Craft (version from trans-GSO-> direct lunar flight)
The same craft with paload adapter
Original version of the craft (5 solid engines)
Craft with paload adapter
payload adapers for different configuration of the cafts (that adapter takes 18
month delay for adaptation satellite to launch vehicle)
Main impulse (for flight from trans-GSO -direct lunar flight)
frame adapter for main impulse engine
brake solid state engine
frame for brake engine with compensation spring
Frame for rover and engine(s)
pyrolock for brake engine
LEO -> direct flight main engine
frame for main engine
Mar 6, 2013 Now, after communication system design done, it is the
time to place everything together
In development Craft's /CubeSat's Mission Control. Follow link below.
ON APRIL 20, 2013 11:46 PM Overview of mission control
Last week's changes
ON MARCH 26, 2013 04:18 PM Radiation according horoscope.
Thanks Boris - crystal diffraction of heavy particles can make some "pockets" with low probability of radiation. And also thanks for another idea - stamped, laminated surface of gold leaf in the form of reflective concave mirror can achieve the same. In that case radiation protection for electronics can be done inexpensively - Holes under microchips (anyway done for soldering ground plates) needs to be filled by indium-lead alloy around 2/3 thickness of PCB. The surface tension of the molten alloy forms a concave surface. All that remains - wash off the solder paste in alcohol and apply gold leaf to the surface. According Alan Turing in his Systems of Logic based on Ordinals
http://fregimus.livejournal.com/212682.html - our intuition needs to be proved before it becomes ingenuity. Of course we are not MIT to check all this in the lab - but the idea is beautiful.
1993, Moscow, friend's house, "space" talks in a kitchen -
- R-7 rocket's engine was... mainframe computers are… and was invented....gyrocompass …..
Little, 6 years old girl, patiently listening to a scientific talk, then suddenly asked:
- Uncle Alex, in what year you were born?
- ?! 1959.
- Uncle Alex! You are the pig! According east’s gyroscope!
Nice article http://haritonoff.livejournal.com/210442.html can be implemented on Xbox (in Russian. Not sure about appropriate/correct translation to other languages)
MARCH 15, 2013 12:51 PM "Communication session" DB.
Thanks, Serg, MySQL is not really my, but yours. To my pleasure language is different from another SQL. StorProcs are different, has a problems with integration to VS2010 and etc. All are hard to learn in one day. Pros- for MYSQL - according Victor - available MS product has limitation - on size - for big DB needs to pay in 3 zeroes, another plus for MySQL- according Boris MySQL perfectly can holds gigabytes, especially on SSD. Cons- According to Victor - time to be spend will the same as pay for commercial product in same 3 zeros.
All what needs from you are - to store from a “ground- station- control- computer” via web interface (i.e. page Post.aspx) record into a table in "Communication session" database with fields (a) session identifier (b) type of record ("uplink"/"downlink"/"downlink-was-with-error) (c) packet number (d) station identifier/backup channel (e) variable data size and two times (f) time of WebControl (g) time of send/receive via ground stations. Simple and non fancy. Also need to have one BD per one table.
I'll take care of”listening" software on a ground station computer. It is includes control for a ground station antenna / rover, and transmitter/receiver.
For you- via web interface (you can see "mockup" page GroundStation.aspx) will be set addresses and another parameters for station control computer. Then on a page
CraftCubeSat.aspx will be formed commands to be sent to a craft/Cubesats. Click on "send" button initiates transfer of uplink data to IP connected GrStn computer (that will be my headache to post/get data from one web server to GrStn software). GrStn software accepts command and send it to uplink (or to Azimuth / altitude controls, GrStn
diagnostics) and posts back to website one records with a new session number, "uplink" / "GrStn_control" / "diag" type, GrStn number, "real" send time, and with packet's number is 0. As a result in DB will be exactly what send to uplink was. Then response (downlink) received on a 2.4Ghz communication will be posted as a second record into DB. Difference is a received time, packet number, and type. One "uplink" can creates many "downlinks". If packed can not be restored by algorithms inside communication system's microprocessor, then record will be with type of "downlink-broken-record". New uplink reset packet number to 0 and assign new session number. Another page to view (list) is a SessionData.aspx on it will be possible to visually check data.
That approach makes initial data to be stored as its. Any derivatives like retriving video or pictures or telemetry can be done separatly from a process of communication and control. Broken packages (impossible to restore in real time) can be restored later (using hash algorithm's matching). In that case additional record (of a restored packet) will be inserted additionally to a original broken packet. All addons can be done separatly, independently, and from different distributed processing computers.
I need for tonight: (a) Post.aspx page (needs to strip site.master's copy, long responce from IIS is not nessesary) (b)another page (for testing and manual entry) SesionPost.aspx (c) C# code for MyrSQL data insertion (d) visualization DB page Sessiondata.aspx (e) StorProc to get new Session number (table do not have "uniqe key" field or whatever terminology used in YourSQL).
SatCtrl web server available from link on a page http://www.adobri.com/ProjectGsC.aspx. If it is not working then ether – I am debugging, or give me a call I’ll reset server.
thnks Serg. Well done. Just to small remind - I need StoreProc. Friday - Are you coming to Korean БАНЯ? I have beer. I have to pick up next portion of just printed lunar money and then we can go.
Sep 30, 2012.
Open letter to CSA done signing in a process,
Sep 24. 2012.
Molds 3D printing.
Complicated carbon fiber parts for rover and ground
stations, require complicated molds. Design such molds require quite a time to
be properly designed. As a solution for fast mold design was choose 3D printing
technology with material withstand temperature for a 1 stage cure (93C) of
epoxy and mold dissolvable in water. Material is PVA (in long spaghetti form),
and printer is Mendel Prusa. It takes a time (1 month) to build such printer
mostly because printer is in experimental stage, source code in 3 different
languages, technology require precise temperature control, adjustments for
experimental filaments like PVA is mandatory / complicated, extruders is in
experimental stage too. But after questions with software and temperature
control was sorted out, 3D printing allows to simplify mold design. Now it
takes 5-7 times faster to make suitable mold then when mold was made from
alumini plastic. Alumini plastic actually better suited to work in molds – it
hold second curing temperature (163C) for epoxy, but saving time is more
in that case. Another advantage is to have 3D printer at hands – this is allow
time savings – ordering via online alumini molds can not be done faster then
two weeks cycle.
You can see captured screen of a printer printing
first (base) part of the dissolvable mold –
For sure software (ether java or pyton ) require improvements – estimated time
to finish print shows
years to finish task, and who knows! – may be software estimation is right ;-)
Sep 24, 2012.
I would say (to my surprise) it is not bad as
I was expected.
Last 3 months it was communication – now I
believe we have good antenna (CTO even insists to get patent), for ground
station sorted out questions with mold creation – it is combined technology with
dissolvable mold and alumini parts (takes a lot of time to make 3D printer
working but now it saved a lot of time in design), amplifiers starts to give a
promising gain. Temperature stabilization for a craft and satellite – study was
done – problems and solutions for electronics temperature control was outlined.
Stepper motors modes was also done study – there some promising development with
precisions in antenna’s orientation. Everything ready for debugging algorithms
on inclination table (simulation moon gravity).
On a business front there is interesting
development – we made attempt to push for Canadian microsatellite launcher – our
proposals/ ideas was in a form of special letter for Canadian Space Agency. Now
we are working on a signing process from supporting business. Nothing is visible
Problems not resolved et – power station,
solar panels, With power it is hard – without experiments on an orbit with
Cubesat everything is in a fog.
Video – not much - we have may be 15-20 min video for communication tests – raw
video – may be next month we can make video with frame (rover/ground station)
creation and wheels testing on inclination table.
Aug 7, 2012.
Questions about amplifiers.
Files with schematics - http://www.adobri.com/misc/stm_bt/RF_GrSt.sch and
http://www.adobri.com/misc/stm_bt/RF_GrSt.pcb file with PCB of 2.4
If provider will block files use .zip
I use ExpressPCB – not best but not bad – most important for me – I get PCB
manufactured on a 3 day after order.
Antenna for ground station -
http://www.shapeways.com/model/322767/2-4ghz-antena.html , Antenna for Cube
Low noise amplifiers -
Voltage regulators 500ma
http://cds.linear.com/docs/Datasheet/1761sff.pdf 100 ma
All components in database - http://188.8.131.52/dbadobri/Search.aspx
In Australia on two antennas like for ground station with regular wi-fi distance
12km without problem – limit is horizon (or height of antenna mount) – no need
for any amplifiers — I think with 20-40mwt for wi-fi distance can be 20-60 km.
If to connect Bluetooth that it must work on 1 km. One man claimed that on 4
antennas together (made less fancy made then my – people used cut PVC sewage
pipes) he get signal from Mars orbiter.
Schematic which I send in last tests was without regulators 3.3v – for power
amplifier and for LNA. I removed them and connect 3.3v from 1amp regulator –
regulators works somehow unstable – I think is just some of my mistake with
capacitors – I need individual regulators for Bluetooth+PIC and separate supply
for a TX and RX
R1 and R2 in power amplifier — I set 2.9V by a separate variable resistor for a
voltage reference – conformed value – in original schematic was bug - fixed
С26 — my inserts — otherwise switches does not works – another bug – fixed
С34 — same bug - fixed
L2 — have no idea - in typical application 12nH – but I put 8.2nH
Then I fixed problem in control signals for TX-RX – was stupid bug in tracing.
Then was a bug in soldering on U4 (I believe this is when I baked voltage
regulator 3.3 100ma)
Then was bug with sizes — capacitors size 402 (1мм х 0.5мм) and resistors 805
(2мм х 1.25мм), but on PCB I placed all sizes 402 – as a result I have to cut
spaces for a bigger resistors and 10uF capacitors.
If two antennas used then stable signal is over open air - 500 -670 m. If one
antenna will be behind window glass then distance reduced to 120m. If I put
amplifiers, then on 500m over glass packets travels fine. Another sign that
power amplifier somehow working – with antennas only (1mwt) behind trees no
receptions on 100m – with amplifiers at the same spot perfect reception. Also
over air 500m is definitely better reception (LED blinking only on packets used
in protocol – period of blinking give rough estimates). But power amplifier
should give 36 dB – am I right or not – input 1mwt – 36dB output 1Wt (current
also shows 350 ma)? If it is 1Wt – than signal must be really strong – even on
another side of downtown with reflection I should see reception? With 1mwt I
perfectly get reflection from nearest building – with 1Wt – no reflection – only
direct open air.
Run with CubeSat prototype at 3AM aroundVancouveris interesting task – all Boms
are curious – usual question – Is that technology can be used to get high? At
day time it is funny too – tourists
I also do not understand – first – capacitors on RF-IN RF-OUT? What sizes of
traces before and after capacitors should be?
And actually what for that capacitors? – lest assume for decoupling. Then why
on power amplifier it is 1.2pF and for LNA it is 2.4pF? – Frequency of a sample
application for a same 2.4GHz, or this picoF just what you have today for
soldering? Or this pico depend on width and length of a trace?
Does they depend on a height? No seriously – after soldering a lot of buggers on
traces – trace like hatchbacks – does it require to clean buggers or make it
bigger – to adjust picoF? Then thickness of PCB – how account this?
And what a hack this L1 – bloody inductor! In recommendation it is stated:
“Could be removed (really?!) if -7dB(?!) return (!?) loss (??!) is acceptable
(?!)” – What loss? Why loss back? — is it -8 is OK and -6 is very bad? What is
ACCEPTaBLE? dB of what? And that documentation is really good – in other specs
for a different chips – only pins – event horizon for a creativity – try what
you like with combinations 16*15*14*… etc.
Now antenna – I soldered it on J3 over capacitor (1.2?.2.4?) – as on a picture –
direct connection of a antenna to a capacitor and without capacitor – What will
be proper way to soldering /connect with or without?
Then connection to a Bluetooth – today it is a coaxial 25cm (2.4Ghz length of
wave) in a future Bluetooth will be on same PCB but now coaxial – question –
from a point of connection length/width of a trace to capacitors should be???
1/6 or 1/32 of wave’s length. Or nature likes different fractions?
Next – “50om /85 mil” or “50om/120mil” or something even more mysterious – “50om
RFIN” “50om RFOUT” – Yes I know – it was a man with a last name Ohm – also Ommm
is Tibetan song – and what? 120mil * 3 = 360 +15 = 375 = small bottle in liquor
store. But how liquor store related to U=I*R?
May be needs additional amplifier? The power amplifier designed to work for
wi-fi. And 1mWt is not enough to work?
Also on schematics present noise canceling QHX220IQT7. But I am fighting with
difference noise - in a head mostly. I even did not soldered it.
If you understand – you help will be appreciated. My knowledge is limited in RF
area - as you know.
Apr 17, 2012,
Replay to Rick.
(a) Ground station – I do not have resources to afford double development, as a
result I adopted idea that rover itself will be a ground station.
It is crazy but that is my only option – rover has to be able orient antenna to
the earth when it will be on the moon (one of the task rover has to perform).
Ground station will be a stripped from solar panels, batteries, camera stand,
wheels will be reinforced (two wheels), power will be delivered by a external
Status is: Design of a two wheel rover ground station: done. Stepper motors
control: ready. Gyro system: ready. Antenna 2.4 GHz: ready. Amplifiers for
receiver/transmitters: investigation where to buy. Transmitter/receiver 2.4GHz:
debugging stage. Software for orbit simulation: ready. Software to control
rover/ground station: in pre design stage. Noise cancelation – in pre design
(b) Main communication will be on 2.4GHz, in Canada do no require to get
operating license on hopping frequency band, no restriction on antennas, and
maximum power for transmitter is 4WT. Otherwise (for different frequency) it
will require license and this can get a time – licenses are two type –
commercial and armatures. On armatures license in Canada I can not transfer data
and receive for that transfer any payments.
2.4Ghz was chosen basically because of that reason – I am in a competition – if
I’ll use radio, is it a commercial or amateurs use? Hard to say. To avoid such
question better to use “open” band.
Another ground station problem – what to do if satellite on another side of
earth and I need to communicate with it – for example to make main engine burn –
well - needs backup communication.
I am planning to use:
http://ftp1.digi.com/support/documentation/90033936_A.pdf it is not
certified to use on planes, but on orbit – it was used couple times.
Problem: it is -40C to +85C range, and manufacturer asking about air pressure
1atm. What I wont to do – is to place it in a compartment – compartment will be
epoxy, I’ll just submerge it into epoxy only connectors needs to be extended.
To minimize noise in British Columbia needs to go some were in wildness – 1-2
hours driving time and big Vancouver noise can be shielded by mountains.
Otherwise ground station can be located outside on any field – theoretically I
need open 10m x 10m area.
(c) If you need a help to tack you satellite , and transmit/receive data – I
need to have yours equipment/antenna (if it is a standard equipment I need to
know where to get it), if it will be not a ‘no license’ frequency band – then I
need in advance to apply for a license to operate ground station.
In a case of lunar mission planning location on earth for a communication will
be Vancouver, Florida, East Brazil, South Africa, Ukraine, Kazakhstan, West
Australia, Japan, Hawaii.
(d) Testing – Two tests spec (and form for certification) I get from a CubeSat
One is vibration test and another is ourgasing at vacuum and with temperature
First is to test for launch vehicle frequency profile, on a website is a profile
for a Dnepr rocket. To pass this test I am planning to send ready for test
CubeSat inside a specified harness to a company (there are a couple around
Seattle area) testing everything on vibration tables.
Second test from one point is little bit complicated but from another point it
is also simple – place to a vacuum chamber ready CubeSat – reach level of a
vacuum – heat till 70C (or 60C) - record pressure during 2 hours, if pressure
did not drifted high then specified level (numbers on website and I also
attached it to a letter) then test passed.
(e) Temperature requirements. I assumes that temperature will be -40C +125C. I
choose epoxy with a range -65C +300C, all electronics -40 +125C, capacitors -40
+85C, stepper motors -40+85C, frame – aluminum, backup communication -40+85C.
Parts with ranges less then -40C +125C will be placed inside epoxy compartment –
and temperature has to be monitored.
Heating / cooling – no chance to control except by rotating satellite.
Reflection - gold layer.
Transferring heat – special epoxy from NASA list.
Temperature performance – not simulated yet.
(f) For shielding from a radiation – I think on orbits till 400-500 km radiation
will be not a big problem. I choose Microchip products – PIC was tested – some
CubeSat team in already did radiation tests – PIC is OK.
To get radiation protection I’ll will use soldering squares under microchip area
(only small part of a microchip holds electronic – this part has to be
protected) soldering will be lead 40, tin 60 – or may be I’ll put more lead.
On top on chip will be placed temperature conductive epoxy – on top of epoxy
will be placed another lead shield, thin layer of gold will be on top of
(g) For a sun sensor (is will be a chamber 9.5cm long with a hole on another
side of chamber) for a sensor it will be
it is -40+85C - but I think I can make it work with a +125C
Well – everything will goes well if to look from a proper point of view.
Any help from my side – do not hesitate to ask
Reply to Jenisha
Question that I am interesting based on (page 481 book by Boris Chertok
http://history.nasa.gov/SP-4110/vol3.pdf in Russian version it is better to
read but anyway):
"During Molniya-1’s third and subsequent spaceflights a rapid
degradation of the photovoltaic converters (FEP) was detected.(33) It was the
manifestation of a little known effect of irradiation when passing through
near-Earth radiation belts. Another factor affecting the performance of the
solar arrays was thermal cycling—a temperature differential from plus 120
degrees in the Sun to minus 180 degrees in the shade during each orbit.
In 1966, to reduce degradation and extend the life of the solar arrays,
Lidorenko’s institute introduced a silica glass coating to the working surfaces
of the photovoltaic cells. We had to give in and increase the mass, but we had
some leeway there, thanks to the efforts of Dudnikov’s designers. Additional
solar arrays shaped like special pleated curtains were installed, which made
them heavier. When needed, the curtains opened and fresh elements that hadn’t
been harmed by radiation or thermal cycling were put into operation."
Because I do not have access to good solar cell – I need to presizley know how
available to me cell will perform. According Boris Chertok in one month I can
loose half of the power on a Lunar mission.
Thanks for information about Sensormetrix. They are out of my reach. I have to
developed everything by myself.
Apr 11, 2012.
Reply to Rick.
Hi Rick, I can not use any available kits. Bunch of reasons.
For power plant it will be:
(a) Solar panels – I tried 3 types -
junction is not available for me now – but everybody tells me that is a best
Question is (my question) which one will be better performed on a orbit –
I am interesting in a particular case – how they will be deteorate under solar
wind at least on a low earth orbit.
I do not have such information for those solar panels – looks like I’ll have to
place all of them on CubeSat to check performance against charged particles.
Solar panels in my case has to be combined to get 3.5-4 Volts – then I’ll use
http://cds.linear.com/docs/Datasheet/1761sff.pdf to get 2.5 volts – I tried
different regulators – this one has low switch-off current and require
additional tantalum capacitor. It is 100ma output current.
Then super capacitor
http://products.nichicon.co.jp/en/pdf/XJA046/e-jc692.pdf main energy
I did not made final choice – second one is better for temperature range, (in a
case of second capacitor needs to use 2.7 V voltage regulator same type).
If temperature will be high then 85 – then capacitors still perform, but
Capacitors will be inside compartment – it is not a compartment truly speaking –
epoxy/carbon shell with another epoxy as a sealant and thermal conductor.
Voltage regulator will be inside compartment – Quiescent current 20mka will help
to fight low temperature.
Layer of gold (gold leafs) will protect form high temperature external.
Amount of layer will be determined by experiments in vacuum chamber.
(b) each solar panel split to a groups with 4V - 100ma capacity (I did not find
suitable voltage regulator with high current but this is advantage not
disadvantage). Grid of voltage regulators connects each group to capacitors.
Control of a grid performs by
http://ww1.microchip.com/downloads/en/DeviceDoc/41341E.pdf it monitors
voltage on individual group and capacitors and switch-on/off desired regulator
to charge capacitors evenly. PIC16f72x can perform that task autonomously with
transferring performance results to a central storage unit (backup computer)
(c) - unknown / no tested part is a bust-up voltage – I did not choose/ tested
any suitable – (http://www.micrel.com/_PDF/mic2250.pdf
http://cds.linear.com/docs/Datasheet/1945fa.pdf some of them does not has
correct temperature range, some of them does not has correct output current,
some does not have low shutdown current).
For bust I need 3.3V 5V, and 12/9V.
3.3V can be 100ma (power for microprocessors), 5V -1 amp pick 1 sec, and for
12/9V needs to have 1.5amp (backup communication) pick 2 sec.
For main communication module 1WT transmission – I do not have requirements yet.
(d) – grid bust-up regulators connects of capacitors from one side and then
distributed to a power suckers – (1) backup/main storage/camera unit
processor(first power level) (2) all another processors unit (second level of
power consumption). (3) orientation system, (4) transmitters/receivers.
Looks like distribution can work be controlled by same power control unit. If it
will be not enough pins on PIC16F72x then may be I’ll put another power’s plant
control unit – for controlling distribution – but I did not made decision yet.
Gathering power control unit will have two powers – one based on six
for power consumption 1mka, and then power plant charged at least one capacitor
it will switch to use power from that capacitor.
(e) all capacitors discharged at launch – simple shot cut resistor grid combined
with remove-before-flight tag will do CubeSat requirements job. Six CPC1822N
(for power plant) will be also shot cut by CubeSat switch. No power until
deployment as it specs.
That all what I have today for a power plant.
Oct 17, 2011.
Ok – this is a deal – lets assumes – (a) power plant on satellite has 6 surface
– each 2 of it can be assumed parallel, looking in opposite direction with solar
panel attached, that makes 3 groups by 2, each group orthogonal each other; (b)
gyro precession is 1 degree per 4 minutes; (c) magnetometer gives readings
(absolute) of a magnetic field with precession 2 degree. First step will be to
stabilize satellite (eliminate rotation) to a level 0.5 degree per hour. It can
be done by monitoring readings from a solar panels and do correction by rotating
stepper motors. Then after 4-5 hours when error is less then 0.5 degree (per
hour) magnetometer readings will be prerecorded during two orbits. This gives
2x3x75x75x2 ~ less 100K of data. Data analyzed for min and max of derivative and
chunks of data with max and min values can be transferred by backup
communication (not require orientation of a craft) to earth control center (~10K
of data). On the ground center simulation program of a magnetic field on a low
earth orbit will match (by brutal computer run) observed readings with
approximated Keplers elements. Basically it will try all possible Keplers
elements and find best match for observation reading. When Kepler’s found then
for a time equal ½ hour before next communication session can be generated
magnetic fields readings which craft has to follow for a proper orientation in a
time of beginning of the session. When session will starts, gyro with precession
1 degree per 4 minutes (session time) can orient satellite to perform session.
Ground station antenna will use the same logic for orientation with assumption
that earth rotates 360 per 24 hours, coordinates and direction to a north
magnetic pole can be reliably verified. The picture taken by camera fixed to the
craft frame using the same method of orientation (30 minutes of prior picture
session) can confirm precision of a method. The same technique will be used to
orient craft for a main impulse on a way to the moon, in this case max 160
minutes of predicted magnetic field values needs to be downloaded. More
complicated procedure on a way to the moon (travel time 7-9 days) for correction
impulses – lets assume the orientation of a magnetic field in each point of
trajectory will be stable during at least 1 hour. Direction to the of a moon and
earth can be preloaded, then using gyro it is possible to orient craft’s camera
(fixed to a frame) to estimated direction to centers of the earth and the moon.
Picture can be taken and gyro will return craft back to the orientation before
picture session with precision 0.25 degree (basically this mean picture has to
be taken in a intervals of 1 minutes). Then magnetometer’s readings can be used
to stabilized previous (before session) direction. Solar plant will give another
vector for orientation to the sun. Two picture can be processed on board to get
direction to the center of the earth and the moon. And calculated earth’s
direction can be used for communication session. On correction impulse it has to
confirm (by picture) direction to the earth, the moon, the orientation of a
magnetic field and using gyro (1 min) before impulse to orient the craft. Brake
impulse performed the same way.
This require – gyro/magnetometer needs to be able to collect data during 160
minutes. Collected data must be processed on board to find max and min of a
derivative. Stepper motors was to able to perform full rotation of a craft
during 1 min. Onboard computer must process images of the earth and the moon to
find direction to center of the body. On ground station (tra application,
additionally) needs to be able to simulate earth magnetic field on low earth
orbit (< 350KM). The same application needs to be able to match trajectory with
magnetic field’s readings. On ground test for power plant and orientation system
needs to include proper rotation elimination. On ground test for magnetometer
needs to include harness with rotation of a magnet in some plane. The readings
can be used to debug kepler’s elements matching and debugging orientation system
stepper motors commands. Full flight tests can be simulated on a ground. GPS on
board system will be used for Kepler’s elements verification. May be trajectory
calculation app will work in a distributed mode to speed up process. If in 1-1.5
year frame gyro with precision better then 1 degree per hour can be obtained
scenario of a flight can be revised.
Oct 1, 2011
Goals for a test flight are:
(a) confirm backup communication;
(b) confirm power plant working and gyro module working for Keplers elements
calculation, data transferred (to earth) using backup communication;
(c) main communication antenna deployment;
(d) main communication session working;
(e) conformation of a GPS readings from Main computer working;
(f) conformation of onboard processing center of earth calculation (from
(g) conformation of onboard processing center of moon calculation, in a failure
of this tests imaginary data and gyro readings has to be transferred back to
earth for additional processing;
(h) update of a software on all onboard microprocessors;
(g) simulation of a main engine firing preparation, this includes upload
magnetic fields targeting values, gyro position target values, center of earth
Sep 1, 2011. Antenna 2.4Ghz – for ground station helical antenna design and
manufacturing – it will be better to print such antenna than make it by hands –
3D printing material suitable is Nylon 12. 3D model: (click on picture to see in
SolidWorks eDrawings, req automatically installed plugins):
two 3D printing facilities : Ponoko:
http://www.shapeways.com/model/322767/2_4ghz_antena.html?gid=sg85851 . After
printing all what will be left - to connect flat reflector made from PCB board,
and 0.3 dia wire. For mockup of a testing antenna on CubeSat (that one has to be
made by hands) 3d model is (click on picture to see in SolidWorks
2.4Ghz band is actually a junk band but according FCC ((http://ecfr.gpoaccess.gov/cgi/t/text/text-idx?c=ecfr&tpl=/ecfrbrowse/Title47/47cfr15_main_02.tpl
) and in Canada
it is possible without license to use 4Wt (Canada) and in 1Wt (USA) transmitters
(assuming BlueTooth as a core), without limitation on antennas.
For local noise cancellation microchip is:
According spec >20dB noise cancellation possible. Manually (which is proved and
possible for ground station) can be achieved 50dB noise (junk) suppression.
Also was discussed the same technique for suppress noise on a transmitter –
transmitter antenna on satellite oriented to the point on the earth at the same
time noise helical antenna looks backward 180 degree. Second (noise) antenna
needs to be with receiving pattern covering the same sky area as a ground
antenna (this depend on a orbit, otherwise needs mechanical adjustment). Mixing
delayed (61.3mm) noise on 2.4Ghz from a particular place on a sky with a
transmitting signal, then subtracting same noise on receiver can give ~50+20=70
dB which will be good as it sounds. Complications: it is hard to get precise
receiving pattern for noise antennas on transmitter and receiver; such scenario
would not work from the Moon.
With active antenna this actually can be done – interesting: does anybody made
5 Aug, 2011.
Small panic about 526N epoxy – by all parameters it is acceptable (-78C +300C
tolerance, flexibility, outgassing, and etc.), but it is not in it
Alphabetical Listing. This (not in list) may be problematic for any composites
used in vehicle and craft. Epoxies 517, 556, 568 present in a section B list but
it is not suitable.
After contacting manufacturer it was confirmed that 526N actually was tested
around 20 years ago for outgassing study and Total Mass Loss at 125C is 0.49
(less then 1%) and Collective Volatile Condensable (at 25C) Materials is (0.00)
with curing conditions 2H (93C) and 16H (204C). Epoxy is good, looks like it
will be require re run ASTM E-595. Anyway satellite (including vehicle) was to
pass vibration tests according launch vehicle profile and thermal-vacuum
bakeouts to ensure level of outgassing at minimum vacuum 5x10-4 Torr.
Manufacturing of composites was to be documented.
July 21, 2011. Backup communication system – pick power 1W based on
Last one is preferable – needs one conformation ORBCOMM would it be able
to communicate with low earth orbit.
July 20, 2011. Main onboard computer and power station. Computer Rabbit with
industrial temperature requirements. 3 Microcontrollers to control orientation’s
stepper motors, flash memory, 5 high-capacitance capacitors, 3 type of a solar
panels with separated method of deployment (flexible one will be extended using
springs). All power capability – 1W.
July 19, 2011. Astro-
navigation system needs pin hole camera (20 g) and software processing captured
image in main onboard computer. Format of a captured image JPEG. Source code for
a extracting image available…. In ground tests can be included processing image
from available space videos converted to exact format from a camera. Also for
matching starts and the Moon it is possible to do series test on a ground and
Photoshop mock-up images of a Moon. Full system must be tested on orbit before
attempt to fly to the Moon.
July 18, 2011.
Communication system 2.4GHz. For a small cube-type satellite it will be
impossible to test full helical antenna (562mm) , as a result all test has to be
done on shorter one (100mm), this will reduced gain from 17dB to 9dB and
increase half power bandwidth from 25 to 60. Analog part of a communication
system was not designed yet. Communication protocol is all software related
topic, nothing complicated. For communication system needs to build 2 portable
ground stations. Station needs orientation platform (using for a amateurs
telescopes) for following satellite on a sky. Each station needs to be equipped
with 4 helical antenna. Stations will be used later for communication with a
probe on path to the moon.
July 17, 2011.
Orientation system will include 3 stepper motors. For a test on 1kg total mass
it will be require motors with bearing. Suitable weight for each 30 g. And
lubrication can be changed to molybdenum. Gyro sensor readings should be
compared with targeting XYZ values. Onboard main computer has to be adaptable to
different type of motors, (a) different inertia momentum, (b) delays of control
to avoid oscillation. Dynamic parameters of a system has to be calculated (not
predetermined) using same gyro-sensor. Placement of 3 stepper motors may be not
in a center of mass – as a result to support orientation along any of 3 axes,
system should calculate parameters by itself. On ground tests for a system can
be done only for one axe, and in plane intersect center of mass. In worth case
scenario it will require to download new binary for orientation system
(effectively for a main onboard computer).
July 16, 2011. GPS any
way needs to be developed for correct main impulses direction. Electronics can
be done in 2 weeks, and software will be major development. First it will
transfer ID29,30 satellites reading using backup communication. Then by formula
on a ground calculates Keplers elements and transfer orbit params back using
backup system. Weight for a sensor 25 g. For ground tests can be used the same
TRA app with 5 GPS satellites Keplers element and simulation of a ID 29,30 data.
When formulas finalized and errors estimated the same source code can be
inserted into a flight main computer. No backup calculations of Keplers elements
July 15, 2011. Check -
which systems from a craft can not be tested on a ground but in orbital flight.
The critical one
(a) GPS data collection from raw data ID 29,30;
(b) orientation system (on a ground it can be only a simulation),
(c) communication system with 2.4 GHz,
(d) orientation system with 3 axes
(e) astro-navigation system (calculation direction to the center of Earth, and
star matching system).
(f) backup communication.
July 13, 2011. NASA
workshop. Talk about active experiments on Apollo sites, no power require for
measure distance using laser from the Earth observatory and reflector on a Moon.
Seismometers data will be really interesting by science community. Why not to
combine all together – small sensor with nanowatt power requirements,
radioactive heater of a couple milligrams to heat electronics at lunar night,
modulation system on reflector, light weight solar panel (even if after couple
years it will might be enough), place everything under surface/dust for thermal
protection (or use power harwested from ceramic elements). Same modulation
system on reflector can be used for transferring any data from a Moon without
big power plant.
July 12, 2011. Fiasco
on Team Plan B speech – engineers do not like to talk about “political agenda”,
even on a first day of a summit was a lot of talks about how to attract
investments, and protection of intellectual property. Common outcome of a first
day was – more likely investment to teams will come in a form of “angel’s
investments”. Anyway one check box in requirements by Boris Chertok done –
Patent System on Software is in a mess, and has to be changed, and this is aside
from our project. Never say – “I made the mistake”, but every time explain –
“What a interesting problem I faced last week!”
July 11, 2011. NASA, big wind tunnel, metal and composites really nice. It is
interesting how was organized tools storage. SOFIA infrared observatory –
impressive (Why I am not a scientist?!). Really impressive Team from Barcelona.
July 10, 2011. San
Francisco is colder then Vancouver
July 9, 2011. TSSOP cases
not only be light weight (twice as SOIC) but size of PCB needs to be smaller –
all electronics will be inserted inside frame (tube) – routes has to avoid side
edges, on perimeter of a board has to be placed crush-protecting holes, two rows
on each side at least.
July 8, 2011.
Soldering passed perfectly. New soldering station works fine (never before use
hot air) SOIC cases is ok. TSSOP was actually better then SOIC. For next round
(after analog part finished) needs to retrace PCB for TSSOP because of different
July 7, 2011 – PCB
arrived – funny conversation with custom – what is it for? What a company doing
with it? How and where it can be used? (On a Moon?!). No need to pay duties and
taxes on PCB! Actually this can be a major problem for a project – today it is
possible to order all rover’s parts from metal shop over internet, each part
will cost around 200USD/CDN (total 35 parts * 200 = 7000CDN), if to order 25
rovers it will be just twice expensive. Development process each time needs to
make one change, or another and order another modified part. What about
delivery?! Delivery costs is not in same proportion as a manufacturing –
delivery for some unknown reason is only available from big parcel service
company. All other companies do not what to listen to a question – Can you send
my carbon fiber supplies by post office or by different delivery company?
Results - with a price of a PCB $51 – delivery costs $40. For $150 solar panels
– delivery, duties, taxes, multiplication calculations was total $100. And this
not all – money not only a problem - time is lost – delivery time of 3 days can
be stretched to 2 weeks. Last month parts costs $1200 and delivery was $357 –
more then 25%.
July 1, 2011
- Source code for debugging (MSVC 2011):
targetver.h. It supports all comands
implemented inside 16LF88 (see STM_LTRX.c)
result of DAC stores in file DatOut.txt
June 31, 2011 – PCB was
ordered - will be ready after holidays in US and Canada, Parts ordered (it takes
a time to check sizes and temperature tolerances for each chip). SOIC case:
U1 - 74AC161 counter 4 bit (just in case – can be 4040)
U2 - 74HC4040 – counter 12 bit
U3 - 74AC00 -
U4 - 74AC04 -
U5 - CY7C199CN (memory)
U6 - HI 3338 (DAC)
U7 - PIC16LF88 (backup microcontroller)
U8 - 74AC174 (bad choice – just was at nearest shop)
U9 - (switch/jumpers for setting delay)
U10 - 74AC161 (again was available - can be 4040)
U11 - 74HC4040 counter 12 bit for receiver
U12 - 74AC04
U14 - 74AC00
U15 - CY7C199CN (memory for receiver)
U16 - ADS930 (ACD 8 bit 30MHz)
U17 - PIC16F88
http://www.microchip.com/wwwproducts/Devices.aspx?dDocName=en010243 Memory and
16LF88 is from -40 + 85C, all another is -55+125
pic16lf88 (not for now):
74HC174 (in a case of one trigger only needed)
will be ready again after 1-4 July.
June 29, 2011 - PCB
layout done: adcdac_1.pcb It takes 3 days. Possible to reduce time to 1, needs to
accommodate to a tool before using it.
June 26, 2011 – Now
attempt to use quartz crystals with 30MHz frequency – Crystal give ether 3 or 6
MHz instead of 30.
June 23, 2011 – Time
was spend on investigation – which software to use for creation PCB and
schematics. Tried Multisim – schematic was created fast, library not full but
similar cases can be used, good but expensive. It gives quick routing, only one
problem - it better be used when needs to make a PCB board with all component
soldered, different PCB manufactures does not accept Multisim format. PCAD was
tried long time ago and it is not preferred schematic tool (may be because I am
not an electronic engineer and do not understand benefits of PCAD). Tried
another two applications, including free software in Windows and Linux. Best
choice is Express PCB. Schematic available:
Routing will be next.
June 22, 2011 – No
ideas how to solve problem – suddenly after connecting LED to random legs on
random chips (for visualizing) receiver became a good boy! Removing one LED
creates a noise, inserting to another place removes noise. Place of insertion
does not matter. Turned out – power supply instead of regulated 5V gives 5.5 –
insertion of a LED suck more current and reduce power to 5V. Ok – forget about
5V – and switch to 3V (which is kind of 3V 3.1, 3.2 - depend of LED, amount of
chips, and etc.) everything become perfect – messages flying, units responding,
June 21, 2011 –
Porcupine was extended to receiver. Nothing works even previously working
transmitter. CY7C199CN require bypass capacitors. Yes - needs to RFM. Each
memory chip gives a noise, 0.1mF helped, but communication with microprocessor
receiver units (terminal app over serial port used) does not work. Applying
bypass capacitors to all chips did not helped. Switching units (the same
software in each did not helped) – transmitter working and receiver persistently
on strike. Was enabled Master Reset in 16LF88 – reset on Transmitter works,
reset on Receiver noisy – garbage on output port without any pattern.
June 19, 2011 –
Unfortunately Alex’s skills in digital design is really rudimentary. A lot of
time was spend on prototyping. Especially in debugging logic NAND and NOT gates.
Good simulation of a digital logic will be beneficiary, but any way -
transmitter is ready in a form of a porcupine’s bread board and works.
June 17, 2011 – IUP
(Inter Unit Protocol) was re-designed from old code – now each unit has serial
output connected to nearest unit’s serial input. I.e. main computer serial
output connect to Transmitter’s serial input, then Transmitter’s serial output
connected to Receiver’s serial input, and Receiver output connected back to main
computer serial input. In passive state bytes from any units travels (looping)
over entire net. Integrity of all network can be verified by any units
individually. NAND gates can be used to bypass un-functional unit. RST in high
state will enable transmit data btw nearest units, and pull-down resistor can be
used to skip broken. Both microcontrollers (16LF88) for LDF (Laser Distance
Finder) transmitter and receiver designed to be exactly the same (difference in
command “wW” and “wV’ – W set W/R high, and V set ports bus into input mode).
Source code: STM_LTRX.c, and for and
June 10, 2011. Digital
U1 - 74AC161 counter 4 bit (just in case – can be 4040)
U2 - 74HC4040 – counter 12 bit
U3 - 74AC00 -
U4 - 74AC04 -
U5 - CY7C199CN (memory)
U6 - HI 3338 (DAC)
U7 - PIC16LF88 (backup microcontroller)
U8 - 74AC174 (bad choice – just was at nearest shop)
U9 - (switch/jumpers for setting delay)
U10 - 74AC161 (again was available - can be 4040)
U11 - 74HC4040 counter 12 bit for receiver
U12 - 74AC04
U14 - 74AC00
U15 - CY7C199CN (memory for receiver)
U16 - ADS930 (ACD 8 bit 30MHz)
U17 - PIC16F88
memory 128kx8 32-SOIC
memory 2Mx8 52-TSOP
memory 1Mx8 44-TSOP
Connector to main computer will be on same bus as 16LF88
16LF88 pin assignments:
RA0,1,2,3,4,7 will be DATA0,1,2,3,4,5
RB0,1 will be DATA6,7
RB2 - serial input RB4 == W/R (out)
RB5 - serial out RB6 == RST (out)
RB7 == CLCR (out) Operations:
1. Transmitter starts with resets W/R = 0, CLCR =0, RST = 0, then set RST=1
sequence “2wcrR2” where
“2” unit == “transmitter” communication tag
“w” set W/R low
“c” set CLCR low
“r” set RST low
“R” set RST high
“2” end communication with transmitter unit (closing tag)
This will reset counter and transmitter ready to accept data into a memory.
2. Microcontroller (or main computer) will store in transmitter memory wave
(data or for laser byte 0xFF coded “on” or 0x00 coded “off” state) to be
transmitted CLCR = 1, W/R = 1 then needs to set on bus 8bit data to store in
memory and set CLCR will clock-in byte into CY7C199CN
Sequence : “2CW88888888wCc2” where
“2” unit tag (transmitter)
“C” CLCR set high
“W” W/R set high
“8” is just a helper command it will store sinus wave in 4096 bytes for
debugging Each byte value will stored on a raising CLCR (first byte is not
important) At the end of 32K bytes needs to set W/R = 0 CLCR = 1 and back CLCR
“wCc” Closing tag for unit is == “2”
3. Now transmitter memory ready – oscillator signal with 30MHz will clock-out
data from memory to DAC (U6 - HI 3338). For debug it can be 2Mhz from 16LF88 on
a pin 12 (RA6) by setting configuration bits, or by disabling clock on RA6 it
can be manually set pin 12 on/off. Sequence is (for oscillator, and 2MHz
debugging ): “2wcrRVC2”:
“2” –transmitter unit
“w” – W/R low
“c” – CLCR low
“r” – RST low – this will reset counter
“R” – RST high
“V” – W/R will be in low and data bus on 16LF88 will be set in input mode on RA
“C” – CLCR high and oscillator drives counters. On each rising edge data will be
send to DAC and on each failing edge of oscillation DAC fill latched previous
4. Transmitter and receiver uses same oscillator. It is possible to set delay
for receiver to start to store DAC data into receiver memory by switch/jumpers
U9. Delay can be useful because on long distances 25 km == laser beam travels
50km time delay == 1.(6) E-4 sec, this for a frequency 30MHz require to store
digitized samples in 5000 bytes (for laser beam travels from a craft and bounced
back from a Moon surface). Error in calculation will be 10-20m. Less data to be
processed (by microcontroller) faster (less delay) to measure distance. Ideally
instead of jumpers possible to set comparator this will trigger receiver to run
by some delay. Roughly 16LF88 running on 8MHz (2Mil or /sec) can process one
measure per sec. This give (at speed 2800m/s) 1 ms accuracy of 0.01 to start
break engine. At the same time main computer will be able to process 50
measurements per sec. To confirm calculation looks like needs to start measure
8-9 sec before impact by main computer – when first measure will be obtained,
needs to switch for a backup microprocessor and compare calculated time. Speed
of a craft does bring error in time travel 1.(6)E-4 of a laser beam around 1 m.
Speed of a craft can be calculated by two different measurements by calculation
difference distances and knowing time btw measurements.
5. Sequence for measure distance is: “1wcrRVC2wcrRC” where
“1” receiver’s unit tag
“w” W/R set low for receiver
“c” CLCR set low for receiver
“r” RSR set low for receiver
“R” RST high (receiver reset done)
“V” W/R set low and data bus on 16LF88 receiver will be set in input mode on RA
“C” CLCR high on receiver and receiver ready to digitize data
“2” unit transmitter tag
“w” W/R low for transmitter
“c” CLCR low for transmitter
“r” RST low for transmitter
“R” RST high for transmitter (transmitter ready t send data to DAC)
“C” oscillator connected to counters on transmitter data clocked out with 30MHz
– delay passed and receiver memory already with measure data
6. Sequence to read receiver memory is “w2whhhhhhhh1” Where :
“w” – W/R low on transmitter
“2” closing tag for communication with transmitter
“w” – W/R low on receiver (prev tag was opened communication with receiver)
“h” – read measurements in chunks of 4096 bytes
“1” – closing tag for communication with receiver
7. Bus data and control signals RST, CLCR, W/R will be sheared with individual
microprocessors (backup) and main computer – this require commands for
microprocessor “L” – listen – in this case both units will put all ports RA,RB
into “input” mode.
June 5, 2011.
will consists from:
Oscillator 30MHz 3-5v
ADC 30Mhz 8 bit
Memory for storing digitized data from ADC 32-64Kbytes.
Counters Microcontroller for ADC
Interface for a main onboard computer
Sensor for a red laser
Amplifier with variable gain
DAC 30Mhz 8 bit
Memory for storing transmitted data to DAC 4Kbytes
Microcontroller for DAC
Interface for main computer
Laser 0.5W (red)
Key transistor for controlling and modulation laser’s impulses.
Optical system for a LDF
June 3, 2011. During
descent, to ignite brake engine at 2800 m/s requires real time distance
measurement to Moons’ surface. We think radar or laser can be used. Weight for
radar is heavier of the two concepts, around 2-3 kg, which is too large for our
small landing vehicle design. Alternative approach is a Laser Distance Finder
(LDF). In this case weight will include only two optical lenses, electronic
equipment and mounting, significant saving in total mass of a probe. Off-shelf
laser range finders can work on maximum distance for a couple kilometres.
Requirements – measure distances from 30 km to 5km, reflection from a lunar
dust, 0.5W laser.
Additionally (a) if it possible the same system can be used to transfer data by
laser beam and (b) be used to measure distance to reflectors on a moon on early
The Craft will be a polygonal aluminum frame mounted on the top of the vehicle
frame. Fixing both frames together will be done via pins removable by inflating
air-bags system. On the craft’s frame will be mounted 4 fixed impulses engines
and a low trust engine. 4 engines will be mounted orthogonal to each other with
center axes project via centre of mass of a probe. Firing sequence – fist orbit
correction, second orbit correction, main impulse, brake impulse will be each
time projected via the centre of mass of the probe.
Hermetically sealed parts of a frame will at the same time tanks for a
liquid (ethanol/alcohol) propellant body for a low trust engine. Pumps will be
require to precisely adjust probe’s centre mass position.
While in a flight, the probe's orientation will be performed via inertia
momentum created by the rotation of three masses. Those masses will be mounted
to stepper motors located close for central mass point inside two fixed engines
frame. Third stepper motor will be mounted directly to a craft’s frame.
Engines will be carbon fibre cylindrical boxes. Small compartments will be
in two of engines to accommodate stepper motors with rotating masses. Graphite
nozzle will be mounted inside rear part of each engine.
Same rotation momentum will be used to control thrust vector direction at
Location of a low-thrust engine not determine et, but it must be
orthogonal to plane of 4 fixed impulse engines.
Requirements for a vehicle’s mass 6 kg on a Moon surface brings requirements for
a total mass on a low earth orbit to100kg. That weight puts a craft onto a
category of amateur satellites. Amateur satellites usually launched as a second
payload for a regular launch. It will be desirable to stay in limits of 100kg,
otherwise price for a launch can be high. Here is a reference for a OSCAR type:
and CubeSat type: