Remote beehive monitoring with no mains power - Arduino + GSM + bluetooth? Any ideas?

[Chris Bingham]

Hi,

Calling any low power Arduino/temperature monitoring experts!

Apologies for the length of this, but it's my first posting and I suspect there is plenty of knowledge people can impart so I thought I'd tell you the full story to make sure my initial decisions were not going to block any future development. I'm very new to RLab (came to the open day last week, and joined up on the day) so apologies if I don't know the correct way of getting advice.

I've dabbled a bit with Raspberry PI B - pretty basic stuff (timelapse videos, traffic survey via pixel difference counting between frames) but I'd like to try to set up a method of monitoring my 3 beehives.

Overview:

No mains power so would need low power. They are sited at the bottom of a farmers field. 

Would need to run 4 weeks or more without battery change during winter to prevent killing bees when you open the hive up at this time of year when it is cold. 

There is Mobile coverage but no Wifi

If need be could have solar/wind charger for possibly leisure 12V battery to run it but preferably only the master, not the temperature units themselves

Phase 1: 

Temperature monitor in each hive to monitor brood temperature, talking to a master unit which then sends via GSM phone data network.

Would like temperature units to be wireless to prevent issues when taking hives apart and reassembling. Have ideas to prevent/limit bees propolysing the electronics. Master unit can have a bigger battery since it would be able to be sited where there is more space.

Initial thoughts have been to use Arduino Mini 3.3V since it is lowest power, however I know that some shields (especially the GSM modules) seem to want 5V min. So I was thinking use Arduino Mini 3.3V waking up every 5 mins, take temperature, send to master. 

Master (possibly not Arduino Mini if can't find GSM module to use 3.3V) would listen in and then forward on the readings from each hive over GSM

Phase 2:

Using load cells (probably 4 of them, one in each corner) and suitable Arduino or similar to weigh each hive every 10 mins - probably sleeping inbetween, communicate with the master for that one to send on via GSM

So, here's where I'm at:

Used to be a programmer so alright hacking around in command line scripts, C/C++ etc. Doesn't have to be GUI!

Zero experience of Arduino - but some with raspberry PI

Zero experience of XBee/Bluetooth/low power wireless

Zero experience of GSM modules using mobile phone network

Zero experience with load cells

Phase 3:

Activity monitor to monitor bees in/bees out of each hive (can be used to determine health of colony and also swarming)

Probably via video analysis but power may be an issue since it would have to be always on.

My Questions:

Which GSM module to go for - which will probably dictate the Arduino I need (There seem to be a large number of different ones so any guidance gratefully received!)

Which low power short range communication and protocol to use between the lower power nodes and the master (which has the GSM module) Range required is under 10 metres with only an inch thick wooden hive wall in between!

Which temperature monitor to go for (are any better than any others!) - need accuracy around +-0.5C ideally or better

Load cells - don't really know where to start - but this is phase 2!

Thanks in advance to anyone reading this and even if you know which modules to avoid that would help narrow the possibilities down from the endless list of possibles I've seen!

[Ryan .]

Sounds awesome!

Block diagrams would really help.

--
You received this message because you are subscribed to the Google Groups "Reading Hackspace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to reading-hackspace+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Oliver]

Ditto Ryan. Sounds great!

[Bob Dunlop]

Hi,

On Mon, Jan 23 at 06:15, Chris Bingham wrote:
> Hi,
>
> Calling any low power Arduino/temperature monitoring experts!

Was planning to subvert OpenEnergyMonitor [1] temperature and humidity
sensors (EmonTH module to be exact [2]) for just this task.

Arduino compatable wireless sensor module runs off a couple of AA.
Someone has already written bee monitoring software for it [3] and
I see they now have a new home [4].

I've been side tracked for the past couple of years but hope to restart
my own project this year. As well as temp/humidity I'd like to include
a MEMS accelerometer with a highish sampling rate to "listen" to the bees.


My bees are in my garden so I don't have the base station issues, but
could probably advise on solar PV sizing etc as it's one of the things
we do at work for science instruments. Beware the light fingered if
putting nice looking panels/kit in a field is an obvious issue.


[1] https://openenergymonitor.org/emon
[2] https://shop.openenergymonitor.com/emonth-v2-temperature-humidity-node/
[3] https://openenergymonitor.org/emon/beemonitor
[4] http://beemonitor.org/

--
Bob Dunlop

[Alexander Ross]

solar mppt (=good) charger for lead acid and lithium. uses decent TI IC.

http://www.ebay.co.uk/itm/BQ24650-5A-MPPT-Solar-Panel-Lithium-Lead-acid-Battery-Charging-Controller-Board-/182042324118?hash=item2a62916c96:g:UskAAOSwPc9W1~Tz

[Vance Briggs]

Chris,

I had similar ideas when I joined RLab, then got sidetracked into other projects.  The bees are in the garden though so slightly easier for me.  I have done some projects that are aligned with your flight path.  I have a battery powered gate monitor that is recharged by solar cell. It worked in the summer :-), but during the winter there isn't sufficient sunlight to recharge the battery fully, so V0.1 doesn't work properly for the whole year, but I proved the principle through the summer where it worked flawlessly.  So I have tackled the low power sleep aspects of an Arduino and the use of XBee radios to send data back (also in low power mode).  Moral of that story is to take Bob's solar PV sizing offer.

I also got as far as buying 4 load cells and some instrumentation amplifiers to amplify the very weak signals from the load cells, but that's as far as I got.  Haven't even prototyped this part.

I have also played a little with an Arduino compatible board (Linkit One - https://www.seeedstudio.com/LinkIt-ONE-p-2017.html ) that has 2G GSM modem included.  However, I didn't find it to be particularly reliable, the board would lock-up every couple of days and the watchdog timer didn't reboot it (Could have been my code...).  So I am trying another solution https://www.adafruit.com/product/3027 which also has a 2G modem, but a much simpler microcontroller (this one is an Atmel chip, as per the Arduino).  This runs on 3.3V and the board has a JST socket to take a Lithium Ion battery and charge it.  So far so good, but I have limited time at the moment and I haven't really battle tested it.

We should talk in more detail in the space.  I try to get down Tuesday evenings, but am not going to make it tomorrow.

I hope to catch you at some point in the near future.

Vance

On Mon, 23 Jan 2017 at 20:01 Alexander Ross <maillist_...@aross.me> wrote:

solar mppt (=good) charger for lead acid and lithium. uses decent TI IC.

http://www.ebay.co.uk/itm/BQ24650-5A-MPPT-Solar-Panel-Lithium-Lead-acid-Battery-Charging-Controller-Board-/182042324118?hash=item2a62916c96:g:UskAAOSwPc9W1~Tz

--
You received this message because you are subscribed to the Google Groups "Reading Hackspace" group.

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

[Chris Bingham]

Hey, good links Bob. Lots to go on there. Looks from the data (at link [3] that the humidity sensor side of things doesn't really give much information below the crownboard since it is pretty much always 95% or above, so I'd be tempted to drop the humidity side of things and just focus on the temperatures. This should also save a bit of power. Looks like the DS18B20 digital on-wire temperature sensor should give us what we need and we can have multiple sensors since they are all uniquely identifiable. Or we just start with EmonTH and a single temperature sensor. However I think the Raspberry PI would be too power hungry for me on remote power so I might need an Arduino based master node receiving the sensor information and then using GPRS or IoT to transfer to the server of some kind.

From the data I would also be tempted to lose the above crownboard sensor too.

This would leave 3 temperature sensors per hive, one outside, one in the cluster and one just under the crownboard. The advantage of the crownboard sensor in addition to the cluster sensor is that if there is a sensor failure, (or they propolyse it up), then we have the under crownboard sensor to check to see if it is a sensor error, or if the bees have all left the hive! Thought would need to be given to the siting of the outside sensor since it would need to be on the north side of the hive, or else the sun hitting the side of the hive would give a false high reading. 

So, the layout would be:

(Apologies for ASCII diagram but knew I'd end up spending ages overcomplicating it if I drew it out and it'd never get done!)

All this column are DS18B20            Prob Arduino Mini

Hive 1 cluster temp           \

Hive 1 crownboard temp      ---> Hive 1 microprocessor ----->\

Hive 1 outside temp           /                                               \  

                                                                                       \

Hive 2 cluster temp           \                                                 \ 

Hive 2 crownboard temp      ---> Hive 2 microprocessor ----->--+---------> master node microprocesor -> IoT or GPRS shield -> server (EconCMS?)

Hive 2 outside temp           /                                                 /

                                                                                       /

Hive 3 cluster temp           \                                               /

Hive 3 crownboard temp      ---> Hive 3 microprocessor ---->/

Hive 3 outside temp           /

Gotta go now. Thanks for all the input so far!

Tusedays are good evenings for me to pop into the Hackspace if anyone is going tomorrow (oops - today) evening. Any takers? If anyone has an arduino I would love a quick crash course on them.

[Chris Bingham]

Thanks for the info Vance - especially which 2G modem to go for which works (so far for you) on 3.3V!

[mikethebee]

Hello Chris, 

I think I mentioned The Things Network (TTN) and LoRaWAN with you last week. We have a separate forum for TVTTNUG where we discuss TTN, but the sensor stuff is equally appropriate here. TTN really need use cases and Bee Hive monitoring has been on my list since I saw This Post. We also have contact with Cisco at Green Park who host beehives on their roof. The hives are maintained by a local bee company who ran a 7 week course last year.

Oliver who posted earlier is developing a number of variants of his DevDot Arduino Zero compatible board that I am looking to use for sensor projects myself.

Finally, elsewhere  in this forum we are discussing the Beet Box (sugar beet growing in raised bed) at theMERL (Museum of English Rural Life) in Reading project. Although this doesn't involve bees (directly) it does involve visual, and other sensing. Hopefully your project can leverage and contribute to,some of these similar projects.

Thanks for posting, 

-MikeTheBee

On Monday, 23 January 2017 14:19:46 UTC, Chris Bingham wrote:

Hi,

Calling any low power Arduino/temperature monitoring experts!

...........snip............

[Mark Stanley]

Hi Chris,

Whereabouts are your beehives located (Reading, Caversham...?)  

You might want to consider LoraWAN or other free to use radio as an alternative to GSM.  Declaring an interest - myself and Mike the Bee have started The Things Network in the Reading area which uses LoraWAN 868MHz radio as a means of getting sensor data onto the internet.  LoRaWan has a range of about 1km in town, and more (Mike has achieved 8km) in open space.  Also there are beehives in Green Park such as on the roof of Cisco's building and we are talking to people in Cisco about the possiblity of connecting their beehives to The Things Network - so there is potential for collaboration.

I'd  be very happy to talk further?  You can find out more about LoRaWAN and The Things Network here - https://www.thethingsnetwork.org 

Cheers, Mark

On Monday, 23 January 2017 14:19:46 UTC, Chris Bingham wrote:

[Chris Bingham]

Hi,

The hives are southeast of Wokingham (pretty much in Gardeners Green if you look on Google Maps) so I am guessing won't be able to use the Reading TTN but might be able to use the one listed as planned by mikethebee - although if in the centre of Wokingham it is probably around 2km to the hives, so might not manage that either!

Happy to collaborate - might even bump into some old work colleagues at Cisco (I used to work for Nortel and there are plenty of Nortel people at Cisco!)

I guess it might be an idea to get familiar with Arduino as a start - are any of them easier to work with to learn on than any others?

Cheers,

Chris

[Oliver]

Chris,

You should be okay with Mike's gateway. It should be too much trouble to carry out a test to confirm.

Picking up an Arduino Uno is probably the easiest way to jump into learning.

Mike mentioned my DevDot above, this is similar, but takes advantage of the SAMD processors and incorporates a few different power options that may be useful when we know what's what.
Obviously there are alternative options as well.

Sounds like an exciting use case with some good challenges

[Chris Bingham]

Hi,

I'm not going to make it down to the Hackspace tonight but hopefully will be able to Tuesday or Wednesday next week if anyone is planning to head down!

<snip>

[Gavin]

Hi Chris - my tuppence-worth written from the perspective of having developed my home / rLab’s temperature monitoring…..

Which GSM module to go for - which will probably dictate the Arduino I need (There seem to be a large number of different ones so any guidance gratefully received!)

What range do you need to get the data back to your house? You may not need to use GSM or even LoraWAN if it is less than 1km. Well-shielded  NRF24L01+ boards with stick antennas have been known to work up to 1km over open ground. You can probably get more with a home-made cantenna. They are cheap-as-chips, low power (certainly compared to GSM) and may help you simplify the power set-up at the hive end. They use a very simple protocol over 2.4GHz with 125 channels to choose from to avoid clashing wifi, DECT, bluetooth etc.

Which low power short range communication and protocol to use between the lower power nodes and the master (which has the GSM module) Range required is under 10 metres with only an inch thick wooden hive wall in between!

For my system I used some cheap NRF51822 boards from eBay (£3.50) - using the same protocol as the NRF24s, though they are also Bluetooth capable. The boards have an NRF51822 chip, all the pins broken out, and a PCB trace antenna with all necessary components to be ready to go. At home, I achieve range >10m even through internal brick walls. The NRF51822 is a ARM Cortex M0 based micro-controller with onboard 2.4GHz radio, runs at 3v, and can be put into a deep, timed sleep so that they only draw 3.5uA in-between readings/transmits - this may open up different power options. Programming them is a little harder than an Arduino - you need to lay your hands on a Segger programmer (we can help here) and program them in C/C++ - but not insurmountable if you have basic coding skills.

Which temperature monitor to go for (are any better than any others!) - need accuracy around +-0.5C ideally or better

To have accuracy of +-0.5C you may need to think again on temp sensors as from memory DS18B20s are only good for +-2C. You can buy DS18B20s in a nice waterproof housing and, as you say, you can chain a few of them together, but as far as I’m concerned that’s just about all that’s good about them. Compared to some other sensors, they are slow and power hungry, and despite what the datasheet tells you, don’t expect them to work reliably down at 3v. Texas Instruments TMP112s are good for +-0.5C but I haven;t seen them in a nice waterproof package - though it may be possible to protect them with some resin / solder on some sort of probe wire to stick into the hive.

Load cells - don't really know where to start - but this is phase 2!

I’ve tried this and found ti very hard. If you can buy a kit to do it for you rather than mess about trying to make your own from a wheatstone bridge and instrumentation amplifier, do that.

Happy to take you through what I did / what I learnt sometime. Should be down later tonight if you’re around.

Gavin

[Chris Bingham]

On 24/01/2017, Gavin <gavi...@gmail.com> wrote:
> Hi Chris - my tuppence-worth written from the perspective of having
> developed my home / rLab’s temperature monitoring…..
>
> Which GSM module to go for - which will probably dictate the Arduino I need
> (There seem to be a large number of different ones so any guidance
> gratefully received!)
> What range do you need to get the data back to your house? You may not need
> to use GSM or even LoraWAN if it is less than 1km. Well-shielded NRF24L01+
> boards with stick antennas have been known to work up to 1km over open
> ground. You can probably get more with a home-made cantenna. They are
> cheap-as-chips, low power (certainly compared to GSM) and may help you
> simplify the power set-up at the hive end. They use a very simple protocol
> over 2.4GHz with 125 channels to choose from to avoid clashing wifi, DECT,
> bluetooth etc.
>

5km from my house to the hives and definately not line of sight so
probably LoraWAN and NRF24L01+ not viable options!

>
> Which low power short range communication and protocol to use between the
> lower power nodes and the master (which has the GSM module) Range required
> is under 10 metres with only an inch thick wooden hive wall in between!
> For my system I used some cheap NRF51822 boards from eBay (£3.50) - using
> the same protocol as the NRF24s, though they are also Bluetooth capable. The
> boards have an NRF51822 chip, all the pins broken out, and a PCB trace
> antenna with all necessary components to be ready to go. At home, I achieve
> range >10m even through internal brick walls. The NRF51822 is a ARM Cortex
> M0 based micro-controller with onboard 2.4GHz radio, runs at 3v, and can be
> put into a deep, timed sleep so that they only draw 3.5uA in-between
> readings/transmits - this may open up different power options. Programming
> them is a little harder than an Arduino - you need to lay your hands on a
> Segger programmer (we can help here) and program them in C/C++ - but not
> insurmountable if you have basic coding skills.
>

Used to be a progammer up to 11 years ago so should be able to pick up
the C/C++ side OK. Not done any Segger Programmer stuff though.

>
> Which temperature monitor to go for (are any better than any others!) - need
> accuracy around +-0.5C ideally or better
> To have accuracy of +-0.5C you may need to think again on temp sensors as
> from memory DS18B20s are only good for +-2C. You can buy DS18B20s in a nice
> waterproof housing and, as you say, you can chain a few of them together,
> but as far as I’m concerned that’s just about all that’s good about them.
> Compared to some other sensors, they are slow and power hungry, and despite
> what the datasheet tells you, don’t expect them to work reliably down at 3v.
> Texas Instruments TMP112s are good for +-0.5C but I haven;t seen them in a
> nice waterproof package - though it may be possible to protect them with
> some resin / solder on some sort of probe wire to stick into the hive.

I have a few of the DS18B20s on order and they 'claim' to be +-0.5 but
maybe I'll have to rethink!

Thanks for the heads up on these and also the suggestion of the
TMP112. There's probably no need to have it waterproof if I keep it
inside the hive and put the outside one in a housing of some kind to
keep it dry.

I think I may even have another sensor in stock which I haven't yet
used for my Raspberry PI - I'll have to have a look!

>
> Load cells - don't really know where to start - but this is phase 2!
> I’ve tried this and found ti very hard. If you can buy a kit to do it for
> you rather than mess about trying to make your own from a wheatstone bridge
> and instrumentation amplifier, do that.
>

Maybe I'll modify some bathroom scales to avoid the complicated bit.
There's probably someone out there who's done it already!

>
> Happy to take you through what I did / what I learnt sometime. Should be
> down later tonight if you’re around.

I'm not going to make it tonight in the end. I'm not feeling great so
going to recharge tonight. I'm busy with work commitments tomorrow
too. So looks like Tuesday or Wednesday next week might be an option
if you are around.

Thanks for all the info,

Chris

>
> Gavin
>
>
>> On 23 Jan 2017, at 14:15, Chris Bingham <chrisbi...@gmail.com> wrote:
>>
>> My Questions:
>> Which GSM module to go for - which will probably dictate the Arduino I
>> need (There seem to be a large number of different ones so any guidance
>> gratefully received!)
>> Which low power short range communication and protocol to use between the
>> lower power nodes and the master (which has the GSM module) Range required
>> is under 10 metres with only an inch thick wooden hive wall in between!
>> Which temperature monitor to go for (are any better than any others!) -
>> need accuracy around +-0.5C ideally or better
>> Load cells - don't really know where to start - but this is phase 2!
>

> --
> You received this message because you are subscribed to a topic in the

> Google Groups "Reading Hackspace" group.

> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/reading-hackspace/8PwIo7imD0k/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to

> reading-hacksp...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
>

--

Regards,

Chris Bingham

[Gavin]

On 24 Jan 2017, at 18:53, Chris Bingham <chrisbi...@gmail.com> wrote:

Used to be a progammer up to 11 years ago so should be able to pick up
the C/C++ side OK. Not done any Segger Programmer stuff though.

The Segger is just the board used to program the chip - it’s no more complicated than loading a program onto an Arduino, just more expense if you’re doing this on a tight budget.  

Which temperature monitor to go for (are any better than any others!) - need
accuracy around +-0.5C ideally or better
To have accuracy of +-0.5C you may need to think again on temp sensors as
from memory DS18B20s are only good for +-2C. You can buy DS18B20s in a nice
waterproof housing and, as you say, you can chain a few of them together,
but as far as I’m concerned that’s just about all that’s good about them.
Compared to some other sensors, they are slow and power hungry, and despite
what the datasheet tells you, don’t expect them to work reliably down at 3v.
Texas Instruments TMP112s are good for +-0.5C but I haven;t seen them in a
nice waterproof package - though it may be possible to protect them with
some resin / solder on some sort of probe wire to stick into the hive.

I have a few of the DS18B20s on order and they 'claim' to be +-0.5 but
maybe I'll have to rethink!

My mistake - just checked the datasheet and you’re right. Don’t know where I got that impression from

[mikethebee]

Hi Chris,

As you are in the Woky area please join us at www.thethingsnetwork.org/community/wokingham/ as this will help us build out the network locally. And if you know others in the area who may be interested in using LPWAN please pass on the link. 

I should be available Tues/Wed evening next week, maybe we can meet to focus on your bee project. Tues may be good if we want a quieter time, but if there is only a couple of us then Weds may be better as others that have posted here may be around anyway and may be able to help further.

I am going to visit our antenna supplier daytime Tues to learn more about the art of their tech. 

Many Thanks,

Mike

[Chris Bingham]

Cool. Thanks Mike. I have joined up to Woky TTN. I will email you the exact location of the hives privately to hopefully assist you in antenna planning!

Who's up for a Tuesday 31/01/17 RLab meet up to discuss the bee project?

[Malcolm Napier]

Hi Chris,

Good to hear that you are alive and well.

Does this presage a Binghams Brewery Honey Beer? If so, count me in as lead tester/taster. Mike has my (slightly problematic) LoraWAN gateway - as I don;t have time to sort it out at present. If he hasn't deployed it anywhere else, and you need one to achieve your objectives, I feel that this would be a very good use.

Regards,
Malcolm

[Chris Bingham]

Done - last June/July

Looking to do another this year - initial honey harvest permitting!

Regards,

Chris Bingham

--
You received this message because you are subscribed to a topic in the Google Groups "Reading Hackspace" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/reading-hackspace/8PwIo7imD0k/unsubscribe.

To unsubscribe from this group and all its topics, send an email to reading-hackspace+unsubscribe@googlegroups.com.

[Malcolm Napier]

On draught, bottled, or both?

Put me down for some please, as high up the waiting list as possible.

:-)

Thanks in advance.

[Vance Briggs]

Chris,

Further to the GSM module running on 3.3V, this looks too good to be true but may be worth a look at the price:

http://www.ebay.co.uk/itm/SIM800L-GSM-GPRS-SMS-Module-With-Antenna-For-Arduino-Raspberry-Pi-ESP-12-/172269328231?hash=item281c0d5767:g:M7QAAOSwYSlXglE9

PS - Put me down for some beer too!  I'll take a case.

Vance

--

You received this message because you are subscribed to the Google Groups "Reading Hackspace" group.

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

[mikethebee]

Hello all interested,

Chris and I are meeting up tomorrow evening to make a plan around beehive monitoring. Come and join in with any ideas. I will be there from 7pm, Chris will be there at 8pm.  

-MikeTheBee

[Chris Bingham]

I'll probably try to make it there for 7-7:30pm, traffic permitting!

If the cheap GSM module turns up tomorrow then I'll bring it along - together with my current prototype using Arduino pro Mini 3.3V 8MHz and a DS18B20 temperature sensor. I'll pack a voltage regulator (LM317) just in case it doesn't have one on-board - the SIM800L chipset needs 3.7-4.2V and apparently is pretty fussy on that point. The ebay posting isn't overly convincing either way, despite the title of the link.

http://www.ebay.co.uk/itm/SIM800L-GSM-GPRS-SMS-Module-With-Antenna-For-Arduino-Raspberry-Pi-ESP-12-/172269328231?hash=item281c0d5767:g:M7QAAOSwYSlXglE9

See you all there.

Regards,

Chris Bingham

--

You received this message because you are subscribed to a topic in the Google Groups "Reading Hackspace" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/reading-hackspace/8PwIo7imD0k/unsubscribe.
To unsubscribe from this group and all its topics, send an email to reading-hackspace+unsubscribe@googlegroups.com.

[Bob Dunlop]

On Tue, Jan 31 at 05:22, Chris Bingham wrote:
> I'll probably try to make it there for 7-7:30pm, traffic permitting!

Ditto. Although I don't have anything to bring for the show an' tell.

--
Bob Dunlop

[Chris Bingham]

Just as a heads up, it looks like the SIM on this one has to go flat end first, rather than notched end first. According to the on board LED it has acquired a signal using a contract O2 SIM. Further  tests tomorrow sending commands to it but first indications are that it has some potential to work! Also does not seem to have a regulator on board - won't take 3xAA power - too high -  so attached up a 3.7V mobile battery to it and it was happy. Annoyingly the LED is on the opposite side to the SIM and pin out silks so you can either see the pin outs and not the LED or the LED and not the pin outs when it is on a breadboard!

Thanks for the tip off on this one!

[Chris Bingham]

Just a quick update:

SIM800L (cheap as chips) can be used to do HTTP POST to a Graphana server using a Arduino Pro Mini and a dedicated RTC to use for low power consumption (putting Pro Mini in sleep mode - waiting for wakeup on a pin). Thanks to Gavin for letting me gatecrash his server, I have proof of concept. However, although I have managed to send 2 parameters it has decided to limit me to one in the last half hour or so. To be fair the compiler did warn me that I might run out of room for locals and that it may be unstable. It was right - at least when trying to do more than one parameter value sent over HTTP POST. So I'll leave it running overnight posting the battery level (analogue input pin level read and the pin is in the middle of a couple of resistors running between VCC and GND). I decided this would be more useful to look at rather than posting how the temperature in my house is fluctuating overnight! This way I can see how it uses the 3 x D-Cell batteries it is running off.

If anyone is interested in doing HTTP POST via the SIM800L I had a great deal of fun getting code to work with this particular function. I ended up using this library:

https://github.com/carrascoacd/ArduinoSIM800L

However, since I am using a Multi-Network M2M SIM (O2 and Vodafone) I needed to clone the code and tell it to look for a ROAMING return code rather than CONNECTED. 

So, although I had hoped that the Arduino Pro Mini was a possibility it seems that I have outgrown its 2K of memory. The Zero has 32K of dynamic memory, so I think I should be OK if it is only just outgrowing the 2K of the Pro Mini!

So, I have decided that the Arduino Zero may be the way to go. I can't seem to find much about what its current draw is in either of its sleep modes so I have decided to add it to my growing Arduino collection (now have 5 x Arduino Mini Pro, 1 x Arduino Uno and 1 x Arduino Zero on its way) and then test it when it arrives. 

I can see how collecting electronics can get to be a habit! Thanks to everyone for their help so far.

See ya

Chris

[Richard Ibbotson]

Hi Chris, great to see you are making good progress on this.

Did you try the usual ways to reduce RAM usage, like putting strings and fixed tables into program memory only?

The Zero is a good way to go, and it should be able to run on about half the power of the ATMega based devices. It also has the RTC inside already. Power consumption is mainly determined by the external circuits and the library controlling the internal module power supplies and clocks.

Oliver and I think Gavin have some other alternative SAMD boards which may have more simple external circuits and make low power more easy.

Richard

--

[Bob Dunlop]

Hi,

On Fri, Mar 17 at 06:33, Chris Bingham wrote:
...

> So, I have decided that the Arduino Zero may be the way to go. I can't seem
> to find much about what its current draw is in either of its sleep modes so
> I have decided to add it to my growing Arduino collection (now have 5 x
> Arduino Mini Pro, 1 x Arduino Uno and 1 x Arduino Zero on its way) and then
> test it when it arrives.

Welcome to the world of ARM programming :-)

Looking at the CPU datasheet [1] circa page 790 it looks like the CPU will
draw approx 4uA in standby maintaining the RTC, and 2-3uA in standby alone.
No idea what the EDBG draws but would hope it is minimal.

Looking at the Zero's circuit diagram it uses a liner regulator to go from
5V to 3.3V, so there's options when chasing the last few uW for a final
unit. For example if you don't need 5V you could bypass the linear reg
and adjust the 5V switching regulator to generate 3.3V directly.

ps. My own nRF51822 based project is moving exceeding slowly. Too many
other distractions. I have at last got a dev system setup how I want it
and got my own code running on the chip, but that's about all.


[1] http://ww1.microchip.com/downloads/en/DeviceDoc/40001882A.pdf

--
Bob Dunlop

[Gavin]

There are alternative dev boards also based on Cortex M0 -based chips, varying in size and also cost - worth also looking at the Adafruit Feather M0, Arduino M0, and I have also used the Atmel D21 Xplained Pro, which is larger but gives you access to more of the pins. The advantage of the Arduino boards is that they are mostly pin compatible with shields, but don’t expect full compatibility with some Arduino libraries - they have sorted the core elements I think, but many community libraries are written for ATMega based boards and have not yet been upgraded.

The M0 cores can put into v low consumption sleep modes, but with boards you can still be left with the current consumed by the regulator and other components such as the programming chip. Two options to address this are:

- look for a stripped back board where you can bypass the regulator and power it directly off a 3v battery supply, something with the barest minimum of components (will likely require an external programmer);

- go for an external power management chip such as the TI TPL5110 (Adafruit tutorial here), which will put not just the MCU but the whole board to sleep for you (credit to Pete T for pointing this one out)

Glad to see you’ve cracked the POSTing

[Dragos Comanescu]

Hi all.
Sorry, my reply is a bit off topic but still related.
I am trying to do an alarm with espruino PiCo, esp8266 and a pir sensor.
Anyone interested in this please let me know maybe we can collaborate.

Andy

[Chris Bingham]

Well, I'm learning. Thanks for the tip on moving strings to program memory - a pretty straightforward change to the code too. Just slap an F() around the sting and voila! Down from globals using 76% of dynamic memory to only 54%. No longer get the compiler moaning. Not sure what is going on now. I might swap the SIM800L module since it doesn't seem to want to connect to the network anymore. There's a big benefit in buying several of each component part! Now, where did I put it.....

[Chris Bingham]

Thanks Bob for confirming that the zero current draw is nice and low!!

[Chris Bingham]

Yep. I've spotted these. I think I'll start off on the Zero itself first, then once it is working on that I can start downsizing. Given the fun & games I have had with the Arduino Pro Mini I think proving the port to the new processor is best done on a genuine to start with. I also like to give them a donation by buying a genuine version once in a while!

[Chris Bingham]

So it looks like my SIM800L was faulty. Swapped in a new one, and it is now working again on the Arduino Pro Mini. Now it is time to try to get the SIM800L to go to sleep.....

[Pete T]

Hi Andy,

Sounds interesting.

I have some experience working with esp8266 modules.  Initially with Arduino Sketch/C, more recently using Micropython. Not used Espruino Pico yet.  

Cheers,

Pete

[Chris Bingham]

Hi,

Quick update on the SIM800L 

The SIM800L draws around 13mA in idle, but if you send it the following command then it goes into Slow Clock state drawing only around 1.3mA - (essentially asleep) and waits for a wakeup: 

AT+CSCLK=2

If you look at the SIM800 data sheet then it only mentions AT+CSCLK=0 and AT+CSCLK=1, and you then have to use the DTC pin to wake it up. However, I spotted that the SIM900 supported AT+CSCLK=2 where it then goes to sleep and awaits a signal on the serial port. I tried it and it works a charm, although best to put a delay(1000) after it before any microprocessor sleep code so that the serial command has time to go through before it sleeps. I spotted that without the delay the command sometimes doesn't make it through in time.