Bharat Rakshak

http://www.planetary.org/blogs/guest-bl ... ssion.html

The data is first transmitted to the Canberra Deep Space Network and from there it reaches the Indian Space Science Data Centre at Byalalu off the Bangalore-Mysore highway.

Q: what is preventing Byalalu to receive them direct?

A single station can only receive data for a period of 8 hours.

If you see the DSN transmission I am not sure all the data is sent via Canberra. For a period of time there is no transmission from the DSN stations to or from MOM probably at that time the Byalalu station takes over. For eg right now (00.55 AM 2/10/2014 IST) there are no transmissions to any JPL DSN center. I presume ti depends on the voume of data, the amount of data cached and transmissable through our transmission window and the sensitivity of the data.

by Saik


Q: what is preventing Byalalu to receive them direct?

"Once a mission gets more than 30,000 km from earth, it is always in view of at least one of the stations."

Check the picture of Deep Space network and its three locations (Canberra, Goldstone and Madrid). Depending upon the earth rotation from Wiki

http://en.wikipedia.org/wiki/NASA_Deep_Space_Network

and Indian Deep Space Network

http://en.wikipedia.org/wiki/Indian_Deep_Space_Network

OT here, but I have been trying to think along the same lines. Indian education builds silos of knowledge. Knowledge never grows in silos, it grows only when there is cross pollination from various other parallel fields. For that you need an enquring, open mind and a community of dedicated researchers. All that depends on basic engineering education in India, and its quality needs a lot of improvement. I believe online learning is the way to solve this.

For ex: Back in my engr. days I went through an entire course on Fourier Analysis without the instructor OR the students having a clue about linear algebra or the physical meaning of complex exponential. The engineering college I went to was a small/new one, and the guys teaching there were "passouts " from the previous year. It was a joke, and everyone knew it. Nowadays, if they can look at an online course from IIT or some massan univ and have a test/gradation done independently of the course by the last years "passouts". It would be a much higher quality education than what I got. In effect, they would be acting as TAs. Similarly, if the newbie lecturers could embark on an online PhD it will help the over all quality of engineering education all over India.

Disclaimer: Arm chair advice onleee

http://www.prl.res.in/~rajiv/planexnews/

How many of you were aware of the existence of this newsletter. Even ISRO does not provide a link to it, even though it is being published from its cradle, PRL A'bad.

There is a summary article on Methane search on Mars in their latest edition.
http://www.prl.res.in/~rajiv/planexnews ... 2014-3.php

so let me try again in asking the question differently,

we would need only one byalalu, if we have say covered all the NASA's 3 DSN equivalent installed by ourselves. in the sense, in my understanding, we could launch ISRO DSN satellites in Geo orbit around Earth, which can do some dual roles - globalGAGAN + DSN relays/routers.

our next mission to moon and mars should actually include DSN routing systems powered by Sun. we could drop them on strategic location on the moon and mars.

Bade wrote:http://www.prl.res.in/~rajiv/planexnews/

How many of you were aware of the existence of this newsletter. Even ISRO does not provide a link to it, even though it is being published from its cradle, PRL A'bad.

There is a summary article on Methane search on Mars in their latest edition.
http://www.prl.res.in/~rajiv/planexnews ... 2014-3.php

Thanks.

SaiK wrote:so let me try again in asking the question differently,

we would need only one byalalu, ...

One can always use more than one.. for navigational purposes (time-delay, Doppler etc..) simultaneous monitoring from more than one is necessary... more the merrier..

(Good trivia question for those who are good in math - why one requires at least 4 satellites to fix a position on GPS)

Amber G. wrote:
(Good trivia question for those who are good in math - why one requires at least 4 satellites to fix a position on GPS)

https://answers.yahoo.com/question/inde ... 559AAOoJCi

http://www.frontline.in/science-and-tec ... epage=true

Nice write up by T.S.S of Frontline, on Mangalyaan. Also, in the same issue http://www.flonnet.com, a scarcely less fine article by R.Ramachandran on Mangalyaan.

Bade: That is interesting.

I would like to draw attention to Fig 2. in that Methane summary article which gives the setup of an experiment for explaining one mechanism of how methane could have been removed from Martian atmosphere. At least to me the setup looks quite simple. Any reasonably equipped UG college in India should be able to reproduce this experiment as a semester/year end group(s) project as soon as they are done with basic chemistry/physics and some knowledge of the lab equipment and lab safety. Surprisingly this experiment was done by one Jensen in 2014. I quote from the news letter:

This appears to be one of the viable explanations for the presence of strong sinks and for the observed spatial variations. However, more theoretical calculations are needed to quantify the removal of methane by this mechanism, to see whether the mechanism can explain the observed intermittent nature of Martian methane. Apart from this, laboratory experiments using fluidized beds may also be useful to simulate such conditions (B. Sivaraman, Personal Communication).

My wild guess is that the computational requirements for the Theoretical caluclations mentioned in the paper would be modest and not out of reach for a reasonably equipped college which teaches BCAs and BSc Physics and BSc chemistry.

That is the kind of "table top experimentation" I was talking about.

Varoon Shekhar wrote:http://www.frontline.in/science-and-tec ... epage=true

Nice write up by T.S.S of Frontline, on Mangalyaan. Also, in the same issue http://www.flonnet.com, a scarcely less fine article by R.Ramachandran on Mangalyaan.

What came to the fore in the MOM was ISRO’s youth power. Deepak Panda, who took part in the simulation of the attitude control done by the eight thrusters and in the development of the electronics needed for the spacecraft’s orbit control, is only 26. The LAM and the eight thrusters boasted a closed loop guidance scheme (CLGS). “If at all any error developed in the attitude control, the CLGS would nullify it,” said Panda. Rishabh Mishra developed the software for display on the big monitors in MOX-1 and 2. The display related to the telemetry data from the spacecraft.

Kamlesh Kumar Sharma, also in his 20s, could not hide his happiness. “I sent the first command to MOM [on November 15, 2013], after it went into its earthbound orbit. I sent the first command during the spacecraft’s trans-Mars insertion on December 1, 2013,” he said. Sharma described himself as “the prime controller during the spacecraft’s MOI” on D-day. “I was also the first to verify the good news that we received” that India’s spacecraft had entered the Martian orbit, he exulted.

These guys have the right environment, the right motivation and were able to achieve what will be a very important milestone for years to come. With a young, inspired and accomplished staff, increasing budgets and a friendly government, ISRO has a bright future.

Amber G. wrote:

SaiK wrote:so let me try again in asking the question differently,

we would need only one byalalu, ...

One can always use more than one.. for navigational purposes (time-delay, Doppler etc..) simultaneous monitoring from more than one is necessary... more the merrier..

(Good trivia question for those who are good in math - why one requires at least 4 satellites to fix a position on GPS)

Did not check the link posted (am sure there are many such links google/bing searches will reveal..without difficulty..) ....but this seems like a geometry-type question (rather than math). A signal goes from the GPS to the satellite. The signal has a timestamp in it. A satellite receives the signal calculates the distance at which the signal originated (time x speed of light). This distance is the radius of a sphere where the signal originated. A second satellite receiving the same signal creates a second sphere similarly....the intersection of the two spheres is ....(perhaps) a circle (it is a closed curve of some sort). A third satellite creates a similar sphere to intersect with the circle/closed curve and you are left with 2 points that intersect the sphere. The 4 satellite's sphere will intersect one of the two points....i.e. location of GPS.

Added later: Checked the link.....the plot thickens.. GPS only receives, does not transmit......aargh! The clean, nice theory goes out of the window (but not entirely). If indeed the GPS only receives, then it is the inverse problem....

Final answer: It should not matter whether the GPS sends or receives, if it receives only, then the satellite has to broadcast its positional coordinates. This plus the signal broadcast time will give the GPS theall the info to 'plot the sphere'. All calculations i.e. intersection of the 4 spheres are done in the GPS (and not the satellite).

srikumar: if the signals are transmitted at some predetermined interval (say every nSec) and the clock of the GPS is synchronized to global time signal to which the satellites are also synced? Of course, everybody knows the location of one or more global time signal transmitters all of which are at fixed locations on earth's surface. One can even theorize about having the time beacons on other planets or on the moon or even satellites which are in geosynchronous orbits.

^^^Thanks... am not a satellite guy but I would assume the above is a requirement for the GPS to perform calculations accurately to get the 'radii of the spheres'.

GPS satellites send time signals. The accuracy of the receiver is dependent on if it is capable of decoding the transmitted time signal in milliseconds, microseconds or even more fine grained time blocks.

GPS receivers are just that - receivers. They have no broadcast ability.

Based on satellite signals, these receivers will estimate their distance from the transmitter, then position themselves on a dimensional location. This 3 dimensional location corresponds to a position on over the surface of the planet.

^^^

Interesting part about this is: (see below):
(and this is why it is an interesting Math question - Fun to figure out the answer without looking it up)

- MOM does NOT have an atomic clock. (It, in itself , can not put accurate time-stamps in the signal).
(Atomic clocks are still very expensive, MOM could not afford it- Ordinary quartz clocks, though accurate enough for everyday business can still drift about 1 part in a million - not good enough, as in 10 months or so, it may be off by a lot (up to a few minutes!) - unless it gets synchronized by atomic clocks located far away - and you have to know how far away to synchronize the MOM's clock)


- Similarly GPS in your car does NOT have an atomic clock. (So how does it measure the distance between it and the satellite accurately?)

- One ought be able to pin point the location, from ground stations, even if it can hear just a radio "beep beep" -the basic math principal is same as in above cases )

Anyway - the fun part if someone does not already know the answer, is to figure out how it can be done.

(Yes, I have used this problem to introduce geometry of hyperbola - and why one needs at least 4 satellites vs 3 even in a perfect world.. (BTW hyperbola is one type of curve in the family of conic section curves )

Shalav wrote:

Amber G. wrote:
(Good trivia question for those who are good in math - why one requires at least 4 satellites to fix a position on GPS)

https://answers.yahoo.com/question/inde ... 559AAOoJCi

I read the link, unfortunately, IMO the first answer given is sloppy (what spheres?) and really does not answer the basic core question -

(mathematically speaking it will be more accurate to say that hyperbolas, not spheres are involved here

(I specially find the following sentences fuzzy

If the receiver finds that it is x miles from one satellite, it knows that it must be somewhere on an imaginary sphere, with the satellites as the center an a radius of x. If the receiver can generate these sphere for two satellites, it knows it can only be located where the surfaces of the two spheres intersect. The two spheres overlap in a ring of possible receiver positions.

Funny part is GPS in your car, having no accurate clock can not find/calculate the radius of such sphere with necessary accuracy.
(And if one assumed that it can find three such spheres, then 3 should be enough - so why a minimum of 4)

(

matrimc wrote:srikumar: if the signals are transmitted at some predetermined interval (say every nSec) and the clocks of the GPS is synchronized to global time signal to which the satellites are also synced?

Matrimc - You raised a very interesting point - just to illustrate how advance the engineering needs to be , and how accurate we have to be, these simple ideas (sync etc) become quite complex.

"Synchronizing" is easier said than done. Just think even if the clocks are 1 feet apart (or electric wire attached to a "sync" logic is few inches..) they will go off by 1 nano second. (The clocks on GPS satellites are typically accurate better than a pico second, and for that you have to know the distance between two clocks with accuracy better than 1 mm or so!!!)

Check out, you talk about 1ns .. so if you move your GPS 1 foot behind, you will be 1 interval away..

Another element, believe it or not, as explained by Einstein, clocks on GPS satellites actually run slower (because it is orbiting the earth) than one on earth. One has to take account for that..(Yes we need Einstein calculations - the needed accuracy is such..).

Simply wow!

The frequency of the signal is probably a part of it. Assuming velocity of the satellite it known, the shift in signal frequency will heard by the receiver (due to Doppler effect) will give the 'approach' velocity relative to the GPS receiver and therefore, it's orientation relative to its velocity vector. Will try to work this angle...

This is what I was thinking.. sorry for my crude MS word drawings. Discount the number of satellites or DSN routing setups.. but just to get an idea why one Bharat based byalalu may be just enough.

(def: byalalu = DSN comm node)

Let me give it a shot (or may be just some really tangential - no pun intended) at an approach to AmberG's poser.

(I am not a satellite guy either but taking off from Srikumar's insight on spheres, circles etc.)

1. First about atomic clock: Three and half decades ago, my cousin used to develop PLL - Phase Locked Loop - synthesizers using very finely cut piezoelectric crystals and MIL spec Opamps (Operational Amplifiers) connected in a difference amplifier configuration with a feedback loop. I am not sure whether something like that work at pico-second pulses? (looks like I am off by three orders of magnitude there - I guessed nanoseconds but that seems to be not enough resolution as pointed out by AmberG)

2. Two solid cylinders intersect in a sphere, is that right? There was a puzzle about this around 1987-88 in Martin Gardener's column in Sci. Am. I wonder what would be the intersection of two solid cones? Three cones? Four cones?

3. What is the projection of the intersection of four cones onto a sphere?

Srikumar: I am assuming that every satellite will use its own carrier frequency which is adequately separated from other sources so that a (very narrow) band-pass filter can distinguish. The alternative is to carry satellite ID in the signal while all of them operating at the same carrier frequency.

1. Byalalu is a place .. it has a 100ft dia antenna used for Deep Space Network ( DSN ). Very large antenna is required because the signal comes from very long distance and is very weak so the large antenna gathers signal over a large surface ( 100 ft dia circle ).

For this to happen it is essential that the antenna face the satellite.

2. The Sun shines continuously but we receive the sunlight only when we are on the surface which faces Sun .. if we go beyond the illuminated portion the sunlight does not reach us.

Similarly signal from MOM ( or from any deep space object ) can be received by Byalalu only when MARS is visible over Indian subcontinent. For the remaining 2/3rd portion of Earth, Byalalu can't communicate with the MOM.
Therefore we have to depend on other stations which can see MARS during this period .. and they have to be on the remaining 2/3rd portion i.e. beyond Indian Territory. So we have to depend on the antennas located in other nation's territories.

Other nations may allow India to build antennas in their countries but it is highly uneconomical and staffing is also not easy. So all countries use the existing network of each other.

3. Saik's concept is a novel idea.. but realisation is very costly, much more than the cost of MOM mission for each node. Instead using the other nation's existing network is always more practicable.

"Bharat Baylalu cud be a conventional Earth station , no need of DSN antenna.



============

@ MatrimC.. PLL by itself does not give a frequency stability. It is based on the stability of the reference oscillator used in the circuit.

GPS requires 4 satellites to resolve XYZand time.(You want to place a point in 3 intersecting spheres and altitude Unless you want to assume you are on the surface of the earth.

prasannasimha wrote:GPS requires 4 satellites to resolve XYZand time.(You want to place a point in 3 intersecting spheres and altitude Unless you want to assume you are on the surface of the earth.

Prasannasimhaji and others - My question was more math/physics or fundamentals related. For example 3 intersecting spheres should be able to pinpoint a point in three dimension (including altitude). If assume that we are on the surface of earth, then only two intersecting spheres are needed (third is earth). So why do we need "at least" four instead of three?

Anyway here is basic part -

- GPS in the car (or MOM) does NOT have an atomic or very accurate clock, synchronized with standard time. (Clocks on the GPS-satellites and earth stations are very accurate - and are synchronized)

- This means that GPS in the car, though it can measure intervals of time (like a stop watch) sufficiently accurately, it can not derive the distance between satellite and the Car.

- But what it can do, is to measure time interval (difference between time signals from two different satellites), hence the difference in distance between two known points.

- The curve where this difference is constant, is well known figure, hyperbola (or hyperboloid in 3d)
. Hence from two satellite signals, GPS can calculate one hyperbola.
- Taking 3 different pairs (thus 4 are satellites are needed), one can calculate the intersection point.

(More satellite obviously give better results, because one reduce experimental error)

Amber G. wrote:

matrimc wrote:srikumar: if the signals are transmitted at some predetermined interval (say every nSec) and the clocks of the GPS is synchronized to global time signal to which the satellites are also synced?

Matrimc - You raised a very interesting point - just to illustrate how advance the engineering needs to be , and how accurate we have to be, these simple ideas (sync etc) become quite complex.

"Synchronizing" is easier said than done. Just think even if the clocks are 1 feet apart (or electric wire attached to a "sync" logic is few inches..) they will go off by 1 nano second. (The clocks on GPS satellites are typically accurate better than a pico second, and for that you have to know the distance between two clocks with accuracy better than 1 mm or so!!!)

Check out, you talk about 1ns .. so if you move your GPS 1 foot behind, you will be 1 interval away..

Another element, believe it or not, as explained by Einstein, clocks on GPS satellites actually run slower (because it is orbiting the earth) than one on earth. One has to take account for that..(Yes we need Einstein calculations - the needed accuracy is such..).

Simply wow!

About clocks. A clock is essentially an oscillator. The quality of an oscillator is defined by frequency accuracy, phase noise, and jitter. Phase noise is the noise level within a 1 Hz bandwidth at some frequency offset from the main frequency. If phase noise is too high, that is the noise level is higher in amplitude than the main frequency, the oscillator can not resolve its main frequency with accuracy. Jitter is a function of phase noise at some offset frequency. Really, no oscillator can be more accurate than its phase noise level which is governed by the thermal noise floor - unless it is cooled by an external source, remember Boltzman's Constant and temperature. Amberji can give a lesson in statistical mechanics to explain this.

The best oscillators in rank are:

1. NIST reference laser which works in a lab environment and is very expensive. Not applicable for commercial use or aerospace/defense applications.
2. Cesium atomic clock. Very accurate to the thermal noise floor. Expensive and are in GPS satellites. They need good current and voltage, and are heavy. Not applicable to MOM due to weight.
3. Rubidium clock. Accuracy 10^-11 sec. Commercially available for around $2K and are used as reference clocks for communication systems and computer networks. If I were designing MOM, I would consider using a rubidium clock.
4. Oven Controlled Crystal Oscillator (OCXO). Accuracy 10^-9 sec. Commercially available and very compact, less than $200. These are used at the board level on aircraft, missiles, and off-the-shelf test equipment. Ideal for MOM as it is light weight, requires less power and very inexpensive.
5. The remaining would be voltage controlled oscillators, crystal oscillators and PLLs. But these would be less than ideal as accuracy would be around 10^-6 to 10^-7 seconds.

SSSalvi wrote:
============

@ MatrimC.. PLL by itself does not give a frequency stability. It is based on the stability of the reference oscillator used in the circuit.

What is the resolution and stability of a very well grown and precisely cut piezoelectric crystal and extremely well made (meaning very tight tolerance) RLC + transistor circuit? May be not enough as the ref clock has to be of the resolution of the order 1e-12. I give up and am going to look up wiki/google later.

---–----------------- added later after Mort's post.

Mort: thanks. BRF has all the expertise without going to chacha.

Thanks for that ranking. So crystal controlled pll clocks are only 1e-7 at most. Those are definitely out or are they? You can have two crystal oscillator clocks out of phase to generate a 1e-12 to 1e-14 time interval. To be safe two rubidium sources then. Also one has to worry about clock skew.

AmberG and Mort: how about a waveguide probably a microwave one?

^^^
Double oven type oscillators are available commercially with a 0.5 ppb stability and ageing.
Other special purpose oscillators are Rubidium ones.
precision Osc specs are very crisp : Apart from regular ones like shape etc other parameters like Harmonic distortion, stability over various periods, Spectral purity, Phase noise start playing important role and have bias based on application.

matrimc,

Crystals will vary in the kind you get and depends on what application used for. We are talking about clocks at 10 or 100 MHz, and if you need to multiply the frequencies up, then you also multiply the error, thus you want more precision. Clock skew is generally an issue with digital circuits. What I'm talking about is in the analog world. Clock skew is controlled if you have low jitter and generally you can design a circuit to limit it.

Microwaves are modulated in pulse, frequency, amplitude, and phase. You need a reference clock to do this. Lasers are used for best timing.

SSSalvi wrote:^^^
Double oven type oscillators are available commercially with a 0.5 ppb stability and ageing.
Other special purpose oscillators are Rubidium ones.
precision Osc specs are very crisp : Apart from regular ones like shape etc other parameters like Harmonic distortion, stability over various periods, Spectral purity, Phase noise start playing important role and have bias based on application.

Spec sheets are good for sales, practice is another. You have to see what would be best for space applications like MOM where you would use a high MTBF oscillator and then design a chassis for EMI/radiation shielding. Ideally, use a precise oscillator, with redundancy, at 10 MHz that can be used as the reference for the payload instrumentation, navigation, and communication.

All - Thanks for very nice discussion. I may add something, but meanwhile let me just post some news for the record and FYI.

MAVEN completes largest Orbital Adjustment Maneuver ..

Following is image from NASA/LASP


>>>
MAVEN executed its biggest orbital maneuver to be performed as part of a series of engine burns to move the spacecraft from its highly elliptical Capture Orbit.

Now it is in its "into its science orbit" (with a low Periapsis to allow studies of the upper atmosphere and a high apoapsis to for a global picture of atmospheric activity.

Initial orbit was f 380 Km - 44,000 Km ( a period of 35.02 hours)
Used four Periapsis Lowering Maneuvers and three Period Reduction Maneuvers,
(Transition to its 4.5-hour science orbit of 150 by 6,200 Km at inclination of 74.2°)

On September 24, first Periapsis Lowering Maneuver (delta V = 8 m/s executed on the apoapsis )

Last Friday Period Reduction Maneuver (delta V = 455 m/s) (executed on periapsis - perion was 5.5 hrs)

Today (Oct 2) - reduced the apoapsis from 8000 to 6400 Km.
Next one is planned is on 10/9 ( Ready for five day science block from 10/17 to 10/22 with close encounter to the comet Siding Spring on 10/19)

It will make about 2 more Maneuvers before settling in on Nov 8.
(For details please see MAVEN page of Nasa)

.

Aren't crystals grown? or are they only naturally occurring? Whatever the source of the crystal, you are saying that 100 MHz is a physical limit even with the best possible crystals?

The question of the clock source in MOM is still unanswered. Is it a laser diode? But the shortest wavelength is 350 nm. So most probably that doesn't work.

AmberG: Are there any natural sources like Quasars or pulsars or other kinds of radiation which would be available in the vacuum of space or there is nothing but white noise in space?

^^^
^^^

True Mort, for Space based applications Mechanical,radiation and Temperature stability is prime but for me the first thing I would see was Phase Noise Purity which affected my n-PSK demodulation performance.

Hence my statement ' each user will have his specs as prime' and therefore the more the specs the merrier.

It is not a sales gimmick but is a requirement for good,clean marketing.

just putting my thoughts .. not to contradict your view.

@ matrimc

In the direction for radio sources the noise level ( frequency specific ) would increase over the background white noise.

However, the white noise level is also not constant in all directions and is also of different levels at different frequencies.

.... Crystal for oscillator puposes are CUT / GROUND to oscillate at a particular frequency. Thinner the slice higher the natural frequency. But above a certain frequency the crystal thickness becomes so small that mechanical problems start overtaking the usability and put a limit on higher possible frequencies.

The MOM clock, in my guess, is probably a an excellent quality 10 MHz OCXO on a PCB that has set reference to UTC and kept alive with a simple CMOS battery. It probably does have excellent phase noise. My guess is that it would be something like what is available from (click the link) Wenzel Associates in the US. There really aren't very many companies making these things. A few in the US, EU, & Japan. I wouldn't trust Russian sources for oscillators as they could vary from batch to batch and then you could design the rest of your system for low phase noise, but never really have it.

Thanks guys. Quite illuminating. May be we should take this Physics thread? scratch that for now. May be space thread?

-------------

GopiN: Can you start with some guidelines? I have a couple starter posts may be. Then we can go on

This GPS question and discussion has taken place at least on one other occasion. Here is a short piece from an old physics today article on the matter.
http://www.fing.edu.uy/if/cursos/fismod/GPS.pdf

The orbiter has sensors and transmitters that may sense “misbehaviour” and change over to the redundant system by itself.
http://igiri.org/health/orbiter-autonom ... n-the-day/

that means the complete path from earth to mars, including those earth orbits, trans orbits, and mars orbits are part of the normal behavior model. but the most important aspect here how does the spacecraft determine its current position given a model or derive its current position to know the deviations?

how do they do this?
i am assuming the path corrections are included here (just keeping redundancy aside for high availability needs - or this is the case of redundant dissimilar system for the same functional objective)

--------------
mil objective: the same system can be used for launch detection of global satellites from massa, russia, eu and china, and early determination of the flight path deviations meaning predicting the target destination within certain acceptable delta-time.. if path corrections, and determination of incorrect flight path corresponds to various launch profile and signature, and online validating the launch with schedule launch to quickly determine that is a first strike by an enemy nation. the second strike under nfu, launches from desh can begin while the first strike is in mid-course.