​ ​ ​ ​ Picaxe-based Satellite: PongSat-18X - Page 2
Page 2 of 6 FirstFirst 1 2 3 4 ... LastLast
Results 11 to 20 of 59

Thread: Picaxe-based Satellite: PongSat-18X

  1. #11
    Senior Member
    Join Date
    Jan 1970
    Location
    Groningen, The Netherlands
    Posts
    467

    Default

    Due to what appears to be a server problem at my internet provider it seems that the links towards the pictures point to the blue sky instead of the pictures meant. I will upload the pictures again this evening, and then respond to the various reactions.

    Sensor testing is now almost ready (and we found a solution to the tilt sensor problem by writing a small subroutine for interpreting the data)

    Jurjen

  2. #12
    Senior Member
    Join Date
    Jan 1970
    Location
    Groningen, The Netherlands
    Posts
    467

    Default

    The whole test circuit is operational now (both hardware and software, including data storage and retrieval) so it's time to begin soldering the pongsat circuit. In the meantime we'll keep the identical test circuit operational in order to have more advanced output options (a StampPlot interface) in the program and do some other final adjustments to the software (in particular sensor settings for daylight conditions). It has already been a very enjoyable adventure, as we have learnt some new things by dealing with the sensors.

    The two sensors that gave us some extra work were the tilt sensor and the use of LEDs as optosensors using the 18X output pins reconfigured as inputs.

    Firstly, the tilt sensor when in the on state appears somewhat surprisingly to act as a vibration sensor, therefore it does not really register tilt. This problem was easily solved by dividing the measurement interval in short ones, register the number of changes per short interval (using the COUNT command) and then define a change in tilt as a difference between one state (0 transistions) and the other (one or more transitions) when comparing the current interval with the previous one.

    The opto-LEDS testing happened in two stages. First Output 6 and 7 had to be redefined as ADC inputs. This is done by first enabling them as inputs, and then (after a waiting period to allow the incident light to discharge the reverse-charged LED) to configure and enable the ADC module and finally performing the conversion. Since a READADC10 command must be simulated by a few PEEKs and POKEs this requires considerably more time to execute. This fact, and the relatively low impedance of the input when the ADC module is enabled, causes the LEDs to be discharged completely before the read is done. As a working remedy I added a 10nF capacitor in parallel with the LEDs (This is not needed when using standard ADC inputs with READADC10). I will publish the codes for enabling the outputs as ADC inputs in the code snippets section (maybe hippy is willing to refer to it in his SFR usage thread).

    Some reactions to previous posts:

    @Mora99: No, we have not tested the DS18B20 at very low temperatures. What we foresee as a more serious problem is the voltage drop if the battery due to low temperature; the minimum operating voltage is 3V. This is one of the reasons why we want to include the voltage reference in order to continuously monitor the battery voltage. We have thougth of isolation experiments in the future.

    @womai: Thanks for the print offer, but maybe we can develop the next version as a joint Picaxe Forum project, using your capabilities to create a very compact print layout and thus an even more powerful version. Maybe we should regard our current project as a feasibility study, and if it appears succesful we can aim really high and define on the forum a much more advanced version, with a work division between several members (we for example could do prototype testing with our philips ee system) and proper planning.

    @mycroft2152 and @slurp: The flight is about 2.5 - 3 hours, plus some "dead time" after the landing. Indeed our pongsat has to be switched on and off, but the places available for these special pongsats is limited. The foam insulation sounds interesting, i haven't made up my mind on this yet.

    @MFB: As we reported on previously, the problem with the (very cheap) tilt sensor has been solved by a few codelines extra and does give satisfactory results now. Note that we included both tilt and vibration sensors as wind effect sensors, primarily because we suspect that the vibration sensor may be too sensitive ...

    Best regards,
    Jurjen
    Last edited by kranenborg; 11-04-2008 at 23:58.

  3. #13
    Senior Member
    Join Date
    Jan 1970
    Location
    Northeast USA
    Posts
    1,441

    Default

    Have you looked at your energy budget on the battery? If there is enough capacity to run a small heat source ( LED or bulb). Along with the insulation you may be able to maintain the internal temperature of the battery.

    Of course if you were NASA, you would have a small radioactive pellet to do the job as is used on the Mars Rovers.
    Last edited by Mycroft2152; 12-04-2008 at 22:53.

  4. #14

    Default

    Interesting Project!


    The clock frequency, as you suggested may be a key part of this... to keep it alive for a bit longer.

    Can I suggest you reduce it with a external resonator, to something like 1 Mhz, and speed up the I2C in the software to 400khz - so It will still be doing 100khz.

    Don't forget, If your 24AA512 fails, no data can be recorded.
    Last edited by xnederlandx; 12-04-2008 at 23:30.

  5. #15
    Senior Member
    Join Date
    Jan 1970
    Location
    Groningen, The Netherlands
    Posts
    467

    Default

    Considerable progress has been made, and we learned a lot about the picaxes internal operation due to a very annoying and persistent failure of operation, that we were finaly able to resolve ourselves ... More on that below.

    The software runs flawlessly on the hardware propotype now, and the identical PongSat hardware is getting in a final stage too, as shown in this picture:



    The PICAXE-18X is on the left, the clock with the crystal on top on the foreground, and right back is the memory chip with the tilt sensor hanging over it. Back in the transistor casing is the DS18B20 temperature sensor, and barely visible on the left side of it is the tiny MAX6018 voltage reference chip (mounted to the wires, which was a horrible job to do). The leds act as light emitters as well as sensors. The vibration sensor still has to be added

    As always there remains some tough problems to be resolved at the end, and we dealt with the following:

    - It appears that the PICAXE firmware for the READADC instruction does not initialize the ADCON1 register (it probably assumes a zero value at startup). Since this register is changed in our software that reconfigures Output 6 and Output 7 as ADC inputs (its MSB bit is set for proper shifting of a 10-bit conversion result in a 16-bit register), it finally appeared to cripple the READADC operations later on (for battery voltage reading). It took quite a few hours to pinpoint and solve it (but with great satisfaction because we did it ourselves), and I added this in the code snippets section (where I posted the Port B Output reconfiguration for additional ADC inputs) as well. Maybe a suggestion to Technical to add the explicit initialization in firmware?

    Soldering gets complicated at the end as well, but most is done now. Due to lack of space the switch for log/readout is not implemented, we will log only, and subsequently after the trip we can download the results by reprogramming (as a download interface will be implemented)

    Regards, Jurjen & Joost
    Last edited by kranenborg; 16-04-2008 at 05:25.

  6. #16
    Senior Member
    Join Date
    Jan 1970
    Location
    Groningen, The Netherlands
    Posts
    467

    Default

    Like in a real space project the PongSat-18X (Code name NLSE-1) has been sent away, unreachable, ... by mail to the USA (last Monday). The picture below shows almost the latest stage (on one of the LEDs a black straw was mounted in order to catch background light only). Just before packaging the circuit was tested fully OK, and the clock is already running now. Lets see how US Customs reacts to this one. Together with this one four other PongSats are packaged containing some plant seeds and marshmallows, in order to do some biological and solid mechanics experiments ...



    Due to the large number of wires at the backplane (as well as some extra components) the circuitry did not fit competely in the pingpong ball as I had wished. As a consequence I decided to leave it largely open at the top. Next time a professionally designed circuit board by Womai and using surface-mount components should cure that problem.

    The picture below shows the positioning of the sensors:



    The back plane quickly became a mess and looks as follows (although the circuit worked flawlessly right away):



    I have tried to increase the chance of getting the flight data unspoiled by asking JPA to cut away one of the wires on the side of the ball prior to shipping back, effectively removing the battery from the circuit.

    The final circuit is shown here:



    The documented code for the PongSat is included as an attachment. Note that in the program body a provision is made for a temporary low-power sleep/wake-up mode when the battery voltage gets dangerously low at very low temperatures (2.3V shutdown and temporary sleep, wakeup at 2.6V).

    If this trip is going to be successful and we get the thing back in good shape, it might fly for a second time; I already got several software improvement ideas (for example the light sensors could be calibrated authomatically at start-up (i.e. sensitivity), and the NTC data could be converted by the Picaxe directly to temperature using table look-up). There is plenty of space for improvement, as less than half of the code space is currently in use. Having done this project I am still amazed by the flexibility and power of the 18X.

    I am now eagerly awaiting a signal from JPAerospace Ground Control, most likely in the beginning of next week.

    Regards,
    Jurjen
    Attached Files Attached Files
    Last edited by kranenborg; 23-04-2008 at 23:04.

  7. #17
    Senior Member
    Join Date
    Jan 1970
    Location
    Groningen, The Netherlands
    Posts
    467

    Default

    The flight that should have been performed at 2 May has been delayed by about a week, so hopefully some reports within 10 days ...

    /Jurjen

  8. #18

  9. #19
    Senior Member
    Join Date
    Jan 1970
    Location
    Groningen, The Netherlands
    Posts
    467

    Default

    The mission was delayed again, but is planned still for the near future. I hope this will be in a few weeks

    /Jurjen

  10. #20

    Wink Interim 28X2

    What a great contribution. Do you know if anyone has done similar work to obtain more than four ADC inputs from a 28X1? I really need the extra analog channels and veriables promised on the 28X2 but can't now remember which summer this product was going to be introduced.

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •