PICAXE in Space

srnet

Senior Member
Does anyone have any experience of using a PICAXE (40X2 in this case) in an environment where the temperature can range from -40C to about +60C ?

I am interested in knowing how stable, in practice, the internal RC processor clock and the Sleep timer are over the above range.

In addition does anyone in the UK have access to a vacuum chamber, I am building an electronics board for a micro satellite, and would like to see if anything goes pop in a vacuum.
 

westaust55

Moderator
Have you had a look at the diagram "HFINTOSC FREQUENCY ACCURACY OVER DEVICE VDD AND TEMPERATURE" in most PIC datasheets.
For the 20M2 see figure 30-3 pafe 351:http://ww1.microchip.com/downloads/en/DeviceDoc/41440B.pdf

There there are ovens for temperature trials.
Some motor manufacturers/repairers have vaccuum impregnation chambers for ensuring the insulation resins/varnishes get into all the mooks and crannies without air voids.
 

srnet

Senior Member
Motor manufacturers, never thought of that.

Wonder if Rev Ed know anyone with a vacuum chamber, I can use for free?
 
Last edited:

manuka

Senior Member
Freezers are handy for quick sub zero temperature checks, but far cooler effects can be quickly investigated with "Dry Ice" (solid carbon dioxide), which is available cheaply (~US$5 a kg here in NZ) from industrial gas outfits. It's around -60 Celsius & pellets of it will last a few days if held in a thermos. Just lower your circuit in & see what happens!

Footnote: Dry Ice is fantastic for entertaining kids, as it boils "mad scientist" style if dropped into warm water!
 

Attachments

hippy

Technical Support
Staff member
Wonder if Rev Ed know anyone with a vacuum chamber, I can use for free?
Not sure if we do but there's always universities and colleges. If you can find a reason why it would be beneficial for them or their students you might find them amenable to helping you out.

Writing a short article on the "effects of altitude, vacuum and low temperatures on low cost microcontroller structures and internal oscillators" and promising to give credit is one way I can think of.
 

srnet

Senior Member
Freezers are handy for quick sub zero temperature checks, but far cooler effects can be quickly investigated with "Dry Ice" (solid carbon dioxide), which is available cheaply (~US$5 a kg here in NZ) from industrial gas outfits. It's around -60 Celsius & pellets of it will last a few days if held in a thermos

What a 'cool' idea.

How much do you think I would need, by weight, the processor PCB with PICAXE and Radio (RFM22) is 40mm x 40mm ?
 

BeanieBots

Moderator
In addition does anyone in the UK have access to a vacuum chamber, I am building an electronics board for a micro satellite, and would like to see if anything goes pop in a vacuum.
No longer able to help out with the vacuum chamber but many things can happen at high vacuum that you might not expect.
Simple LEDs can overheat because there is no heat conduction. Resistors can also suffer the same fate.
Motors which use carbon brushes fail very quickly because the water within the brushes boils off and then they wear very quickly.(same for almost all common lubricants).
Not to mention that coil heat can only escape through the bearings.
Regular cable insulation outgases and becomes very brittle but you can use PTFE cable instead.
Probably a whole host of other things I've forgotten but I'm sure you'll discover your own issues once you start testing. The only thing to be aware of with most vacuum issues is that the effects take time so don't be too hasty with your test results.
 

manuka

Senior Member
I agree re long term vacuum effects being devious. Just how deep & prolonged is this space voyage?

A kg of dry ice will be more than enough. I bought this amount recently for chilly bin frozen food use while our freezer was out. There was still plenty left after 2 days so I popped into the local school & entertained several classes of new entrant kids with it. Even the neighbour's teens were impressed. With any luck this may serve to inspire a few future Nobel Prize winners!
 

Attachments

srnet

Senior Member
Regular cable insulation outgases and becomes very brittle but you can use PTFE cable instead
I have found short lengths of PTFE 0.1" pitch ribbon cable at Farnell, I need a short length to connect the battery\power board, but have yet to find any at 0.05" pitch, I need some in 23way to connect up an I\O expansion board, although I could always use short lengths of single wire.
 

KeithRB

Senior Member
I think radiation is more of a problem than vacuum, however plastic parts are generally not used in space because they tend to outgas and interfere with other electronics. You might have to put your electronics in a hermetic box.
 

MFB

Senior Member
You should also take into consideration the effects of low temperature on battery characteristics. LiPo batteries have increased internal resistance/voltage drop at low temperatures and reduced capacity. However, I have not read of any ill effects due to low pressure. Attempt to use electrical power to warm the battery may be a case of diminishing returns. I was wondering if mounting the battery at the top of the electronics stack and then leaving a window in the thermal insulation might help warm the battery directly from solar energy (no solar cells required).
 

srnet

Senior Member
Last edited:

srnet

Senior Member
You should also take into consideration the effects of low temperature on battery characteristics. LiPo batteries have increased internal resistance/voltage drop at low temperatures and reduced capacity
Indeed, not much to be done about the low temperatures, but as long as there is enough power to keep the PICAXE going in low current mode, a Lithium Ion should be OK, remembering not to charge (from the solar panels) when it is below 0C.
 

Dippy

Moderator
Drop an email to Astrium.
Put on your shorts and say it's a school project.:)
If they answer it'll take the guesswork out of things.
And it'll save Google getting worn out.
 

rq3

Senior Member
I have a very low volume product which uses the surface mount 20M2, and it has been tested to MIL-STD-810.
The environmental details are here: http://www.ackemma.com/tech.html

I have no idea what these conditions do to the internal clock rates, as that isn't relevant to me.
 

srnet

Senior Member
Drop an email to Astrium.
Put on your shorts and say it's a school project.:)
Well in a way it is a 'School' project.

I have been assisting the space satellite guys in Moorhead State University;

http://www2.moreheadstate.edu/ssc/

to produce one of these;

http://www.space.com/8838-tiny-satellites-big-science.html

Why me ? Well in that form factor, about the only suitable transceiver, really small and cheap, is the RFM22!. And PICAXE was one of the choices for the processor. The guys in Moorhead have all the test gear needed, but I have been developing the processor\radio and want to test it before sending it to the US for assembly.
 

srnet

Senior Member
I have a very low volume product which uses the surface mount 20M2, and it has been tested to MIL-STD-810.
The environmental details are here: http://www.ackemma.com/tech.html
Interesting and impressive.

The internal clock rate is not massively significant, all that would alter would be the audio frequency of the Morse beacon part.

The PICAXE sleep instruction is used as part of power management, low battery, longer sleep time between beacon signals, and it would be useful to know how much that sleeps time varies in practice, if only to confirm what the data sheet claims.
 

Dippy

Moderator
What an interesting project.

Does this mean that your device is being fitted into a (sealed?) cube?

What about your batteries? How will they perform with vibration and temp and vacuum?
Are there regulations and minimum requirements?
I can't imagine the 'organisers' being impressed when your 20p cells explode ;)

I'm a bit rusty on latest PICAXEs:
If you were using PIC you could pop it out of Sleep more accurately with an external trigger.
As it is you have to rely on INTRC/WDT which, according to datasheet, varies quite a lot.
I've never tested it but my first step would be polythene bag and pop it into freezer for a few days.


Seriously though, get advice from people who know rather than us lot who are guessing.
 

srnet

Senior Member
Yes, the board will be going across to the US to be fitted to the PocketQube satellite and have the solar panels attached and the like by the guys at Moorhead. I am working with a couple of others in the US who are looking at the power\battery board.

The regulations etc are being handled by Bob Twiggs team at Moorhead (google him)

Yes, batteries, interesting, one of the guys in the US working on the power board has access to an environmental chamber, so we can do some cold tests on a battery. If we can get away with it we might use nothing more than a Lithium Ion battery like you use for a camera (such as Canon NB6L) which has enough capacity and is the right size. It may need to be dismantled to remove the protection circuit however, that depends on whether we need a high current to melt some plastic to release the antenna.

PICAXE code is working.

The RFM22 has a crystal controlled NIRQ driven Wake Up Timer (ms to about 3 months) code for that works also, so I can use that if I need to. But I would prefer to use the Sleep command if possible as there is less code to go wrong, and no NIRQ to miss. I have tested it my freezer, no significant change, but then my domestic freezer only goes down to about -17C.
 

Dippy

Moderator
I don't think my Google is working properly :)

How long will it float around for?
I'm just thinking of precautions for when your circuit is (obviously) unattended for long periods with possible power supply glitches or code get's stuck.

Are you going to have an external reset watchdog?

Can there ever be occasions when power is lost and then is slowly restored?
If that could happen then you should look into delayed MCLR (or even external trigger) at power up.
Always best to cover all sod's law possibilities.
 

srnet

Senior Member
How long will it float around for?
Few months, dont want to litter space.

code get's stuck.
External watchdog, may even be a 08M2, to reset the PICAXE

Are you going to have an external reset watchdog?
Can there ever be occasions when power is lost and then is slowly restored?
If that could happen then you should look into delayed MCLR (or even external trigger) at power up.
External power surpervisor holds the PICAXE in reset, if the voltage falls below 2.9v (3V regulator) and does not release reset until volts rises above 2.9V again. Picaxe will run with about 3ma down to 2.2v, so can monitor the battery itself with minimal power above 2.9V

However this should not happen, and if it does it probably means the solar panels or chargeing have failed anyway = dead satellite.

They are no doubt lots of other tricks to pull but the idea was to keep this real simple, real small, cheap and not impossible for other to build or program.
 

Dippy

Moderator
Oh simplicity yes - and often goes hand in hand with reliability.

Supervisory sounds good, though I would add the (Microchip suggested) simple RC+diode on MCLR as it is dual function and costs 10p.
And I would be reluctant to use another micro as a master external watchdog.

But, just from a personal perspective, I wouldn't compromise reliability for over-simplicity merely for others to copy. Nor would I choose cheapest components like some people will (not you).
Good luck with it, sounds like a great project.
 

hippy

Technical Support
Staff member
And I would be reluctant to use another micro as a master external watchdog.
That can just shunt the problem down the chain and if the main micro has failed you have to hope the watchdog micro hasn't as well, while hoping watchdog failure doesn't mess with the otherwise working master.

It might help simplicity, rather than have a true watchdog, have something that resets the PICAXE every so often regardless.

It's a well known problem that simplicity introduces single points of failure that can be catastrophic, and avoiding single points of failure adds complexity so it can be hard to say what's best.
 

srnet

Senior Member
On the watchdog I would agree, and there is a rather nice MAXIM one in a SOT package, MAX6369–MAX6374. But as the device will use sleep mode to conserve power (20uA) and the sleep might be for extended periods, the max watchdog time of the above device (60secs) may not be enough. We have these web conferences on Wednesday evening, its one of those things to discuss.

Price of components for the electronics is not a major issue as such, the RFM22 is real cheap (£9 or so) for a frequency agile transceiver, and part of the exercise is to see if that devices can work. The only real quality item (apart from solar cells) we might need is the battery, a £1.80 eBay camera battery may just be a step too far.

The power output is enough, 100mW, for the Morse beacon to be heard from 2000km away, and the uplink budgets appears to show that with 10-100W of uplink power and a reasonable yagi, we can command it from the ground, your supposed to be able to shut it down on request.

There is also a fair chance that students will be able to decode the fast Morse data, 60WPM, using nothing more than a UHF tranciever, a small yagi and a pre-amp. Whether the data telemetry will work on the ground is less clear.
 

Paix

Senior Member
There is also a fair chance that students will be able to decode the fast Morse data, 60WPM, using nothing more than a UHF transceiver, a small yagi and a pre-amp.

At 20wpm you can include a lot of operators who will be able to copy your data aurally. Pretty much none at 60wpm. If you have the available time in your pass window, then that is perhaps worth considering. Much above 20wpm you start losing capable operators quite rapidly.

Are students perhaps ace CW operators I wonder :) best to play to the larger crowd I think.
= = =
I'm not sure how busy AMSAT US are at the moment due to ITAR interpretation. Good wishes for your project.
 

srnet

Senior Member
There is also a fair chance that students will be able to decode the fast Morse data, 60WPM, using nothing more than a UHF tranciever, a small yagi and a pre-amp.
I forgot to add "and a PC\Netbook with free software"

The callsign goes out at 15WPM, for humans, the 60WPM is for machines.

The 60WPM will alternate between battery volts, solar panel volts and temperature, I might even sneak in "PICAXE"
 

geoff07

Senior Member
Is it going to have any transponders? or purely experimental plus morse? I'm in the process of resurrecting my old pacast station and this might be an encouragement! Not that there is room for too much, but your transceiver module seems quite capable if you have the power budget and antennae.
 

srnet

Senior Member
Is it going to have any transponders? or purely experimental plus morse? I'm in the process of resurrecting my old pacast station and this might be an encouragement! Not that there is room for too much, but your transceiver module seems quite capable if you have the power budget and antennae.
No transponders, yet. This is very much proof of concept for probably the cheapest, smallest and simplest satellite ever. Its been nicknamed $50Sat. We are confident the Morse 15WPM beacon and 60WPM 'telemetry' can be heard and should be decodable on a PC with modest gear (437Mhz). The power budget for the packet\data uplink looks good, we might get away with a modest 10W and average yagi. The downlink of the packet\data is marginal, a packet will be transmitted every 30 seconds so we will be able to check that out, if it survives launch of course.
 

Pauldesign

Senior Member
Nice to hear a space related project on PICAXE forum.

Probably might be the first known PICAXE to space :)

Well, i'm also working on a nano-sat related project (due for launch this October) but something more useful than yours :)

Keep us posted ;)
 

srnet

Senior Member
Nice to hear a space related project on PICAXE forum.

Probably might be the first known PICAXE to space :)

Well, i'm also working on a nano-sat related project (due for launch this October) but something more useful than yours :)

Keep us posted ;)
So tell us more ........
 

Pauldesign

Senior Member
So tell us more ........
Sorry, i wished i could blow the news first but is classified for now until after October; if all goes well :)

If u've noticed, that's why i haven't been spamming the forum of recent because i've been busy ;)

By the way; i reckon the $50 dollars sat exclude launch cost as launch is the most costly (almost 10000 times that price) :D

Still cool stuff though!
 

srnet

Senior Member
By the way; i reckon the $50 dollars sat exclude launch cost as launch is the most costly (almost 10000 times that price)
I have seen figures that suggest that you can get a Cubesat (10cmx10cmx10cm) launched for $50000, and allegedly you can get a tubesat kit (includes 'free' launch) for $8000.

But this satellite is tiny, you can fit 8 of them into the space of one Cubesat, so the launch costs are much reduced.
 

srnet

Senior Member
Just been out testing the radio for this satellite, the RFM22B.

From a hill just to the North of Cardiff to another in the Mendip hills by Cheddar across the Bristol Channel, a Line of sight Distance of 40kM. Simple 1/4 wire antennas.

For Data packets (19bytes);

100% success rate at 100mW, 50% success rate at 50mW, 15% success rate at 25mW.

At 100mW I was able to remotely command the 'satellite' to turn its transmitter on and off.

For FM Morse audio;

Using a FT60R UHF tranciever and 1\4 wave vertical antenna the morse was decodeable by ear at only 1.2mW from the TX, giving an equivalent LOS range of 365kM
 

srnet

Senior Member
I have two satellite PCBs across in the US for the radio\processor board to be added to the power board and then assembled into the chassis, I think we are using 3mm alluminium for shielding.

Launch was postponed a month, so I might look at making a few improvements to the PCB.
 

Attachments

srnet

Senior Member
I was testing out using the RFM22 to send AFSK RTTY, 8 bit ASCII 1000hz\1850hz at 100 baud.

It works OK, but as far as machine decode is concerned, Mmtty for RTTY and FLDIGI for Morse, the 60WPM Morse has about a 3dB advantage for signal strength.

One way it would be worth the effort is if the was a RTTY decode application that would display the hex values of the received ASCII characters, sending the number $FF\255 as 2 or 3 ASCII characters is not very power efficient.

Anyone know of a RTTY application that will display the hex value of decoded characters ?
 
Top