EME Moon Bounce Query

On Tue, 18 Feb 2014 10:43:51 -0800 (PST), makolber@yahoo.com wrote:

> >The 3dB (Rx) bandwidth is 22MHz. > 
>
>Really?
>
>Why use such a wide BW?
>
>Mark

In radar, range resolution is proportional to bandwidth, so maybe that
was a consideration.


Eric Jacobsen
Anchor Hill Communications
http://www.anchorhill.com

On 2/17/2014 10:08 PM, Eric Jacobsen wrote:
> On Mon, 17 Feb 2014 15:33:31 -0800 (PST), clay@claysturner.com
> wrote:
>> I've managed to find some of the details on the Haystack radar as
>> used for Shapiro's experiments for Venus.
>>
>> Frequency 7750MHz (3.8cm)
>>
>> 120 foot diameter dish - cassegrain inside of a radome!
>>
>> Antenna gain is 66.1dB
>>
>> Angular resolution is 0.4 arcseconds (half power points) At this
>> resolution you can pick out parts of Venus!
>>
>> Transmit power is 500 kW using two parallel klystrons
>>
>> Transmitted signal is a coded pulse of 10 to 20 minutes duration.
>> I'm trying to find the coding that was used.
>>
>> The receiver is liquid helium or nitrogen chilled and uses a maser
>> amplifier
>>
>> The 3dB bandwidth is 22MHz.
>>
>> Not bad for 1964 - 1968
>>
>> Clay
>
>
> Nice!   I suspected it'd have to be something really gnarly like
> that. 500kW with 66dBi gain.  123dBW EIRP!!  Woot!
>
> I can also see why the extremely long pulse was needed for both
> power and ambiguity resolution.    Now I'm wondering how they
> integrated that in the receiver?
>
> Very cool stuff!
>
>
> Eric Jacobsen Anchor Hill Communications http://www.anchorhill.com

This make me wonder what a link budget our neighbors at Alpha Centauri
(Robinson family perhaps?) would look like with this setup...

Assuming a receive antenna with the same gain and a 4.3 light-year
distance, I calculate a 250.2 dB path loss.

Since we're transmitting +87dBm, the received signal at Alpha Centauri
is -163 dBm.

That's actually higher than I expected.

It won't be a fast conversation but with a 8.6 year round trip, who cares?

Rob.

On 19/02/14 04:51, Eric Jacobsen wrote:
> On Tue, 18 Feb 2014 10:43:51 -0800 (PST), makolber@yahoo.com wrote:
>
>>> The 3dB (Rx) bandwidth is 22MHz. >
>>
>> Really?
>>
>> Why use such a wide BW?
>>
>> Mark
>
> In radar, range resolution is proportional to bandwidth, so maybe that
> was a consideration.

Also consider the now-standard "counterintuitive noise reduction
miracles" worked by a high bandwidth spread-spectrum signal plus a
matched filter. That was an advanced military-grade technique back
in the 60s.

On Wednesday, February 19, 2014 4:07:00 AM UTC-5, Tom Gardner wrote:
> On 19/02/14 04:51, Eric Jacobsen wrote: > On Tue, 18 Feb 2014 10:43:51 -0800 (PST), makolber@yahoo.com wrote: > >>> The 3dB (Rx) bandwidth is 22MHz. > >> >> Really? >> >> Why use such a wide BW? >> >> Mark > > In radar, range resolution is proportional to bandwidth, so maybe that > was a consideration. Also consider the now-standard "counterintuitive noise reduction miracles" worked by a high bandwidth spread-spectrum signal plus a matched filter. That was an advanced military-grade technique back in the 60s.

as discussed recenlty in another thread, SS imparts no advantage against white noise.

and I would think that on the interplanetary size and distance scale, a few kHz BW would give more than enough resolution?


Mark

On Wed, 19 Feb 2014 09:22:04 -0800 (PST), makolber@yahoo.com wrote:

>On Wednesday, February 19, 2014 4:07:00 AM UTC-5, Tom Gardner wrote:
>> On 19/02/14 04:51, Eric Jacobsen wrote: > On Tue, 18 Feb 2014 10:43:51 -0=
>800 (PST), makolber@yahoo.com wrote: > >>> The 3dB (Rx) bandwidth is 22MHz.=
> > >> >> Really? >> >> Why use such a wide BW? >> >> Mark > > In radar, ran=
>ge resolution is proportional to bandwidth, so maybe that > was a considera=
>tion. Also consider the now-standard "counterintuitive noise reduction mira=
>cles" worked by a high bandwidth spread-spectrum signal plus a matched filt=
>er. That was an advanced military-grade technique back in the 60s.
>
>as discussed recenlty in another thread, SS imparts no advantage against wh=
>ite noise.
>
>and I would think that on the interplanetary size and distance scale, a few=
> kHz BW would give more than enough resolution?
>
>
>Mark

I know, it seems a little odd.   They may have wanted/needed good time
resolution to discern the gravity effects, though, which could explain
it.


Eric Jacobsen
Anchor Hill Communications
http://www.anchorhill.com

On 19/02/14 17:22, makolber@yahoo.com wrote:
> On Wednesday, February 19, 2014 4:07:00 AM UTC-5, Tom Gardner wrote:
>> On 19/02/14 04:51, Eric Jacobsen wrote: > On Tue, 18 Feb 2014 10:43:51 -0800 (PST), makolber@yahoo.com wrote: > >>> The 3dB (Rx) bandwidth is 22MHz. > >> >> Really? >> >> Why use such a wide BW? >> >> Mark > > In radar, range resolution is proportional to bandwidth, so maybe that > was a consideration. Also consider the now-standard "counterintuitive noise reduction miracles" worked by a high bandwidth spread-spectrum signal plus a matched filter. That was an advanced military-grade technique back in the 60s.
>
> as discussed recenlty in another thread, SS imparts no advantage against white noise.

If the transmitter is peak power limited but not mean power limited
then there is an advantage to being able to have a longer pulse: you
get more power-per-pulse in the aether. More power means an
increased range, but longer pulse implies a reduced range resolution.

Chirped radars (where chirping is effectively a very simple form of fast
frequency hopping SS) "recover" the range resolution by having
a receiver filter that matches the transmitted chirp profile.
The matched filter "de-smears" the original long
transmitted pulse into a much narrower and stronger pulse. The noise,
not matching the filter, "destructively interferes" with itself. Hence
you do get a processing gain over white noise.

I strongly suspect the transmitters in the moon/planet bounce systems
were peak power limited but not mean power limited.

I have no idea whether chirping techniques were used, nor other
forms of FFHSS.

For examples, see
http://en.wikipedia.org/wiki/SHARAD
http://www.radartutorial.eu/19.kartei/karte111.en.html
http://www.radartutorial.eu/08.transmitters/Intrapulse%20Modulation.en.html
and theory at
http://alcatel-lucent.com/bstj/vol39-1960/articles/bstj39-4-745.pdf

On Wed, 19 Feb 2014 18:34:43 +0000, Tom Gardner
<spamjunk@blueyonder.co.uk> wrote:

>On 19/02/14 17:22, makolber@yahoo.com wrote:
>> On Wednesday, February 19, 2014 4:07:00 AM UTC-5, Tom Gardner wrote:
>>> On 19/02/14 04:51, Eric Jacobsen wrote: > On Tue, 18 Feb 2014 10:43:51 -0800 (PST), makolber@yahoo.com wrote: > >>> The 3dB (Rx) bandwidth is 22MHz. > >> >> Really? >> >> Why use such a wide BW? >> >> Mark > > In radar, range resolution is proportional to bandwidth, so maybe that > was a consideration. Also consider the now-standard "counterintuitive noise reduction miracles" worked by a high bandwidth spread-spectrum signal plus a matched filter. That was an advanced military-grade technique back in the 60s.
>>
>> as discussed recenlty in another thread, SS imparts no advantage against white noise.
>
>If the transmitter is peak power limited but not mean power limited
>then there is an advantage to being able to have a longer pulse: you
>get more power-per-pulse in the aether. More power means an
>increased range, but longer pulse implies a reduced range resolution.

Klystrons can be operated as Class A, B, AB or C, but my understanding
is that they usually (as in almost always) have been used in
saturation as a Class C amp.   I haven't gone through the HayStack
stuff closely enough to know how they used it, but I'd be surprised if
it wasn't Class-C, especially back in those days.

>Chirped radars (where chirping is effectively a very simple form of fast
>frequency hopping SS) "recover" the range resolution by having
>a receiver filter that matches the transmitted chirp profile.
>The matched filter "de-smears" the original long
>transmitted pulse into a much narrower and stronger pulse. The noise,
>not matching the filter, "destructively interferes" with itself. Hence
>you do get a processing gain over white noise.

It's the usual process of getting processing gain using correlation
with a known waveform.   Chirps are useful from many perspectives,
especially in that time frame, in that they're not difficult to
generate, the bandwidth is easily controlled, there does not need to
be any AM so they work well with a Class-C amp, and the dispersive
filters used for  demodulation are practical to construct in analog
implementations.

>I strongly suspect the transmitters in the moon/planet bounce systems
>were peak power limited but not mean power limited.
>
>I have no idea whether chirping techniques were used, nor other
>forms of FFHSS.
>
>For examples, see
>http://en.wikipedia.org/wiki/SHARAD
>http://www.radartutorial.eu/19.kartei/karte111.en.html
>http://www.radartutorial.eu/08.transmitters/Intrapulse%20Modulation.en.html
>and theory at
>http://alcatel-lucent.com/bstj/vol39-1960/articles/bstj39-4-745.pdf
>

Eric Jacobsen
Anchor Hill Communications
http://www.anchorhill.com

Eric Jacobsen <eric.jacobsen@ieee.org> wrote:
> On Wed, 19 Feb 2014 18:34:43 +0000, Tom Gardner

(snip)

>>If the transmitter is peak power limited but not mean power limited
>>then there is an advantage to being able to have a longer pulse: you
>>get more power-per-pulse in the aether. More power means an
>>increased range, but longer pulse implies a reduced range resolution.
 
> Klystrons can be operated as Class A, B, AB or C, but my understanding
> is that they usually (as in almost always) have been used in
> saturation as a Class C amp.   I haven't gone through the HayStack
> stuff closely enough to know how they used it, but I'd be surprised if
> it wasn't Class-C, especially back in those days.

Hmmm. I wouldn't have thought of them as any of those classes.

They start with a DC beam, velocity modulate it in the input cavity,
then it goes through a drift tube (no field) where, due to velocity
differences, the electrons bunch. That is, turn velocity difference
into density differences. Then, going through the second cavity, 
the density difference generates the output signal. 

Because of the velocity modulation, the time delay from input
to output isn't constant. As far as I know, you either run them
as an amplifier with an external source as input, or with feedback
to make an oscillator. I have never known them to be used in 
push-pull form, as I would have expected for class B or AB.
(At least that is how you make audio amplifiers B or AB.)

As you increase the input signal, the velocity modulation increases,
and, as it goes through the drift eventually the front of the bunch
will pass the back of the bunch. That seems somewhat different than
the usual class C definition, but maybe.

On the other hand, they are often used pulsed. The SLAC klystrons
run at 0.1% duty cycle, 20MW pulses or 20kW average, pulsed at 360Hz.
(The result of using a three phase power supply.)  They generate
a fairly loud 360Hz audio signal when running. 

>>Chirped radars (where chirping is effectively a very simple form of fast
>>frequency hopping SS) "recover" the range resolution by having
>>a receiver filter that matches the transmitted chirp profile.
>>The matched filter "de-smears" the original long
>>transmitted pulse into a much narrower and stronger pulse. The noise,
>>not matching the filter, "destructively interferes" with itself. Hence
>>you do get a processing gain over white noise.
 
> It's the usual process of getting processing gain using correlation
> with a known waveform.   Chirps are useful from many perspectives,
> especially in that time frame, in that they're not difficult to
> generate, the bandwidth is easily controlled, there does not need to
> be any AM so they work well with a Class-C amp, and the dispersive
> filters used for  demodulation are practical to construct in analog
> implementations.

Klystrons have resonant cavities for the input and output, but
I don't know how high the Q is. You can presumably modulate them
somewhat around resonance. 

-- glen

On Wednesday, February 19, 2014 1:07:00 PM UTC-5, Eric Jacobsen wrote:
> On Wed, 19 Feb 2014 09:22:04 -0800 (PST), makolber@yahoo.com wrote:
> 
> 
> 
> >On Wednesday, February 19, 2014 4:07:00 AM UTC-5, Tom Gardner wrote:
> 
> >> On 19/02/14 04:51, Eric Jacobsen wrote: > On Tue, 18 Feb 2014 10:43:51 -0=
> 
> >800 (PST), makolber@yahoo.com wrote: > >>> The 3dB (Rx) bandwidth is 22MHz.=
> 
> > > >> >> Really? >> >> Why use such a wide BW? >> >> Mark > > In radar, ran=
> 
> >ge resolution is proportional to bandwidth, so maybe that > was a considera=
> 
> >tion. Also consider the now-standard "counterintuitive noise reduction mira=
> 
> >cles" worked by a high bandwidth spread-spectrum signal plus a matched filt=
> 
> >er. That was an advanced military-grade technique back in the 60s.
> 
> >
> 
> >as discussed recenlty in another thread, SS imparts no advantage against wh=
> 
> >ite noise.
> 
> >
> 
> >and I would think that on the interplanetary size and distance scale, a few=
> 
> > kHz BW would give more than enough resolution?
> 
> >
> 
> >
> 
> >Mark
> 
> 
> 
> I know, it seems a little odd.   They may have wanted/needed good time
> 
> resolution to discern the gravity effects, though, which could explain
> 
> it.
> 
> 
> 
> 
> 
> Eric Jacobsen
> 
> Anchor Hill Communications
> 
> http://www.anchorhill.com

For the gravity dilation experiments, they were looking for a dilation of around 200uSec and they ended measuring to within around 10uSec.

These days with interplanetary probes (both orbiting a planet and/or sitting on the planet's surface) a transponder method is now used to measure the distances.

The moon's distance is measured optically (via a retroreflector left there by Armstrong & Aldrin) to fractions of a meter!

Some of the planet's distances are measured to within a km or so!

The Haystack microwave system had three modes of operation. 

1) very short duration (100 - 200 uSec) pulses for the moon mapping

2) a long "cw" signal with phase modulated pulse compression for Venus, Mercury, and Mars. This sometimes had pulse rates of up to 60Hz.

3) a radiometry mode (no transmission at all, and you just listen to the target). This is how you measure the object's temperature.

The Haystack's klystrons were driven with a very precise frequency source.

The 22MHz wideband spec is for the receiving maser amplifier. The signal is narrowed after that as needed.

When JPL first echoed Venus in 1961, it took them a year to process the data!


Clay