wave to medium wave radios use a ferrite rod, but with a tuning cap.
Amazing how much signal such a -magnetic- antenna delivers,
sensitive to direction though.
Potato guns aside, the last rocket launched by me happened a long time
ago. If another small rocket is ever again launched by me it may carry
<https://crystalradio.net/misc/rocket/index.shtml>
Yes that is the one!
Don
2024-02-15 15:40:07 UTC
Permalink
Post by Jan Panteltje
Post by Jan Panteltje
Post by Don
A good crystal set needs to squeeze as much precious power as possible
out of a radio wave. A solid connection to ground fulfills a fundamental
function - it firmly anchors the signal. The NEMA ground cable provides
a handy ground connection, even though the power supply remains unused.
I had crystal portable radio back then
it used a moving ferrite rod in a coil for tuning and antenna,
took it with me on long walks.
Most AM long wave to medium wave radios use a ferrite rod, but with a tuning cap.
Amazing how much signal such a -magnetic- antenna delivers,
sensitive to direction though.
Potato guns aside, the last rocket launched by me happened a long time
ago. If another small rocket is ever again launched by me it may carry
<https://crystalradio.net/misc/rocket/index.shtml>
Yes that is the one!
The TL431 needs to be repurposed into a power supply before my crystal
set can accommodate another adornment: an audio amp in the form of a
LM386. It turns out the power supply in my "bone yard" chassis supplies
only 5 VDC.

So a simple SMPS, similar to the flyback converter Win Hill reversed
engineered for Figure 9.74 in _Art of the Electronics_, needs to be
prototyped:

<https://archive.org/details/the-art-of-electronics-3rd-ed-2015_202008/page/657/mode/2up>

Applications found in both the CSC72XX and UC3843 datasheets also fit the
bill.

Danke,
--
Don, KB7RPU, https://www.qsl.net/kb7rpu
There was a young lady named Bright Whose speed was far faster than light;
She set out one day In a relative way And returned on the previous night.
Jan Panteltje
2024-02-16 06:34:12 UTC
Permalink
On a sunny day (Thu, 15 Feb 2024 15:40:07 -0000 (UTC)) it happened "Don"
Post by Don
Post by Jan Panteltje
Post by Jan Panteltje
Post by Don
A good crystal set needs to squeeze as much precious power as possible
out of a radio wave. A solid connection to ground fulfills a fundamental
function - it firmly anchors the signal. The NEMA ground cable provides
a handy ground connection, even though the power supply remains unused.
I had crystal portable radio back then
it used a moving ferrite rod in a coil for tuning and antenna,
took it with me on long walks.
Most AM long wave to medium wave radios use a ferrite rod, but with a tuning cap.
Amazing how much signal such a -magnetic- antenna delivers,
sensitive to direction though.
Potato guns aside, the last rocket launched by me happened a long time
ago. If another small rocket is ever again launched by me it may carry
<https://crystalradio.net/misc/rocket/index.shtml>
Yes that is the one!
The TL431 needs to be repurposed into a power supply before my crystal
set can accommodate another adornment: an audio amp in the form of a
LM386. It turns out the power supply in my "bone yard" chassis supplies
only 5 VDC.
So a simple SMPS, similar to the flyback converter Win Hill reversed
engineered for Figure 9.74 in _Art of the Electronics_, needs to be
<https://archive.org/details/the-art-of-electronics-3rd-ed-2015_202008/page/657/mode/2up>
Applications found in both the CSC72XX and UC3843 datasheets also fit the
bill.
Danke,
I am wondering, would not a non-switching converter (mains transformer + rectifier followed by maybe a LM317)
give less RFI?
Don
2024-02-16 17:04:17 UTC
Permalink
<snip>
Post by Jan Panteltje
Post by Don
The TL431 needs to be repurposed into a power supply before my crystal
set can accommodate another adornment: an audio amp in the form of a
LM386. It turns out the power supply in my "bone yard" chassis supplies
only 5 VDC.
So a simple SMPS, similar to the flyback converter Win Hill reversed
engineered for Figure 9.74 in _Art of the Electronics_, needs to be
<https://archive.org/details/the-art-of-electronics-3rd-ed-2015_202008/page/657/mode/2up>
Applications found in both the CSC72XX and UC3843 datasheets also fit the
bill.
I am wondering, would not a non-switching converter (mains transformer +
rectifier followed by maybe a LM317) give less RFI?
Your assessment's absolutely accurate (according to my gut). Regardless,
my tinkering involves more than one goal, and fiddling with SMPS along
the way is acceptable.

And any advice about DIY SMPS transformers from readers is appreciated in
advance.

Danke,
--
Don, KB7RPU, https://www.qsl.net/kb7rpu
There was a young lady named Bright Whose speed was far faster than light;
She set out one day In a relative way And returned on the previous night.
Jan Panteltje
2024-02-17 09:54:35 UTC
Permalink
On a sunny day (Fri, 16 Feb 2024 17:04:17 -0000 (UTC)) it happened "Don"
Post by Don
<snip>
Post by Jan Panteltje
Post by Don
The TL431 needs to be repurposed into a power supply before my crystal
set can accommodate another adornment: an audio amp in the form of a
LM386. It turns out the power supply in my "bone yard" chassis supplies
only 5 VDC.
So a simple SMPS, similar to the flyback converter Win Hill reversed
engineered for Figure 9.74 in _Art of the Electronics_, needs to be
<https://archive.org/details/the-art-of-electronics-3rd-ed-2015_202008/page/657/mode/2up>
Applications found in both the CSC72XX and UC3843 datasheets also fit the
bill.
I am wondering, would not a non-switching converter (mains transformer +
rectifier followed by maybe a LM317) give less RFI?
Your assessment's absolutely accurate (according to my gut). Regardless,
my tinkering involves more than one goal, and fiddling with SMPS along
the way is acceptable.
And any advice about DIY SMPS transformers from readers is appreciated in
advance.
It is hard to do better than the small wall-warts that you can buy for just a few dollars.
Those use advanced chips with all sorts of protections, are often CE certified
saving you a lot of work, safer too:
Loading Image...

For the switchers I designed I usually use a Microchip PIC as driver chip,
it has 2 build in hardware comparators that you can use for cycle by cycle current limiting
and it has several ADC channels, needs programming though.
As a rule of thumb for transformers: 1V per turn for 15 kHz switching frequency
for small transformers like this one:
Loading Image...
But look for inductance, core material, possible air-gap, current, DC bias, frequency, etc etc.
Or grab one from an old defective wall-wart?

Switchers have a lot of negatives, yesterday I repaired a satellite receiver box that had, among several other
regulators, a 5V switcher for the memory and CPU chips and USB memory stick.
it would abort recording on random occasons.
Super small box, the one on the right, rather new:
Loading Image...
was thinking as it got worse over time, could be a filter cap in one of the switchers.
Opened it and scoped the ripple on the filter caps:
about 1.5V RF spikes on the 5V output cap.
Replaced cap, with one from the junk box, recording works again....
Need to get some low ESR caps, and maybe bias it with a better smaller capacitor.
But even then the thing runs of a 12 V wall-wart... with overload protection.
If you want to get a bit bigger, then get an old PC power supply [transformer],...
Plenty of circuits for that online too.
I often use what I find laying about in the junk box...
Don
2024-02-18 05:14:16 UTC
Permalink
Post by Jan Panteltje
Post by Don
<snip>
Post by Jan Panteltje
Post by Don
The TL431 needs to be repurposed into a power supply before my crystal
set can accommodate another adornment: an audio amp in the form of a
LM386. It turns out the power supply in my "bone yard" chassis supplies
only 5 VDC.
So a simple SMPS, similar to the flyback converter Win Hill reversed
engineered for Figure 9.74 in _Art of the Electronics_, needs to be
<https://archive.org/details/the-art-of-electronics-3rd-ed-2015_202008/page/657/mode/2up>
Applications found in both the CSC72XX and UC3843 datasheets also fit the
bill.
I am wondering, would not a non-switching converter (mains transformer +
rectifier followed by maybe a LM317) give less RFI?
Your assessment's absolutely accurate (according to my gut). Regardless,
my tinkering involves more than one goal, and fiddling with SMPS along
the way is acceptable.
And any advice about DIY SMPS transformers from readers is appreciated in
advance.
It is hard to do better than the small wall-warts that you can buy for just a few dollars.
Those use advanced chips with all sorts of protections, are often CE certified
https://www.panteltje.nl/pub/floor_warts_IXIMG_0790.JPG
For the switchers I designed I usually use a Microchip PIC as driver chip,
it has 2 build in hardware comparators that you can use for cycle by cycle current limiting
and it has several ADC channels, needs programming though.
As a rule of thumb for transformers: 1V per turn for 15 kHz switching frequency
https://panteltje.nl/pub/new_transformer_test_setup_img_3153.jpg
But look for inductance, core material, possible air-gap, current, DC bias, frequency, etc etc.
Or grab one from an old defective wall-wart?
Switchers have a lot of negatives, yesterday I repaired a satellite receiver box that had, among several other
regulators, a 5V switcher for the memory and CPU chips and USB memory stick.
it would abort recording on random occasons.
https://panteltje.nl/pub/sat_box_sizes_IXIMG_0907.JPG
was thinking as it got worse over time, could be a filter cap in one of the switchers.
about 1.5V RF spikes on the 5V output cap.
Replaced cap, with one from the junk box, recording works again....
Need to get some low ESR caps, and maybe bias it with a better smaller capacitor.
But even then the thing runs of a 12 V wall-wart... with overload protection.
If you want to get a bit bigger, then get an old PC power supply [transformer],...
Plenty of circuits for that online too.
I often use what I find laying about in the junk box...
My mindset mostly mirrors yours. Here's the story so far:
The chassis came from my bone pile. Its NEMA grounded bulkhead
connector enables an easy connection from a proto-board to ground.
Adding an audio amp along the lines of a LM386 seemed a fitting way
to wrap up the AM receiver before moving on to FM. Unfortunately the
chassis' built-in SMPS only outputs a meager 5 VDC instead of the 12
VDC or, better yet, 24 VDC sought by me.
A quick rummage through my surplus wall warts (over a hundred in
all) unearthed a 12 VDC jewel. The plan was to liberate its board from
its plastic enclosure then swap it into the chassis in place of the
built-in. But the jewel's board didn't cooperate - its footprint was
too small to accomodate the chassis' SMPS mounts.
The jewel seems like simplicity itself. It's scarcely more
complicated than the typical 555 circuit:

<Loading Image...>
<Loading Image...>

Why not assemble a similar circuit from spares on a larger universal board
and also mod it to output 24 VDC?

# # #

Alternatively, all analysis and reverse engineering can be avoided by
using AI to design it:

PI Expert is an automated, graphical user interface (GUI)
driven program that takes power supply specifications and
automatically generates a power conversion solution designed
around the most appropriate Power Integrations’ IC family.
PI Expert provides all necessary information to build and
test a working prototype, including the full schematic and
BOM. PI Expert also provides a complete transformer design
that includes core size, number of turns, appropriate wire
selection and winding construction. Detailed winding
instructions for mechanical assembly are also generated.

<https://pi-expert-suite.software.informer.com/download/#downloading>

Danke,
--
Don, KB7RPU, https://www.qsl.net/kb7rpu
There was a young lady named Bright Whose speed was far faster than light;
She set out one day In a relative way And returned on the previous night.
Jan Panteltje
2024-02-18 06:15:08 UTC
Permalink
On a sunny day (Sun, 18 Feb 2024 05:14:16 -0000 (UTC)) it happened "Don"
Post by Don
Post by Jan Panteltje
Post by Don
<snip>
Post by Jan Panteltje
Post by Don
The TL431 needs to be repurposed into a power supply before my crystal
set can accommodate another adornment: an audio amp in the form of a
LM386. It turns out the power supply in my "bone yard" chassis supplies
only 5 VDC.
So a simple SMPS, similar to the flyback converter Win Hill reversed
engineered for Figure 9.74 in _Art of the Electronics_, needs to be
<https://archive.org/details/the-art-of-electronics-3rd-ed-2015_202008/page/657/mode/2up>
Applications found in both the CSC72XX and UC3843 datasheets also fit the
bill.
I am wondering, would not a non-switching converter (mains transformer +
rectifier followed by maybe a LM317) give less RFI?
Your assessment's absolutely accurate (according to my gut). Regardless,
my tinkering involves more than one goal, and fiddling with SMPS along
the way is acceptable.
And any advice about DIY SMPS transformers from readers is appreciated in
advance.
It is hard to do better than the small wall-warts that you can buy for just a few dollars.
Those use advanced chips with all sorts of protections, are often CE certified
https://www.panteltje.nl/pub/floor_warts_IXIMG_0790.JPG
For the switchers I designed I usually use a Microchip PIC as driver chip,
it has 2 build in hardware comparators that you can use for cycle by cycle current limiting
and it has several ADC channels, needs programming though.
As a rule of thumb for transformers: 1V per turn for 15 kHz switching frequency
https://panteltje.nl/pub/new_transformer_test_setup_img_3153.jpg
But look for inductance, core material, possible air-gap, current, DC bias, frequency, etc etc.
Or grab one from an old defective wall-wart?
Switchers have a lot of negatives, yesterday I repaired a satellite receiver box that had, among several other
regulators, a 5V switcher for the memory and CPU chips and USB memory stick.
it would abort recording on random occasons.
https://panteltje.nl/pub/sat_box_sizes_IXIMG_0907.JPG
was thinking as it got worse over time, could be a filter cap in one of the switchers.
about 1.5V RF spikes on the 5V output cap.
Replaced cap, with one from the junk box, recording works again....
Need to get some low ESR caps, and maybe bias it with a better smaller capacitor.
But even then the thing runs of a 12 V wall-wart... with overload protection.
If you want to get a bit bigger, then get an old PC power supply [transformer],...
Plenty of circuits for that online too.
I often use what I find laying about in the junk box...
The chassis came from my bone pile. Its NEMA grounded bulkhead
connector enables an easy connection from a proto-board to ground.
Adding an audio amp along the lines of a LM386 seemed a fitting way
to wrap up the AM receiver before moving on to FM. Unfortunately the
chassis' built-in SMPS only outputs a meager 5 VDC instead of the 12
VDC or, better yet, 24 VDC sought by me.
A quick rummage through my surplus wall warts (over a hundred in
all) unearthed a 12 VDC jewel. The plan was to liberate its board from
its plastic enclosure then swap it into the chassis in place of the
built-in. But the jewel's board didn't cooperate - its footprint was
too small to accomodate the chassis' SMPS mounts.
The jewel seems like simplicity itself. It's scarcely more
<https://crcomp.net/smps/a.png>
<https://crcomp.net/smps/b.png>
Nice circuit.
I like they use the TL431, it is a good reference.
Post by Don
Why not assemble a similar circuit from spares on a larger universal board
and also mod it to output 24 VDC?
Maybe not enough space for extra windings on the transformer...
but you could try...
Post by Don
# # #
Alternatively, all analysis and reverse engineering can be avoided by
PI Expert is an automated, graphical user interface (GUI)
driven program that takes power supply specifications and
automatically generates a power conversion solution designed
around the most appropriate Power Integrations’ IC family.
PI Expert provides all necessary information to build and
test a working prototype, including the full schematic and
BOM. PI Expert also provides a complete transformer design
that includes core size, number of turns, appropriate wire
selection and winding construction. Detailed winding
instructions for mechanical assembly are also generated.
<https://pi-expert-suite.software.informer.com/download/#downloading>
I wonder if as kid I was brought up with that Pi Expert stuff if I would
have chosen an other field.

Sure AI can help, but you need to know the basics.

My first 'learning' came from this book:
https://www.veron.nl/nieuws/zoo-werkt-de-radio/
'that is how radio works'
Aisberg was a very good teacher.
He later also wrote 'that is how television works'
and IIRC same for transistors...

Much later when I worked in TV an other book given to me by somebody
there went into the US NTSC color system.
And then one day I had 20 minutes to repair a converter for the French SECAM
Eurovision system.. Did it, because I had read up on it out of interest.
You need to know basics, and how to use tools.
Had a TV repair shop for several years that grew out to a bigger one...
Fast fault finding requires in dept knowledge.
I am not sure if AI would cope ;-)
I remember fault finding in an ESA ground station, in missile control equipment,
on navy ships... Where not.
But I am just a neural net, a proper trained AI could out-perform me?
work 24/7?
But then one high altitude nuke or heavy solar solar storm and no more
electrickety, we organisms are likely more durable than computahs.
All emerging from what was there? the big-bang and then before that..
Loading...