Re: [CCHS] Garduino garden monitoring and automation

[John Spencer]

Hi Greg

I'm certainly interested in this, even though my garduino project has been stalled for over 12 months!

I have at least one (possibly two) seeed studio moisture detectors I'm happy to donate.

John

On Fri, Apr 5, 2013 at 2:18 PM, Greg Nash <nash...@iinet.net.au> wrote:

Hi All,

 

It seems there is some interest at the moment in garduino projects so I thought I start a discussion, my personal interest is in building moisture, temperature and light and whatever else we can think of monitoring and watering automation for a 4 bed vegetable garden based on the diggers club publication: http://www.diggers.com.au/the-australian-fruit-and-vegetable-garden.aspx 

 

The plan was to use some of the basic techiques linked in linked here:  http://www.instructables.com/id/Garduino-Gardening-Arduino/ and here: http://www.instructables.com/id/Garduino-Automated-Gardening-System/ 

 

Now that I've met a few of you at Tuesday sessions I know there's some related interests/projects on the go and it would be good to share ideas, I plan to bring in a new mascot (some kind of easy to care for plant) for Garduino fun on Tuesday so our first challenge will be to keep it alive!

 

To kick off the discussion I'd like to know if any of you have had experience with soil moisture sensors.. especially the cheap ones sold on ebay for arduino or the old two nail trick used in the links above. if anyone has one of these or other garden related sensors lying around and you don't mind us having a go at it I'd be grateful if you could bring it in to play with.

Cheers

Greg Nash

P.S. can someone please let me know if the payment gateway is working..

--
You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.
To post to this group, send an email to connected-commu...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msg/connected-community-hackerspace/-/eZm2KzihD48J.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[ajfisher]

Hi Greg & John.

I've done a lot of playing on this front over the last couple of years and although my project has stalled as a result of a few other things (namely work!) I'm definitely still interested.

I've got an 8 channel - hard wired set up that does moisture, light and temperature. My one uses the 2-nail trick which works remarkably well though you have to change the nails every once in a while. I've been meaning to switch my sensors over so that I alternate the probes being + & - so they don't rust but a pair of galvanised nails last for about 9 months with readings every 10 seconds so the pressure hasn't been great. Though I'm about to update all of them again so maybe I'll "do it properly" and rebuild my shield again. 

My set up / code etc is all on github: https://github.com/ajfisher/plantuino

It all works with Cosm as well so the data just goes up there and then I have scripted triggers to do stuff like ping me a tweet if something needs water or an email if it goes critical. I've also got code + a shield that activates a water pump to turn on the watering system too though I haven't had time to link those two things together.

So there's a fair bit there already that more or less just works and whilst hasn't had a lot of recent work I was hacking on it every couple of days for  8-10 months straight so I'd like to think it isn't a bad base. I'd be happy to chat anyone through what's there - bring some bits in if there was sufficient interest.

Cheers

Andrew

[ajfisher]

Should have added this to the original reply actually. So in terms of the stuff I wanted to do to extend things on the project I started:

- update the probes so they flipped polarity to stop the rusting issue.

- make individual probes a little more smart with eg moisture, above & below ground temp and light for each probe. That could be contained in some sort of shell with an attiny to control it.

- as a result of that, have some mechanism for the probe to report back to a base station. I was actually thinking wired using something like spi/ i2c etc rather than wireless which may be a bit of over kill given distances invoked (living in St Kilda my back yard isn't very big so one end to the other only requires a few metres of cable). Also means the probes would have to be independently powered so cost / complexity goes up somewhat.

- link watering system to the plant state.

- create an alternative to using cosm. I like cosm but would be handy to have a different end point and not have to use their api quite so much.

- have ability to create little meshes of probes + base stations which then link back onto a wider network. So maybe base stations are wireless (ZigBee?) and solar powered and report back to a bridge in the house.

So there's a few things there I was planning. I've got partial prototypes for some of them. Not sure if that's in line with what you guys are looking to do...

Cheers
Andrew

[Luke Weston]

I like the Seeed Grove moisture sensor with the PCB shaped into prongs... cheap and not labor intensive to scalably manufacture.

But you don't even need the two prongs, you could do it with just a single "prong"  shaped board where the copper layers on either side of the board comprise the two electrodes. And all your other electronics, eg. current amplification for the moisture measurement, power supply, temperature and light sensors if desired, microcontroller, communications bus etc. could go on the same board making for a compact unit.

It would make for easy manufacturing if all you've got to do once the components are loaded onto the PCB is apply some insulation such as potting compound or epoxy or silicone to some areas and stick it in the ground without the need to make separate moisture sensors or enclosure etc.

Ideally the communications bus would be something that only uses a couple of wires and doesn't mind being cabled over a relatively long distance... i2c maybe, or something differential like RS-485 or CAN. RS-485 is probably simpler/cheaper.

I2C might work if it's only a handful of meters, but I'd be personally reluctant to endorse something professionally that was pushing I2C past a few meters.

You'd probably end up with 4-wire cables, with power ground and two wires for communications, so something like 4-core telephone cable could be used.

Alternatively, if you wanted to, you could go for a radio network. There are lots of different options here like those cheap generic 70cm FSK modules (not so good if you want more than a single point-a-to-point-b link though) or the Nordic 2.4 GHz chips, or 802.15.4, or 802.11.

Every node will of course need a battery powering it if you want to go wireless, but the battery life should be good if the system is designed to be power efficient. If designed properly you could get good battery life out of an AA battery, but you'd need to keep it in a sleep state most of the time, and periodically wake up the microcontroller (or wake-on-radio if the chipset supports it), turn on the radio and poll the sensors and transmit the data, just once in a while, and go back to sleep.

If the overall system power consumption was kept very low, you might even be able to get away with a small solar cell (think solar garden light) that keeps an AA NiMH cell charged, if the unit was in a location with good access to sun. If it was designed efficiently enough (to keep the solar cell small and cheap) it might just work forever, wirelessly, without maintenance.

You'd also need some sort of gateway that connects the network (wired or wireless) to the Internet if the intention is to do datalogging with Internet-of-Things web services like Cosm. That gateway could take different forms - it could just be a microcontroller plugged into a PC over USB, or it could be a microcontroller connected to the LAN over WiFi or ethernet, or it could be a more powerful embedded computer like an OpenWRT box or a RaspberryPi connected to the LAN and running some appropriate code.

You could in theory avoid the gateway by having every sensor node directly connected to the LAN, either via ethernet (this would require an ethernet cable to every outdoor node and probably adds a fair bit of node BOM cost, so it's probably not attractive, although you could power over ethernet) or WiFi in every sensor node (Not as expensive as you might think, but relatively power intensive compared to other embedded radio alternatives, so the battery system is probably going to be annoying.)

--
You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.
To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msg/connected-community-hackerspace/-/cVwlJiD0AywJ.

For more options, visit https://groups.google.com/groups/opt_out.

--
This email is intended only for the personal and confidential use of the human(s) named above. If intercepted by an extraterrestrial civilization, all opinions expressed in this email are my own and do not necessarily reflect the opinion of mankind as a whole.

[ajfisher]

I like the idea of a prong shaped board. Not least as it will keep the probes at a consistent distance from each other which would be handy. I had attacked this using a very basic pair of screw terminals and then two pieces of galvanised wire - did the job though considerably more DIY. That model would allow for an underground temperature sensor very easily as well which allows you to account for temperature at the point where the probes are actually measuring the moisture and also if you have any plants that require a certain ground temperature to germinate etc it will give you a good idea there.

I'd be interested in having a crack at a solar powered option that charged a battery as I've always wondered how to do that and make it totally self contained. I have 6x 6v, 500mA panels I harvested off something ages ago so if this was something a group wanted to pursue I could donate them to the cause so we could see how they went in a range of conditions etc.

[damie...@gmail.com]

I'm interested.

Looked into it a while ago, but it got put in the back-burner.

Moisture sensors and a feedback loop seem pretty straightforward. Was originally thinking about using a solenoid and a 

small reservoir/tank that waters with the aid of gravity (not a pump), but ran into difficulties -- low-power solenoids usually require mains water-pressure.

A week or so ago at the Hackerspace, I saw an ingenious hack: a generic 9g servo connected to a mini automated watering system tap. Was that yours, and how well does it work?

As for the moisture sensors made from nails (or any other type) being prone to oxidation, do you _really_ need to take a reading every 10 seconds? The environment of a potted plant or garden doesn't change that quickly. Could you prolong the lifespan of sensors to many years by changing the sample-rate to once or twice an hour, with an on-demand button if you're curious for a live update?

Cheers,

Damien

On 5 April 2013 14:18, Greg Nash <nash...@iinet.net.au> wrote:

Hi All,

 

It seems there is some interest at the moment in garduino projects so I thought I start a discussion, my personal interest is in building moisture, temperature and light and whatever else we can think of monitoring and watering automation for a 4 bed vegetable garden based on the diggers club publication: http://www.diggers.com.au/the-australian-fruit-and-vegetable-garden.aspx 

 

The plan was to use some of the basic techiques linked in linked here:  http://www.instructables.com/id/Garduino-Gardening-Arduino/ and here: http://www.instructables.com/id/Garduino-Automated-Gardening-System/ 

 

Now that I've met a few of you at Tuesday sessions I know there's some related interests/projects on the go and it would be good to share ideas, I plan to bring in a new mascot (some kind of easy to care for plant) for Garduino fun on Tuesday so our first challenge will be to keep it alive!

 

To kick off the discussion I'd like to know if any of you have had experience with soil moisture sensors.. especially the cheap ones sold on ebay for arduino or the old two nail trick used in the links above. if anyone has one of these or other garden related sensors lying around and you don't mind us having a go at it I'd be grateful if you could bring it in to play with.

Cheers

Greg Nash

P.S. can someone please let me know if the payment gateway is working..

--

You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.
To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msg/connected-community-hackerspace/-/eZm2KzihD48J.

For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
Damie...@gmail.com

[Clifford Heath]

I looked into automatic soil moisture sensors a few years back, 2005 I think,
and checked a number of commercial designs and perhaps even patents,
and discussed it with some folk who seemed to know.

The difficulty with simple prongs is avoiding electrolytic effects, meaning
both DC voltages arising from the chemistry of the soil water interacting
with the electrodes, and also with corrosion of the electrodes themselves.
Even if you use a very low AC current, you'll still accelerate natural
corrosion.

I think the solution being used at the time was to embed the electrodes
into some medium - plaster comes to mind - and allow the soil moisture
to gradually penetrate the block. The alkalinity of the plaster keeps the
electrodes from corroding, along with the use of AC sensing, to maximise
their life (which was still only a couple of years at best). It might have been
something more porous than plaster, or perhaps it was loaded with
something, I can't recall.

Something to consider anyhow. Find out how the commercial units work,
and you're more likely to have a happy outcome.

Clifford Heath.

> --
> You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.
> To post to this group, send an email to connected-commu...@googlegroups.com.

> To view this discussion on the web, visit https://groups.google.com/d/msg/connected-community-hackerspace/-/5yGR3ujSzBEJ.

[Jan Schmidt]

On Thu, 2013-04-04 at 20:18 -0700, Greg Nash wrote:
> Hi All,
>
> It seems there is some interest at the moment in garduino projects so
> I thought I start a discussion, my personal interest is in building
> moisture, temperature and light and whatever else we can think of
> monitoring and watering automation for a 4 bed vegetable garden based
> on the diggers club publication:
> http://www.diggers.com.au/the-australian-fruit-and-vegetable-garden.aspx
>
> The plan was to use some of the basic techiques linked in linked here:
> http://www.instructables.com/id/Garduino-Gardening-Arduino/ and here:
> http://www.instructables.com/id/Garduino-Automated-Gardening-System/

I've wanted to do the same project at our place here in Wodonga for a
while now, but not started to. Like you, I want soil moisture monitoring
and automatic irrigation from that. I'm still keen on finding a cheaper
water solenoid - the ones I've seen all seem more in the $20-$30 per
valve, which is a bit steep for many zones.

I've been logging temperature, rainfall etc from the wireless weather
station on the roof (results go to http://widgetgrove.com.au/). Adding
ambient light recording would be a cool extension to that.

J.

> Now that I've met a few of you at Tuesday sessions I know there's
> some related interests/projects on the go and it would be good to
> share ideas, I plan to bring in a new mascot (some kind of easy to
> care for plant) for Garduino fun on Tuesday so our first challenge
> will be to keep it alive!
>
> To kick off the discussion I'd like to know if any of you have had
> experience with soil moisture sensors.. especially the cheap ones sold
> on ebay for arduino or the old two nail trick used in the links above.
> if anyone has one of these or other garden related sensors lying
> around and you don't mind us having a go at it I'd be grateful if you
> could bring it in to play with.
>
> Cheers
> Greg Nash
> P.S. can someone please let me know if the payment gateway is
> working..
> --
> You received this message because you are subscribed to the Google
> Groups "Connected Community HackerSpace" group.
> To unsubscribe from this group and stop receiving emails from it, send

> an email to connected-community-hackerspace
> +unsub...@googlegroups.com.

> To post to this group, send an email to
> connected-commu...@googlegroups.com.
> To view this discussion on the web, visit

> https://groups.google.com/d/msg/connected-community-hackerspace/-/eZm2KzihD48J.

> For more options, visit https://groups.google.com/groups/opt_out.
>
>

--
Jan Schmidt <tha...@noraisin.net>

[Greg Nash]

Hi Andrew,

 

This sounds like you are pretty far advanced down the road I want to tread so any contribution you can make would be greatly appreciated, I think for the group we could concentrate on building something which could be implemented in community gardens for low cost then indivicuals could taylor the build for their own projects? anyway I'm happy to see there's some live projects happening even if they are stalled!

 

Cheers Greg

 

[Greg Nash]

Good to hear there are some keen folks out there.

 

I'll be at the space on Tuesday (9/4/13) and I have a plant! The theme will be soil moisture so I also made some nail/gypsum sensors for tooling around and I'll bring some soil pots to test wet/dry soils for any of the sensors. If you are interested in a garden project then come on down and share your experiences.. If we can get enough interest we might even go for our own dedicated day.

 

Cheers Greg

[Keith Franks]

I'm interested in this also.

What time will you all be kicking off from?

Keith

--

You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.

To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.

To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msg/connected-community-hackerspace/-/sfkhEGV6piIJ.

[Greg Nash]

 

I'll aim to get there around 6:30pm with some stuff to get us started in the right direction.

 

Cheers

Greg

[Stuart Young]

Definitely interested in this. I've got a "huge pile of garbage cans" as water storage (about 800L all up) so automating the watering will hopefully save me more of that water (got to Christmas before I needed to start using the tap to water the veggies).

--
You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.
To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msg/connected-community-hackerspace/-/-fKex5dW0lUJ.

For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
Stuart Young (aka Cefiar)

[ajfisher]

Hi Greg,

Cool - I'll try and get down to the space tomorrow night and I'll pull a couple of my bits apart so I can show what I've got. I think I also have some sketches somewhere of what I was going to do re probes as well. Would be good if we could get Luke's advice on how a PCB solution might work on that side. 

A run through of the bits I have built already would take about 10 mins for probes, control board, web integration and watering system as it's pretty basic. As I mentioned above I'm keen to get my teeth back into this after a long stall so this might be the catalyst and I'd view it as a rebuild rather than a tweak given some of the learning I've had. Definitely keen on having a nice base that would work in the general sense (eg for a community garden) that could then be extended further for specific cases.

I can probably get down to the space for about 6:15 / 6:30 tomorrow night.

Cheers

Andrew

[damie...@gmail.com]

Greg, that sounds great. 

Good to see so many people interested in collaborating on this type of project.

Looking forward to tomorrow night for some discussion and demo/test of various setups.

Cheers,

Damien

--

You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.
To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msg/connected-community-hackerspace/-/sfkhEGV6piIJ.

For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
Damie...@gmail.com

[John Spencer]

I'll make sure to bring in my seeed studio sensors and my servo driven taps.

John

[tubular]

Hi Greg

I've got some little LTC3105 energy harvesting designs, that might be of use.  They work happily with a 0.55v single solar cell, or some other sources, and put out microcontroller friendly voltages, reset/power good signals etc.  

I also have some boards based on the LTC3108, and associated magnetics from Coilcraft and Wurth 

You can use them to charge a supercap (or two) which gives you the energy to fire off radio bursts, even with GSM/GPRS modules.  

A few years back I had a look at the Vegetronix VG400 with some friends.  We can't seem to locate it now.   In any case I believe something with two prongs should be superior.   Look forward to seeing how the various sensors perform.  

Btw thanks for seizing initiative and running with this.   Great to see!   I had noticed many hackerspace members had a peripheral interest in plant stuff.   Terry and Jenny last week with their orchids too.

cheers

Lachlan

[Greg Nash]

 

We had a pretty good garduino session last night, thanks to all who came down to contribute and show interest. I think we have started to get a good idea of what we are trying to acheive and how we'll do it. I have to say it hasn't thrown up any significant challenges yet so this could be a good project for young hackers too... I'll do my best to document it so we'll add it to the project wiki on the hack melbourne website http://www.hackmelbourne.org/wiki/index.php/Main_Page and I'll put together a project brief with a list of what we've discussed insofar as deliverable products... if any of these tickle your fancy then feel free to put your hand up and I'll pencil in your name againt it.

 

So far I've got Andrew down to look at wireless solutions

 

I've left an indoor plant (peace lily) at the space with a basic soil moisture sensor stuck in the dirt, the next challenge is to hook the sensors up to the arduino and build a protoype of simple soil monitor, if we can build a v0.1 tiny soil monitor with peizo buzzer (or led notifcation that water is low) thats a pretty good and usable outcome.. and we can plug it into the plant :)

 

I'll do a hunt this week for suitable temperature and light sensors which will be the subject of upcoming dicussions.

 

Cheers

Greg  

[Greg Nash]

Someone suggested checking out Sarah Sharpes open source garden tools and it looks pretty good, she has started developing tools since building a Garduino a while back.

here's a link to her promising looking projects page. http://www.gardengeek.org/projects/

 

[damie...@gmail.com]

Apologies for being a no-show last night (was looking forward to it, but work had other ideas).

Sounds like you got into some interesting stuff. Curious to see what sorts of data get logged, and how it can vary from one situation to another (can stress-test it with my north-facing balcony during summer).

Keen to see more next time. :)

Is there anyone here from http://www.growstuff.org/? 

Might be a useful to share info.

--

You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.
To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msg/connected-community-hackerspace/-/6ixgFNtTXn8J.

For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
Damie...@gmail.com

[Geoff]

Hi Greg

> ...v0.1 tiny soil monitor with peizo buzzer...

I have done a ATtiny85 based resistance soil monitor with peizo buzzer, phototransistor, and with an I2C option, PCB. It may be more like v0.01 :-) The PCB may not be back by next Tues so we may need to breadboard it.

Have you done any wet to dry resistance measurements?

Regards

Geoff

[Greg Nash]

I haven't had as much time as I'd like to work on it this week apologies but we did some basic measurements with a couple of different soil monitors and had some interesting results, but it wasn't really a structured test so I think I'll set that up next week..

 

Basically we had two plaster monitors and a nice rig from Russell who has a working version of the famous servo tap and his soil monitor used two nails in a screw terminal... the nails were not covered with shrink or gal and he used no flipping code so the rig might be a little prone to rust but they are easy corrections to make.

 

The screw terminal rig was highly responsive to changes in moisture from our little test and was mechanically rigid and really easy to service so I think that's a front runner for our moisture monitor.

The "hard to solder" plaster of paris monitors did not have much variation in reading (probably due to my setup/code) and did not respond very quickly to change. I'm thinking these wont cut it.

I made up another screw terminal sensor and we had success on my code too so its looking the goods... I left the one I made there for people to check out - it's in the plant.

 

I didn't test the formed board monitor from john, we can try it next week.... in the structured test we'll run them all off the same code a post the results.

 

I will also be updating the wiki this week with a project brief and some goals... I'll post when I have news on this.

 

[Geoff]

Greg

Would you be willing to bring your plaster and nails? And I bring the plastic netting, I would like to try to make some of the plaster type sensors. I am happy to contribute to the cost of the plaster and nails.

The plaster type sensors I think are better for plants in the ground as they may reduce local effects and even things out over time. 

Regards

Geoff 

[Clifford Heath]

On 17/04/2013, at 9:50 PM, Geoff <geoff.le...@gmail.com> wrote:
> Would you be willing to bring your plaster and nails? And I bring the plastic netting, I would like to try to make some of the plaster type sensors. I am happy to contribute to the cost of the plaster and nails.
>
> The plaster type sensors I think are better for plants in the ground as they may reduce local effects and even things out over time.

One thought. Plaster (… of Paris) is fairly brittle, so will tend to crack if dropped or your nail
starts to rust.

You might get better longevity using cornice adhesive, which is basically just plaster with
additives. I've used cornice adhesive for a lot of things, including casting. It's great stuff,
a surprisingly strong adhesive, etc. IIRC you can buy a 20kg bag for fewer dollars, which is
better value than buying it in 1kg lots. If you anticipate ever living with teenage boys, you'll
need it sooner to later to patch up walls, doors, etc (I know I did!).

I don't think the additives in cornice adhesive would prevent moisture ingress, and the
sensors would definitely live longer.

Clifford Heath.

[Russell Keenan]

I have been wondering how to get data off our little watering systems so we know what our automation is doing 24 * 7.  There is irony in there somewhere, but it's late...

With a focus on keeping total system cost down, I have been trying to weigh up the wireless vs wired options.

The wired solution has the upside of being able to centralise power provisioning and being able to use standard serial connections to a central point.
The downside of wired is aesthetic (but, hey, it is a garden), vague unease about what the children might trip up on / play with and cost of proper outdoor cabling and connectors. 
The connector costs were enough to make me pause.

To cut a long story short, I have been looking at a Bluetooth SoC - the nRF51822 as an alternative to an ATtiny in the designs discussed previously in this thread.  It is small & cheap.
The chip is quite capable (32bit ARM Cortex M0 processor, 128Kb usable flash, 31 pin GPIO) and designed for low-power.
Overall system could be much smaller & cheaper than a wireless shield or an outdoor connector and cabling.

Some info:

http://www.nordicsemi.com/eng/Products/Bluetooth-R-low-energy/nRF51822
http://au.mouser.com/Search/Refine.aspx?Keyword=nRF51822

There has also been a Kickstarter that uses these to make an arduino-like board - the rfduino
http://www.kickstarter.com/projects/1608192864/rfduino-iphone-bluetooth-40-arduino-compatible-boa

Anyone else interested?

Russell

--
You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.
To post to this group, send an email to connected-commu...@googlegroups.com.

[Geoff]

Hi Russell

I agree the rfduino looks cool but I could not find how far it could go. For a garden at lest 100m (class 1) would be needed.

There are two types of sensors needed for a community garden, a simple one that beeps when its daytime if the trees in the common area need water (there are always people around to hear the beeping), and a more complex system, for individuals garden patch, to plead (or provide water) for the plants before they became crispy critters.

http://maribyrnongediblegarden.weebly.com/meg-gardeners.html

Another option is the RFM22-S2 SMD Wireless Transceiver (AU$13) from Ocean Controls. It works on the 430 to 480 MHz band. The lower bit rate of this band should be adequate for this application (some pre-processing could be done if required). The battery based system at the garden patch would send data back to a mains powered web connected unit (like a RPi, pcDuino, etc) that collected all the data from the garden patches (up to 60 at MEG) and emailed, txt, displayed on a web page, set of alarms, etc. as required.

Or

If only one garden patch is to be monitored a RPi , pcDuino, etc that is connected to the web could be used at the patch. An efficient power design would be required if working of battery's is required.

While I do have a particular use in mind (MEG community garden) the requirements should be common to most community gardens that I know.  

Does this help?

Regards

Geoff  

Russell

To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-hackerspace+unsubscribe@googlegroups.com.
To post to this group, send an email to connected-community-hacke...@googlegroups.com.

[Russell Keenan]

Hi Geoff,
I'd be surprised if the Bluetooth chip could go 100m. I'll have to check the specs and see if anyone has some real-world experience yet.

I have a goal of $10 per node, so I can have lots of these all over my backyard. Even $13 per shield would mean I'd want to share the connection between a couple of sensor nodes.

It sounds like we need to capture the layout requirements of our target gardens (/windowsills) and gateway points. 
For example: My garden is in two 'zones' each with about 10 planters/pots. Zones are about 20m apart. I'd like to capture this graphically if possible.

Cheers

Russell

To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.
To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msg/connected-community-hackerspace/-/UFoDBuyKleYJ.

[ajfisher]

So I did some research based on some pointers Andy G had provided yesterday. From playing with this a year ago and getting stuck the challenge is getting something that will work at a reasonable price point that can be embedded in each sensor. You can go with things like bluetooth (range limited) or wifi (more costly) which is good in a point to point arrangement and can get good results. I have have tried standard XBees as well as switched them for Bluetooth and WiFI versions. WiFi with a real antenna I managed to get across 2 backyards to the base station (admittedly I live in a strip of terrace houses so not as impressive as it sounds). Bluetooth was just rubbish.

Personally I think if you were going to go wireless (my efforts to date have been wired) then going something that supports native meshing is a better option. This means you don't have to worry about long point to point comms, you can use the mesh to get you back to your base station. Sure that might mean if you have a massive garden 100m long you might need a couple of nodes between your veggie garden in the corner back to the house (arguably if you have such a problem then the cost of a couple of extra nodes probably won't matter).

There is an Atmel chip that has 6LoWPAN support right off the bat. In volumes this can go quite cheap - sub $10 by the look of it. I'm just trying to investigate what the toolchain looks like to be able to work with it and then what we'd need to do to manufacture boards to work with it. 

As an aside @Greg - I managed to get Moisture, Temp and Light all working off one Xbee and a watch battery fairly reliably the other day on a bread board so the sensors themselves would be dead simple to hook up and control in a "pull" fashion as we discussed if a reasonably priced chip could be found.

Russell

To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-hackerspace+unsubscribe@googlegroups.com.
To post to this group, send an email to connected-community-hacke...@googlegroups.com.

[ajfisher]

So this raises something I'd been thinking about for a while as well which is the notion of localised "wireless" nodes, connected with wired ones more locally. I've got a similar issue in that I have 4 zones. Two in the backyard about 5m apart, one in the side and one in the front. Each zone would have up to 15 or so pots but that doesn't matter too much, it's the separation that's the issue. 

Potentially this would mean that you could have a hub node that has whatever radio solution we go for, attached with some extremely cheap and small individual pot sensors (moisture + light + temp on an ATTINY could make these only several bucks each I guess). Then your wireless one could deal with power and comms back to your base station...

You'd probably need to model this with the optimal number of sensors per wireless node in order to determine whether this is cost effective or not. My sense is that I tend to group plants together into regions based on whether it's warm / cool, shady / sunny so you naturally end up with a clustering of plants together that are then separated from another zone. Seems like others do this also...

Russell

To post to this group, send an email to connected-community-hackerspace...@googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.

--
You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-hackerspace+unsubscribe@googlegroups.com.
To post to this group, send an email to connected-community-hacke...@googlegroups.com.

[Geoff]

I agree, a practical combining of wired local and wireless non-local comm's could offer a flexible solution and reduce cost.

One would not need to have sensors in every pot in a zone, a minimal setup would be just the one in the pot with the highest transpiration rate. Another sensor in the pot with the lowest transpiration rate to see if it is being over watered would be a good idea.

Does the following sum up the thinking so far?:

1. A local hub could do a number of watering/environment zones. 

2. Each local hub could do point to point or mesh comm's back to a base that is on the net. 

3. The base could also have a number of sensors.

4. Both hubs and base could be run on batteries, note: efficient use of power is needed. 

5. Each hub and base could also control a number of solenoids to water the plants.
6. Cost of hubs need to kept as low as possible.


Regards

Geoff

[Keith Franks]

This is awesome guys, keep it up.

I just wanted to mention something a little bigger picture...

There should be two phases to this approach - the Research phase and the Monitoring phase.

During the Research phase data should be collected to enable a routine and pattern to be set up (this might be sensors in every one of 6 pot plants in one area). The Monitoring phase would then alert if the assumed conditions changed (only one sensor would be needed to cover those six pots as light, moisture and temperature should be the same).

This then means that a smaller amount of research collectors can be used and moved around over time - eg you have 6 garden beds and each week the research sensors get moved to the next bed. Only one monitoring sensor would be needed for the other beds to alert a change.

Also, other factors can be added in, for example google weather that predicts if it will rain. If the alert sensor says it hasn't rained, but should have, they can then be watered. This would also work seasonally as adjustments can be made to increase/decrease watering and monitored through the temperature.

So, the six garden bed scenario means that you need only 5 alert sensors and, say, 6 research sensors that would move around. The six research sensors could be wired to a central hub since they will be in one specific location for only a period of time and can be removed if required without endangering anything (just may need to sit longer). The other 5 alert sensors could then be slightly larger and more expensive per unit, since it would still be cheaper than the 6 x 6 units needed if there were research sensors all over.

Keith

To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.
To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msg/connected-community-hackerspace/-/wR8MDTBF3kwJ.

[damie...@gmail.com]

Geoff, I agree that it's good to consider implications of making the sensors more complicated. If you don't need large-scale, it's quite easy to Keep It Simple...but you need power for the sensors if they have a microprocessor and maintain a wireless link. 

If you're going to wire in power, you may be able to include a data line and skip wireless. 

If you're going to go with wireless / remote, factor-in batteries as a consumable, and include some sort of heartbeat signal in your software/sensor protocol.

To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.
To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msg/connected-community-hackerspace/-/wR8MDTBF3kwJ.

For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
Damie...@gmail.com

[Greg Nash]

A few threads of discussion here and I'm stoked we have so much interest.

 

@geoff I've got heaps of nails and plaster and I'm happy to donate some to you. I think we can have those stand alone basic buzzer sensors working pretty easily so I'm keen to push toward that while we work out wireless.

@clifford yep Im worried about how brittle it will be too over time but they come out pretty solid so I think we could be OK (my test sensor survived a week in the soil with no sign of ill effect so far) my current favourite sensor doesn't use the plaster so it may not be an issue.

 

re: wireless.. this is a great discussion thread, I think everyone is pretty much on the same page with cost vs range being the major issue, I like the idea of a low cost mesh network of wireless hubs combined with wired local sensors option.

   

@Keith I agree it would be cool to have a way for the device to research conditions and perhaps even auto adjust watering thresholds based on findings... I reckon we could look at this in software (maybe we could switch a hub/node into research or profiling mode to establish conditions or have common plant/climate profiles downloadable from the web and use live weather forecasts) 

 

I'll add it to the want list :)

 

P.S. this is going to be awesome by the time we are through.

 

 

[Geoff]

I will update the list tonight with the new points made, to give people a chance to add and chat about the requirements. 

And I will have a go at doing a UML Use Case. This is as much for my learning as it is to document the progress.

Regards

Geoff

Geoff

To post to this group, send an email to connected-community-hackerspace...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msg/connected-community-hackerspace/-/UFoDBuyKleYJ.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-hackerspace+unsubscribe@googlegroups.com.
To post to this group, send an email to connected-community-hacke...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msg/connected-community-hackerspace/-/wR8MDTBF3kwJ.

For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
Damie...@gmail.com

[David Lyon]

I have some of the seeed studio water detectors.

Another way of approaching the problem might be to use a Raspberry-Pi. Which is what I'm doing.

Instead of monitoring the soil temperature and humidity I'm trying to come up with an algorithm that
turns on the pump depending on the hotness of the day and the (internet) weather forecast.

So for example if it's hotter today, water more. If the internet says it is going to rain, water less.

Just a few idea's for the from my weed garden..

[Luke Weston]

There are stacks of different options for RF.

It depends...
What microcontroller core do you prefer? (AVR, ARM, 8051? 8-bit OK or 32-bit better?) How much microcontroller speed/memory do you need? What microcontroller development environments do you like?
What RF wavelength do you want? (2.4 GHz, 900 MHz, 433 MHz...)
Do you want something like 802.11 or Bluetooth where it just talks to your phone or WiFi network at the other end of the link without another "gateway" in the RF network connecting it to the LAN?

Do you just want simple, dumb point-A-to-point-B communication, or do you want a more complicated protocol stack for a mesh network of many devices, eg. 802.15.4?

Are you able and willing to design and manufacture a board yourself from the chip level, which usually includes a little bit of RF stuff like the balun or matching network, RF filter, PCB antenna, 50 ohm microstrip lines to connect them, etc? Even if the RF would work (which it won't) you can't stick a 48-pin QFN on a breadboard. :)

Or do you instead want a pre-made board/module, which is going to be easier to design with, but bulkier, and usually a blob of expensive and proprietary and/or badly documented software and hardware?

The Texas Instruments CC3000 will give you 802.11b/g for ~$10 in small volume with minimal external hardware (other than the antenna, and microcontroller with SPI interface) with the IP stack (v4) on-board so you don't have to get it inside the microcontroller's firmware (unlike some others).

Here are a few more integrated RF/microcontroller SoCs to look further into if you're interested.

ATmega256RFR2 and the other devices in that family

Freescale MC13224 and their range of other wireless system-on-chip.

TI CC430F5137, CC2538, CC1110/CC1111, and a few others depending on what sort of microcontroller and what sort of RF you want.

ST STM32W108

Silicon Labs Si1000 / Si100x

To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.
To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msg/connected-community-hackerspace/-/TFicAYYInxwJ.

For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
This email is intended only for the personal and confidential use of the human(s) named above. If intercepted by an extraterrestrial civilization, all opinions expressed in this email are my own and do not necessarily reflect the opinion of mankind as a whole.

[Russell Keenan]

So many to choose from!

Echoing the comments above, I'd like to first look at the ATmega128RFA1 offering - I assume that it was the one referred to as it has '6LoWPAN' on the data sheet.

http://www.atmel.com/images/8266s.pdf

It would be hard to go past if it really does give us wireless networking & the arduino toolchain.  There is a section in the datasheet on backward compatibility with the Atmega 1281, which in turn just has less io than the 1280 (see section 2.2 of http://www.atmel.com/Images/doc2549.pdf )

I hope I'm not missing something obvious, but just using it as a starting point should help us flush out any hidden requirements.

Cheers

Russell

[ajfisher]

Thanks for that Luke.

RF is definitely my weakest knowledge area. Historically I go for "dumb and expensive" with my Arduino based setups - ie XBees using either Zigbee or WiFi. This is largely reflective of me being more a network and software guy than a hardware one so I like the familiarity of stuff I'm used to (don't we all?).

The biggest concern I had about the RF side was really the accessibility of the toolchain. I think it's a fair assumption that the skill of those of us involved so far is largely based on Arduino. So even though there are probably better chips available from other manufacturers, it may raise the bar required to get involved. (And yes I'm aware of being demanding and saying it needs to be powerful and simple at the same time).  Of course, this is a perceived concern - so were someone to go "just install this, this and this, create a make file like this and here's some examples in C" and you're up and running in Win, Mac and Linux then my concern disappears obviously.

The ATmega256RFR2 looked like the front runner for me based on the above point (though there's some assumptions there around having a similar toolchain as a standard Arduino) as well as cost (sub $10) and reasonable RF potential. 

Going down this path implies the design of a custom board with the inherent bits that go along with that (especially the added complexity of RF) - so we'd be pretty reliant on other members of the space to provide some support on this (you interested in the Mentoring role there @Luke??)

Cheers

Andrew

Russell

To post to this group, send an email to connected-community-hackerspace...@googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.

--
You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-hackerspace+unsubscribe@googlegroups.com.
To post to this group, send an email to connected-community-hacke...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msg/connected-community-hackerspace/-/TFicAYYInxwJ.

For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Andy Gelme]

hi All,

On 2013-04-19 12:12 , ajfisher wrote:
> The ATmega256RFR2 looked like the front runner for me based on the
> above point (though there's some assumptions there around having a
> similar toolchain as a standard Arduino) as well as cost (sub $10) and
> reasonable RF potential.

It is possible to acquire a Zigduino now
(http://logos-electro.com/zigduino) based on the ATMega128RFA1 to begin
software development and experimentation. It is rather expensive ($70)
... so designing / manufacturing a less expensive version ourselves
would be excellent. There are several people who are seriously
considering this already.

Whilst the Zigduino can be a regular Arduino ...
http://tronixstuff.wordpress.com/tag/atmega128rfa1 ... to unleash a
proper mesh-network based around open-standards like IPv6/6LoWPAN, one
needs to consider operating systems like Contiki ...
http://www.contiki-os.org

Unfortunately, there is a higher barrier to entry (in terms of software
experience) to Contiki compared with Arduino. However, that strikes me
as a very useful problem to solve.

--
-O- cheers = /\ /\/ /) `/ =
--O -- http://www.geekscape.org --
OOO -- an...@geekscape.org -- http://twitter.com/geekscape --

[Geoff]

Points so far 19/04/13:

Does the following sum up the thinking so far?:

1. A local hub could do a number of watering/environment zones. 

2. Each local hub could do point to point or mesh comm's back to a base that is on the net. 

3. The base could also have a number of sensors.

4. Both hubs and base could be run on batteries, note: efficient use of power is needed. 

5. Each hub and base could also control a number of solenoids to water the plants.

6. Cost of hubs need to kept as low as possible.

19/04/13

7. The wireless comm's, that is each hub point to point or a mesh network back to a web connected base, needs to cover the maximum expected area of the garden.

8. The hubs need to be suited to fixed location and mobile use. That is the hardware needs to be suited the both fixed and mobile use but the software may be tailored for fixed or mobile as required.

9. The system needs to be able to measure, control, and facilitate research.

Note: The discussion of the wireless comm's has raised some very interesting and valuable points and devices but that discussion is more to do with implementation. I believe the requirements need to be more settled before we go to far into the implementation of the project.

[Luke Weston]

"Going down this path implies the design of a custom board with the inherent bits that go along with that (especially the added complexity of RF) - so we'd be pretty reliant on other members of the space to provide some support on this (you interested in the Mentoring role there @Luke??)"

I'm happy to help, yes, sort of. Discussion and emails and stuff are of course always a pleasure, but I don't want to commit to large amounts of time-intensive commitment.

Cheers,

  Luke

To unsubscribe from this group and stop receiving emails from it, send an email to connected-community-h...@googlegroups.com.
To post to this group, send an email to connected-commu...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msg/connected-community-hackerspace/-/G_yOF8sCHTYJ.

For more options, visit https://groups.google.com/groups/opt_out.