800MHz signal source...

  • Thread starter Klaus Vestergaard Kragelund
  • Start date
Mike Monett VE3BTI wrote:
Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

How do you get the exponential decay?




By switching off the incoming RF and letting the filter ring, or by
making the envelope at baseband and running it into the IF port of a
mixer.

Mixers and RF switches are also in the MCL catalogue, as are suitable
amplifiers. Old RF guys usually have a drawer full of that sort of
stuff, which is great for one-offs.

I\'ve also used medium-power amps from RF Bay, which gave great bang for
the buck.


That would work if there is only one tuned circuit in the path. What would
happen if there were multiple tuned circuits? The Radiotron Designer\'s
Handbook might give some clues.
Click to expand...

Well, one would need to make the other sections with much lower Q,
anyway. Using the mixer trick you can get whatever envelope you like.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com
 
Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

Well, one would need to make the other sections with much lower Q,
anyway. Using the mixer trick you can get whatever envelope you like.

Cheers

Phil Hobbs
Click to expand...

You need fairly high Q for multipliers and bandpass filters.

I think the most straightforward method would be to shock an 800MHz circuit
and use the exponential decay to get the desired response.



--
MRM
 
On 29/08/2022 20:19, Mike Monett VE3BTI wrote:
Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

Well, one would need to make the other sections with much lower Q,
anyway. Using the mixer trick you can get whatever envelope you like.

Cheers

Phil Hobbs

You need fairly high Q for multipliers and bandpass filters.

I think the most straightforward method would be to shock an 800MHz circuit
and use the exponential decay to get the desired response.
Click to expand...

We don\'t know the time scale for the exp decay - the t axis in the graph
posted doesn\'t state units. May not be ns but us, ms or even seconds?

piglet
 
On Monday, 29 August 2022 at 22:15:25 UTC+1, erichp...@hotmail.com wrote:
On 29/08/2022 20:19, Mike Monett VE3BTI wrote:
Phil Hobbs <pcdhSpamM...@electrooptical.net> wrote:

Well, one would need to make the other sections with much lower Q,
anyway. Using the mixer trick you can get whatever envelope you like.

Cheers

Phil Hobbs

You need fairly high Q for multipliers and bandpass filters.

I think the most straightforward method would be to shock an 800MHz circuit
and use the exponential decay to get the desired response.



We don\'t know the time scale for the exp decay - the t axis in the graph
posted doesn\'t state units. May not be ns but us, ms or even seconds?

piglet
Click to expand...
The OP did say it was an 800MHz signal.
John
 
On 29/08/2022 22:24, John Walliker wrote:
On Monday, 29 August 2022 at 22:15:25 UTC+1, erichp...@hotmail.com wrote:
On 29/08/2022 20:19, Mike Monett VE3BTI wrote:
Phil Hobbs <pcdhSpamM...@electrooptical.net> wrote:

Well, one would need to make the other sections with much lower Q,
anyway. Using the mixer trick you can get whatever envelope you like.

Cheers

Phil Hobbs

You need fairly high Q for multipliers and bandpass filters.

I think the most straightforward method would be to shock an 800MHz circuit
and use the exponential decay to get the desired response.



We don\'t know the time scale for the exp decay - the t axis in the graph
posted doesn\'t state units. May not be ns but us, ms or even seconds?

piglet
The OP did say it was an 800MHz signal.
John
Click to expand...

Yes but so what? If one takes that waveform literally it shows the wave
decaying to 1/e in six cycles. But six cycles of 800MHz takes 7.5ns yet
the t axis about that time is labelled \"50\". We are just assuming
nano-seconds. It is equally possible to imagine a 800MHz decaying at a
different rate. So I think the t axis units need to be clarified.

piglet
 
piglet <erichpwagner@hotmail.com> wrote:

We don\'t know the time scale for the exp decay - the t axis in the graph
posted doesn\'t state units. May not be ns but us, ms or even seconds?

piglet
Click to expand...

You can tell the time scale by counting the time per cycle. You can tell the
Q by counting how many cycles it takes to halve the amplitude and multiply
this number by 4.53.

See \"The ring-down method\":

https://www.giangrandi.ch/electronics/ringdownq/ringdownq.shtml



--
MRM
 
piglet <erichpwagner@hotmail.com> wrote:

The OP did say it was an 800MHz signal.
John

Yes but so what? If one takes that waveform literally it shows the wave
decaying to 1/e in six cycles. But six cycles of 800MHz takes 7.5ns yet
the t axis about that time is labelled \"50\". We are just assuming
nano-seconds. It is equally possible to imagine a 800MHz decaying at a
different rate. So I think the t axis units need to be clarified.

piglet
Click to expand...

Obviously the waveform pictured has nothing to do with the final result.
\"t\" and \"f(t)\" are textbook illustrations, not signal waveforms.



--
MRM
 
On 29/08/2022 23:13, Mike Monett VE3BTI wrote:
piglet <erichpwagner@hotmail.com> wrote:

The OP did say it was an 800MHz signal.
John

Yes but so what? If one takes that waveform literally it shows the wave
decaying to 1/e in six cycles. But six cycles of 800MHz takes 7.5ns yet
the t axis about that time is labelled \"50\". We are just assuming
nano-seconds. It is equally possible to imagine a 800MHz decaying at a
different rate. So I think the t axis units need to be clarified.

piglet

Obviously the waveform pictured has nothing to do with the final result.
\"t\" and \"f(t)\" are textbook illustrations, not signal waveforms.
Click to expand...

Exactly so. Everyone has jumped to the conclusion that the decay
ringdown is related to the 800MHz itself and decay is on a scale of
nano-seconds. If the harvested energy source was being transferred by an
800MHz oscillator powered by some mechanical event the decay could be
scaled in milliseconds, if wave power it could be scaled in seconds, if
tidal power it could be scaled in minutes. We just don\'t know.

piglet
 
On 29/08/2022 23:04, Mike Monett VE3BTI wrote:
piglet <erichpwagner@hotmail.com> wrote:

We don\'t know the time scale for the exp decay - the t axis in the graph
posted doesn\'t state units. May not be ns but us, ms or even seconds?

piglet

You can tell the time scale by counting the time per cycle. You can tell the
Q by counting how many cycles it takes to halve the amplitude and multiply
this number by 4.53.

See \"The ring-down method\":

https://www.giangrandi.ch/electronics/ringdownq/ringdownq.shtml
Click to expand...

Thanks, good to know.

piglet
 
piglet <erichpwagner@hotmail.com> wrote:

You can tell the time scale by counting the time per cycle. You can
tell the Q by counting how many cycles it takes to halve the amplitude
and multiply this number by 4.53.

See \"The ring-down method\":

https://www.giangrandi.ch/electronics/ringdownq/ringdownq.shtml

Thanks, good to know.

piglet
Click to expand...

This took a while to find, but the other method is to count the number of
cycles until the amplitude falls to 1/e (37%) of peak, then multiply by 2pi.

https://www.rp-photonics.com/q_factor.html


--
MRM
 
piglet wrote:
On 29/08/2022 23:04, Mike Monett VE3BTI wrote:
piglet <erichpwagner@hotmail.com> wrote:

We don\'t know the time scale for the exp decay - the t axis in the graph
posted doesn\'t state units. May not be ns but us, ms or even seconds?

piglet

You can tell the time scale by counting the time per cycle. You can
tell the
Q by counting how many cycles it takes to halve the amplitude and
multiply
this number by 4.53.
See \"The ring-down method\":

https://www.giangrandi.ch/electronics/ringdownq/ringdownq.shtml


Thanks, good to know.

piglet
Click to expand...

That\'s been used in laser spectroscopy for yonks as well. Thirty or so
years ago I used it with a crystal oscillator to calibrate the amplitude
digitizer of a laser microscope.

Your average free-running 80 MHz crystal decays about 1 dB per
millisecond, which is a super convenient time scale for a calibrator.

The key is to run the crystal at its series resonance, using a high-Q
parallel inductor to get rid of the electrical capacitance of the
crystal, and to pick the right topology so that a single diode switch
can remove the oscillator components from the circuit during ring-down.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top