Quantcast

Interferometry =advanced=

The Rocketry Forum

Help Support The Rocketry Forum:

DynaSoar

Well-Known Member
Joined
Mar 15, 2004
Messages
3,007
Reaction score
0
Talk of lasers elsewhere got me rethinking about an old idea.

Would it be possible to devise an interferometer using a reference laser on the ground and a target laser on a rocket, so that you could determine the speed (and given a time stamp, altitude)? There's no reason that for this purpose the target beam would have to remain narrow, would it? After all, it's the frequency shift you're measuring not the beam strength. I assume the receiver would be the limiting technical factor. What does it take to make an interferometer like this go?
 

Hospital_Rocket

Well-Known Member
Joined
Dec 25, 2003
Messages
3,990
Reaction score
1
Been a real long time since I worked with this. A couple of points I'd like to make:

Somehow I think you would have to have an optical tracking system to keep the lasers aligned.

I kind of think spin would be your worst enemy. I don't see how you would keep the phase lock with the interferometer.

You would also need some serious computing horsepower on the tracking end to keep up with the size and speed of the target. Otherwise you would need a serious Optical Theodolite to record the data for post flight analysis.

I think some sort of doppler phase shift measurement may be more in line with this. However you are still hampered by the need to track the airframe.


Of course I may be misunderstanding what you are trying to do. If I have, please clarify.
 

bobkrech

Well-Known Member
Joined
Jan 20, 2009
Messages
8,351
Reaction score
26
It's not practical, nor is it cheap.

A more cost effective method, relatively speaking, would be to use a radar or lidar (laser radar) system to measure distance (time difference) and/or velocity (doppler shift).

An even simpler system would be to measure the doppler shift of a very stable microwave signal transmitted from the rocket as a function of time. (Heterodyning is an analog form of interferometry.) This would measure the velocity vs time.

Either method would cost a few bucks. A possible starting point might be the radar system from Ramsey electronics below.

http://www.ramseyelectronics.com/cgi-bin/commerce.exe?preadd=action&key=SG7

This would require some serious modifications and a transponder in the rocket, but the theory is there in the downloadable manual.

Bob Krech
 

jwagner61

Well-Known Member
Joined
Jul 3, 2004
Messages
96
Reaction score
0
Interferometer is designed to measure distances with a resolution comparable to the wavelength of the given radiation.

For visible light this around 500 nm or 1/200th the thickness of a human hair. For microwaves, the wavelength is closer to 1 cm. This is overkill for measuring the velocity and height of a rocket since there are a lot easier ways of doing this at a fraction of the cost and complexity.:cool:
 

rstaff3

Oddroc-eteer
Joined
Jan 18, 2009
Messages
11,716
Reaction score
8
Seems like the tracking issue is a biggie for interferometry.
 

DynaSoar

Well-Known Member
Joined
Mar 15, 2004
Messages
3,007
Reaction score
0
Darn, I'd written a response in here but I must have it cancel instead of send.

Originally posted by bobkrech
It's not practical, nor is it cheap.

A more cost effective method, relatively speaking, would be to use a radar or lidar (laser radar) system to measure distance (time difference) and/or velocity (doppler shift).

An even simpler system would be to measure the doppler shift of a very stable microwave signal transmitted from the rocket as a function of time. (Heterodyning is an analog form of interferometry.) This would measure the velocity vs time.

Either method would cost a few bucks. A possible starting point might be the radar system from Ramsey electronics below.

http://www.ramseyelectronics.com/cgi-bin/commerce.exe?preadd=action&key=SG7

This would require some serious modifications and a transponder in the rocket, but the theory is there in the downloadable manual.

Bob Krech
Thanks, that's a good start. This unit doesn't go fast enough nor far enough. And it's discontinued. But at least I know things are out there. Actually I'd gotten the idea from laser speed guns.

There are laser range finders available for $150 to $200. That might be workable for altitude events, but it looks like they're single measurement devices.

Ah, here's what I'm talking about:
http://www.opticsplanet.net/osprey-...-range-with-speed-finder-camo-comet-emc4.html
600 yard range on nonreflective objects (1600 yards on reflective: Can you say "chaff"?), speed to 1000 MPH.
$179. I'm betting that's way cheaper than trying to build one.
 

rstaff3

Oddroc-eteer
Joined
Jan 18, 2009
Messages
11,716
Reaction score
8
One of these, initiated at apogee, and *lots* of rolls of aluminum foil to cover the launch field. I bet it would be tough to get a reading from a poof of chaff at apogee.
 

bobkrech

Well-Known Member
Joined
Jan 20, 2009
Messages
8,351
Reaction score
26
dynasoar

You missed my point. If you want continuous readings, you have put a transmitter in the rocket, and throw away the transmitter and the counting electronics in the radar gun, and record the gun's audio output to a tape recorder or computer. You don't use the radar gun as the transmitter, that's located in the rocket, you simply use the analog receiving electronics and the audio output. Heterodyne detection is an interferometric technique since you are beating two frequencies together and generating a difference frequency where 1 Hz ~ 11.5 cm/second ~ 4.5 "/second ~ 0.25 mph. (Did you call Ramsey, there is no indication that the unit has been discontinued that I saw on their website.)

At 2.6 GHz, the Doppler shift for the radar gun is 7.7 Hz/mph. and for my simplified method is more exactly 7.7/2 = 3.85 Hz/mph. If you have a stable oscillator on a 2.6 GHz transmitter in the rocket, you manually tune the gun's receiver to get a 1000 Hz tone (or what ever frequency you like) output from the heterodyne circuit in the radar gun. The radar gun receiver is then placed at the bottom of the launch rod pointing up.

At lift off the frequency output of the receiver will change smoothly as a function of the rockets velocity. Lets use the example of a rocket that goes as constant aceleration from 0 --> 1000 mph in 2 seconds.

If you record the audio tone output into the CD quality audio input to a PC (sampling at 44 KHz), you would hear the tone increase it's frequency from 1000 Hz to 3850 +1000 = 4850 Hz in 2 seconds, and then decreases back to 1000 Hz as the rocket reaches apogee. Upon descent, the tone will drop below 1000 Hz s the rocket descends towards the receiver. You really shouldn't need any tracking for this doppler method of velocity determination.

You should be able to get very accurate velocity data (probably +/= 2 mph or so, from this method if you do the correct mathematical analysis to the recorded data.

You could even use commercial 24 dB gain flat antennas or a turnstile antenna and an off the shelf data preamplifier as the front end to the electronics.

Measuring the doppler velocity is much simpler than using conventional radar since you don't need tracking or fancy electronic circuitry. You could use a laser (lidar) system, but you still have to heterodyne the output, and you will need tracking. It's a much harder problem.

I have a fair bit of knowledge about lidar systems and interferometers, having built a $700,00 system for a homeland secuirty application I can't discuss and an infrared interferometer for a DOD application. The off-the-shelf laser rangefinder solution won't work for anything other than apogee detection without a complete revamping of the electronics, and you would have to accurately track it to have a chance of getting rather poor quality ranging data.

Bob Krech
 

cls

Well-Known Member
Joined
Jan 21, 2009
Messages
2,217
Reaction score
0
If you have a stable oscillator on a 2.6 GHz transmitter i
given the desired resolution, where do you get a 2.6Ghz +/- 1pbb (3.85Hz) oscillator?
 

rstaff3

Oddroc-eteer
Joined
Jan 18, 2009
Messages
11,716
Reaction score
8
You don't need one. You receive the tone and discriminate the Doppler from that. You'd have to have some level of accuracy in the receivers reference.
 

cls

Well-Known Member
Joined
Jan 21, 2009
Messages
2,217
Reaction score
0
let me be more clear: I bet the tx and rx frequencies together will drift more than 1ppb. the total drift limits the possible accuracy of the measurement, making the error too large for this application.
 

rstaff3

Oddroc-eteer
Joined
Jan 18, 2009
Messages
11,716
Reaction score
8
I agree, for what it's worth. There are ways of syncing clocks, but we are getting way our of line with a hobby project IMO. Anyway its not easy/possible with gear that is not set up for it.
 

Hospital_Rocket

Well-Known Member
Joined
Dec 25, 2003
Messages
3,990
Reaction score
1
Originally posted by cls
given the desired resolution, where do you get a 2.6Ghz +/- 1pbb (3.85Hz) oscillator?

About the only way I could see it would be to lock the LO to a Cesium or Rubidium standard. I'm not sure it's necessary though.

If I understand the direction the thread is going we would have a Microwave transmitter on board (probably have to be a Gunn diode or similar to create the RF energy). Then a microwave receiver is tuned to the frequency of the transmitter and upon launch the reciever discriminates the doppler phase shift leaving an audio tone that can be measured to determine speed.

Ok

Some points.

3+ GHZ is very line of sight. Any small oscillator I have worked with required a resonance cavity that almost always terminated in a waveguide. The net result was a directional beam. The antenna theory to create an omnidirectional radiator for this is way beyond me.

in any case, I would probably choose X-Band (1CM) as the transmission line components will be smaller than anything in the S band (10cm) as is being discussed.

You could probably gut the oscillator components out of a broke radar detector. or you can try here.

Of course you would need to license this. I don't think you want the BATF and the FCC chasing you.

Somehow "Radio Free Dynasoar" just does not sound like the basis for a solid grass roots movement....

A moderately good piece on the theory of Gunn oscillators can be found here.

I was wondering when we would get a real hard core tech thread here;)
 

DynaSoar

Well-Known Member
Joined
Mar 15, 2004
Messages
3,007
Reaction score
0
Originally posted by rstaff3
One of these, initiated at apogee, and *lots* of rolls of aluminum foil to cover the launch field. I bet it would be tough to get a reading from a poof of chaff at apogee.
Oh I don't know. I pumped about 20 in^3 of chaff out of my Renegade-X at 1600' over the weekend. Didn't glitter worth a darn but made a farily visible cloud about one or two degrees across. Could ceartinly hit it. Getting a reading back, might be worth testing.

Whew, another fast mover thread. I'm going to have to go back over this all.
 

Hospital_Rocket

Well-Known Member
Joined
Dec 25, 2003
Messages
3,990
Reaction score
1
Originally posted by DynaSoar
Oh I don't know. I pumped about 20 in^3 of chaff out of my Renegade-X at 1600' over the weekend. Didn't glitter worth a darn but made a farily visible cloud about one or two degrees across. Could ceartinly hit it. Getting a reading back, might be worth testing.

Whew, another fast mover thread. I'm going to have to go back over this all.
The only thing about a chaff deployment is that I would be mildly concerned about any ecological or littering impact. I know we have a lot of discussion concerning the biodegradeability of wadding, and with that in mind, I think it might be a bit weird for you go through that kind of care to deploy a wad of aluminum foil at apogee.

And on a real non-sequitur....

I realize all this stuff about miniature microwave oscillators may concern Dick and the MDRA Cowboys.

Fear not, I found a suitable microwave source for you guys...


Photo swiped with no remorse from WikPedia

I figure it weighs in at about 650 pounds ~ Nopthing a couple of EX P motors can't deal with....

A
 

bobkrech

Well-Known Member
Joined
Jan 20, 2009
Messages
8,351
Reaction score
26
I think you guys are trying to make it too complicated.

You should be able to use a lot of standard-off-the-shelf 2.6 GHz communications gear for WALAN applications as the backbone of a doppler system.

You can a use crystal controlled frequency generator to generate the carrrier frequency and a wide band amplifier to transmit it from the rocket, and a crystal controlled frequency generator in the receiver on the ground to decode it.

Another way you could do it is to have a transmitter on the ground, a transceiver on the rocket, and a second receiver on the ground. The transceiver receives on one frequency and transmits on another separated by a fixed amount just like a ham or commercial repeater does. This type of system requires no special stability requirements what so ever provided you have decent AFC circuitry in the ground receiver as the doppler velocity information is proportional to the AFC error correction voltage.

This is a ham problem and I don't think the solution is all that difficult. If Bill Schworer's around he could add a lot of useful comments.

But then again, if the object of the exercise is to measure the speed and altitude of the rocket, a GPS unit such as the GPS flight system that will directly measure speed below 60,000 ft under most conditions coupled to an 8 channel RDAS flight computer using Dave Schultz's Kalman filter along with a 3 axis accelerometer system (for locating the earths g vector as well as the rockets acceleration) and a 3 axis mad sensor (for rocket orientation information) and a modified Adept apogee sensor for pressure reading above 37 KFT to 130 KFT would give excellent data to 120 KFT to determine the flight trajectory of the rocket.

If Mach number is sufficient for speed then the solution is very simple. Checkout this reference for details of a Mach meter near the bottom of the article.

http://www.womanpilot.com/past issue pages/2000 issues/jan feb 2000/airspeed.htm

Bob Krech
 

DynaSoar

Well-Known Member
Joined
Mar 15, 2004
Messages
3,007
Reaction score
0
Originally posted by Hospital_Rocket
The only thing about a chaff deployment is that I would be mildly concerned about any ecological or littering impact. I know we have a laot of discussion concerning the biodegradeability of wadding, and with that in mind, I think it might be a bit weird for you go through that kind of care to deploy a wad of aluminum foil at apogee.
Wasn't aluminum foil. Was cellophane glitter. It's biodegradeable. I got the MSDS on it from the maker before I bought it.
 

Hospital_Rocket

Well-Known Member
Joined
Dec 25, 2003
Messages
3,990
Reaction score
1
There is a neat idea Bob. I never considered using the rf source out of a WLAN access point. Needs some thought. My initial concern right off the bat would be scavenging the actual RF cirrcuit. Since most WLAN is spread spectrum and some, actually use a frequency hopping scheme (I belive Proxim employed this). It might be rather challenging to get the fixed frequency you need for a doppler system.
 

Zack Lau

Well-Known Member
Joined
Feb 13, 2009
Messages
538
Reaction score
0
Stability is a significant problem for hams on these bands.
At 10 GHz, even a short term stability of 1 kHz can be difficult to achieve. Time period is the length of the flight. This is 10e-7, and doesn't take into account G forces that a rocket payload must handle. If size and weight aren't important, one can phase lock your microwave source to a Z3801A GPS locked reference. This is a quality HP 10811 SC cut crystal oscillator (10MHz) phase locked to GPS--available on the surplus market for $250. I think it would be easier to fly a porta potty than to fly a powered Z3801A as a payload.
Gunn diode oscillators are limited by the quality of the cavity, as well as thermal considerations. A stability of 100 kHz would be exceptional at 10GHz.
The Ramsey unit claims an accuracy of better than 1%, so its frequency stability is probably on the order of 10e-3.
About 10 years ago I built a high power 10GHz radio system that could have been used for rocket radar--2 watts and a 3 dB noise figure to a pair of 16 inch dish antennas. No TR switching--the leakage between the deep dishes provided the oscillator signal for receive hetrodyning. All I'd have to do would be to listen to the mixer diode connection at audio. It would be cheaper to build the system today--you can buy a 2 watt 10GHz amplifier from www.downeastmicrowave.com for just $350.
 

Zack Lau

Well-Known Member
Joined
Feb 13, 2009
Messages
538
Reaction score
0
I found a page with Z3801A specifications

http://nms.lcs.mit.edu/~dga/time/z3801.pdf

5"x12"x14"
weight: 3.6kg
<25 watt power consumption, +27 and -54V supplies

However, if you really intend to fly a reference oscillator, it makes more sense to just fly the HP 10811 without the GPS locking--I've seen these go for around $100 on Ebay. This would significantly cut the size and weight.
 

bobkrech

Well-Known Member
Joined
Jan 20, 2009
Messages
8,351
Reaction score
26
Al

I did not mean to use the RF modem, just a wide band power amp which are pretty cheap. You would have to make the oscillator and feed it's output to the amplifier.

Zack

You can get very stable ppb OXCO such as the one here http://www.ecmelectronics.co.uk/pdf/vf1000.PDF

You could use this type component as the clock in a microwave frequency synthesizer to obtain ppb stability for a microwave transmitter.

To all

In thinking about the problem some more, I think I may have come up with a $100 solution, but I'm not sure. It would involve 4 FRS radio, 2 on the ground and 2 on the rocket.

You would put a 2 KHz tone into the microphone in radio 1 on the ground transmitting on any one of the lower channels 1 to 7;

The signal from radio 1 would be received by radio 2 in the rocket, and the earphone output from radio 2 would be connected to the microphone input of radio 3 in the rocket transmitting on any one of the upper channels 8-14.

Radio 4 is on the ground and receive the signal from radio 3.

Since the radios transmit FM, I believe the tones would be shifted by the doppler and the tone frequency would change as a function of the rocket velocity, but I don't know enough about FM theory to say this with certainty.

If this were true, you could take the earphone output from radio 4 and digitize it into the audio input of a PC. If you then run a windowed FFT (128, 256 or 512 points) across the data file, you should be able to obtain the velocity vs time with pretty good spatial resolution.

FRS operates from 462 to 467 MHz, so a 2 way shift gnerates a resoluion of 12.7" per Hz. So if I'm correct (that's a big if), then if the rocket were traveling at 1000 fps, the tone should shift from about 2 KHz to approximately 1 KHz.

Since the radios have AFCs, you don't have to worry about RF frequency stability, and the information would be in the slope function of the demodulated signal (I think).

Am I correct on my asumptions.

Bob Krech
 

Zack Lau

Well-Known Member
Joined
Feb 13, 2009
Messages
538
Reaction score
0
Yes, that little SC cut 10MHz crystal oscillator should do just fine--I wouldn't mind have a few of them myself :)

However, the received audio is not significantly changed, as you suggest. If this were so, hams couldn't use conventional 2M FM radios for school to astronaut communications--the audio would sound too weird. The worst case doppler for ISS on 146MHz is typically +/- 3.5 kHz.

http://www.arrl.org/ARISS/
Details on the SAREX program.
 

Zack Lau

Well-Known Member
Joined
Feb 13, 2009
Messages
538
Reaction score
0
I might add that receiving on one FRS radio while transmitting on another right next to is not easy--the strong nearby transmitter will overload the receiver unless additional filtering is added. I looked up a representative filter by Celwave. Six cavities with 60 dB of isolation that weighs 4.5 lbs. It measures 2"x8.5"x11" inches. The lower power of FRS is likely to be offset by the low battery consumption radio--you may need even more isolation out of your filter with a cheap FRS radio than an expensive land mobile transceiver designed for full duplex service. The latter often features excellent shielding not found on consumer electronics--you may need to put the radios in shielded enclosures with proper filtering for the supply, signal, and control leads.
 

bobkrech

Well-Known Member
Joined
Jan 20, 2009
Messages
8,351
Reaction score
26
Zack

I was afraid that the FM audio probably would not be effected as much as the carrier frequency.

In talking to ISS, the horizon distance is about 1230 km assuming a 360 km orbit, so for 146 MHz I calculate a maximum doppler shift of about 2.3 KHz assuming LOS and no skip. With the AFC on current electronics, you wouldn't even notice it.

I think you are overestimating the level of isolation required between the FRS radios. The lower 7 frequencies at 462 MHZ are separated by 5 MHz from the upper 7 channels at 467 MHz. That's a pretty big frequency separation which I believe was done to permit future repeater operation and/or duplex operations which is allowed on GMRS which shares some of the frequencies.

Sound like to do my experiment, you have to have a ham license and do it on 420-450 Mhz band with a CW tranmitter on the ground, a wideband repeater/transceiver in the rocket and a receiver with a BFO on the ground.

As I stated before, I think I'd stick to an RDAS or ARTS running at 200 Hz and use the accelerometer and pressure data to get the velocity. It's simple and accurate.

Bob Krech
 

Zack Lau

Well-Known Member
Joined
Feb 13, 2009
Messages
538
Reaction score
0
Hi Bob,
I was just pointing out the typical technology used to obtain full duplex GMRS operation--they use a big heavy filter made out of metal that occupies a rather large volume.
http://www.arrl.org/news/features/2004/10/06/1/?nc=1
This story describes how hams actually do full duplex with satellites--the repeaters are crossband--I don't think hams have ever put a single band repeater in orbit.
However, a non-ham bird, ATS-3 does have a 135 to 149MHz repeater in near geostationary orbit.
http://radioscanning.wox.org/Scanner/miscfreqs/some_satellite_frequencies.htm
I have noted that a lot of experimenters have what I call "microphobia," an irrational fear of working with microwaves.
In practice, working at 2GHz may actually be easier than working at 450MHz. The reality is that s-parameters used for characterizing modern semiconductor devices work much better for 2GHz devices than 450MHz devices (you run into infinity/zero problems when doing measurements, so the manufacturers often don't characterize devices below 1GHz.) But the frequency is too high to ignore device parasitics (non-ideal behavior) like lower frequencies.
 
Top