@Yveaux I'm embarrassed to admit I did buy a RTL-SDR last year on account of the problems I had with 433MHz sensors on the RFLink gateway and wanting to decode wireless m-bus. But I never really learned how to interpret the data from SDR#. I booted it back up just now and am flailing around in the interface 
@evb Thanks I'll try that.
I was thinking about setting up a second gateway also (I have another sensebender gateway I'm not using) and running a different frequency for the network. The new driver allows you to set basically any frequency that the module supports right down to the Hz?
And the 868MHz modules are able to go as high as 915MHz (even if it isn't optimal) and vice versa? I do recall something from Lowpowerlabs from years ago that said the 915MHz ones could at least be run at 868MHz? Felix and other users on their forum even confirm it. So that would rule out the possibility that I have radio modules of the wrong frequency,
I am starting to question some of these radio modules I bought off of AliExpress 3-4 years ago. I know for a fact that I have RFM69W modules that are 915, don't remember why I even have those. The signal scanner used a good and known *W 868 module. All the *CW modules are marked as 868. But I am nevertheless starting to wonder if there could be some foul play here. Either knockoffs, or real duds. It would explain a lot of the problems I've had with HopeRF modules over the years, with them often being outperformed by knockoff nrf24s.
No neither the temp sensors nor the mailbox sensor are working on 2.3.0 with the new rfm69 driver at any serious range at least. The hardware is this (https://github.com/EasySensors/ButtonSizeNode) and this (https://github.com/EasySensors/SwitchNode). Have tried different types of antenna.
The hardware difference would be that the scanner is running on AA batteries with no regulator and the sensors above use 3V button cells and some sort of regulator I think. Plus they use RFM69CW while the scanner uses RFM69W.
Like I say above, I've really only disabled debug to save some memory because I read in some other thread that it in fact helped. I have not tried setting TX power manually on either the gateway nor the sensors.
Still 5bdm on the gateway doesn't sound too bad though. What was the gateway TX dbm on 2.2.0 and old regular driver?
So I've been a little bit inactive for the last couple of years. I recently updated my nodes and gateway to 2.3.2 to see if it would fix some of the range issues I've been having and because it seemed convenient to be able to see RSSI and TX power. I also built the signal scanner (https://forum.mysensors.org/topic/7822/portable-rfm69-signal-scanner/6) so I'd be able to walk around and see if communications were working in certain spots.
Gateway (rfm69hw) initialization:
#define MY_IS_RFM69HW
#define MY_RFM69_FREQUENCY RFM69_868MHZ
#define MY_RFM69_NEW_DRIVER
Example node (rfm69cw) initialization:
#define MY_NODE_ID 0xAC
#define MY_PARENT_NODE_IS_STATIC
#define MY_PARENT_NODE_ID 0
#define MY_RADIO_RFM69
#define MY_RFM69_FREQUENCY RFM69_868MHZ
#define MY_RFM69_NEW_DRIVER
I have a couple of nodes in my garage that are just damn near impossible to get to work. It's odd because I have deployed nrf24 modules there in the past and while transmission has been spotty, 868MHz should be superior in this regard, at least not perform worse. The nodes are all the same, but are using three different antennas. I originally had #define MY_TRANSPORT_UPLINK_CHECK_DISABLED and #define MY_DEBUG with those nodes but commented it out this afternoon on two of them. Still only one is coming through and that is the one with the commercial thick coil antenna that you can find on ebay. I recall reading somewhere on the forum that people had good results with those (https://forum.mysensors.org/topic/7845/rfm69-antennas-comparison).
Another node that I have out in the mailbox and that had been performing flawlessly on 2.2.0 I think it was and the old driver is now completely out of reach. I did the same commenting out as above but it still wont work.
I guess my first question is regarding ATC and how it works on 2.3.2. In particular in this case where nodes are either initialized within range of the gateway and brought outside or initialized out of range and installed. I understood the system as starting at the lowest dbm and working itself up until it reached an acceptable RSSI? How does it proceed if it cant reach the gateway?
Also I was wondering about disabling ATC and setting another modem config? Would it be a better troubleshooting strategy?
#define MY_RFM69_ATC_MODE_DISABLED
#define MY_RFM69_MAX_POWER_LEVEL_DBM (13) // 13dbm/20mw max on rfm69cw?
#define MY_RFM69_MODEM_CONFIGURATION RFM69_FSK_BR9_6_FD19_2
Anything else I can tweak in the settings?
I should add that the "signal scanner" (with a rfm69w) works really well with the code in the link. Transmit % is about ~0% next to the gateway. About 50-60% in the garage. And 70-80% at the mailbox. Perhaps a little bit lower once it has settled in those spots. Nothing special about the hardware either. Built it exactly like described.
As previously noted, Sensebender Gateway sketch from 2.3.2 with MySensors SAMD Boards 1.0.6 does not compile with Arduino SAMD Boards 1.8.9. Rolling back to Arduino SAMD Boards 1.8.8 or earlier version works.
@Denke Any updates to this project? Did you deploy the sensors? Any issues or has it been smooth sailing for the last couple of years?
Is there a suggested replacement for the TSL250? While it can still be found, it seems to be discontinued for the most part.
I've had no luck with the LM393 solution for my new power meter. Just can't tune it to pick up the led impulses. What sort of diode solution is everyone else using? I'm aware of the possibility of IR communication (https://wiki.hal9k.dk/projects/kamstrup) but it seems a little over the top.
While looking into the water meter I realized that we are getting a "smart" meter replacement soon. I'll see if I can speed up the replacement schedule. Anyhow, the replacement is a Kamstrup Multical 21 (https://www.kamstrup.com/en-en/water-solutions/water-meters/multical-21) which has several options for reading:
@flopp Did you ever manage to stabilize the readings? Am I understanding you correctly that the readings were fine but the transmission got the reading scrambled?
Did you try to find a magnetic field with the meter instead? I have a similar meter and I'm wondering how best to approach the problem. I've seen people hook up Raspberries and cameras to do OCR basically but that seems like way too complex for an easy problem.
@koresh Package went under the "Postnord" radar! Woohoo, perfect! I will start replacing the old capacitors and finally be able to deploy the sensors. Very nice of you to include some product samples as well. Really excited about these sensors!
@alexsh1 I may have made a bit of a mistake with some glitchy chinese test probes. De-glitched cables and connected to a proper lab power supply gave very different readings (with the default git-sketch) ...
Still a bit confused over the differing readings. I get that the board I stripped from regulator and capacitors would perform better though. So do I have to do the same for all the other boards as well? Of course, I could save the worst performing boards for later and run them on 2xAA or something in places where space isn't an issue.
@alexsh1 Thanks for the suggestions! Reassuring to know that you succeeded.
You're probably right about something not sleeping. Or else I've done a horrible soldering job on the four radios I've attached so far. All four sensors behave pretty much the same.
Also, at least on Domoticz they changed things around now so you have to use V_LEVEL instead of V_LIGHT_LEVEL or what? https://www.domoticz.com/forum/viewtopic.php?t=18577 With V_LIGHT_LEVEL I get some incomprehensible percent value reported.
@yury No I ran them on just a 2032 battery at the time, just had it in an external battery holder. I also removed the LDO on one of the boards and bridged the two top pads, with no apparent effect.
Unless there is something wrong with my DMM.
@Koresh I still have a bunch of v1 boards that I never got to work because of power consumption (I presume). I thought I'd make one more attempt.
I tried bridging the jumper and I'm using a 2032 battery. Consumption varies between 2 and 4mA on different boards. On some boards it fluctuates a lot. The instructions above are not immediately clear. Do I have to remove the LDO AND capacitors as well?
I'm sold on Fusion360 as it is basically professional 3DCAD that can be used for free. Full on parametric with a powerful history tree. As you've no doubt seen it has built-in CAM functionality as well, equally professional and it can output G-CODE for most CNC-controllers. So very much all in one. And you'll be glad you learned real 3DCAD instead of meddling with 123D, Sketchup or anything else.
Personally I prefer the Vectric programs (Aspire) for making G-CODE for the CNC as they are geared towards woodworking with features like v carving. Fusion360 CAM feels more like it was made for mills and lathes but can do most things as well.
@rfm69 I think the mechanical stuff is ~$1400 and then ~$600 for steppers and spindle and a few other things.
Of course the machine we're building at the makerspace is the biggest suggested model. Working area ~100x150cm with steel reinforced Y and X and with laser cut steel plates instead of aluminum. So I don't think he ever sold that as a kit internationally because of weight and size. A smaller machine would be much cheaper obviously.
@neverdie Yeah the guy behind the RawCNC has been a bit difficult. This is sort of his side project and he keeps changing his mind on how to proceed. He actually rolled back some of the features to make the machines lighter and easier to ship. But then he didn't sell internationally anyway. At least not for long.
Yes but can he do this? 
https://youtu.be/Nz8k2MSAasI?t=47s
The MPCNC is a smart project but I would assume it would struggle with a bigger motor.
The 2.2kW we are going to use weighs A LOT. Around 5kg at least I'd guess. The 1.5kW is not much better. With a smaller spindle, or router like a Kress things would be different. Still I'm glad to have the extra power, and standard ER20 collets up to 13mm. Plus these chinese spindles are really quiet compared to a regular router. And the speed can be set and spindle can be started from the control panel or from the computer. The downside is the low quality VFD I guess, but if it works it works. And you can always use shielded cables everywhere.
@neverdie Yeah that is a surprisingly good machine as it seems to be built from scratch with no plans or whatnot. It says in the Youtube comments that he spent $4K-4.5K on it though, with 1K of that being for the extra cost of servos as compared to steppers. Nice to have those linear rails and ball screws though.
By comparison the machine we're building is maybe $2K. It will have worse precision (than ball screws) as it is rack and pinion. But a work area of ~100x150cm.
The RawCNC uses rods that are glued or pressed into a shaft clamp and then a u-groove bearing on top of that.
Well, we looked into this for our local Makerspace. Especially if you're in the US, the Shapeoko or the Xcarve can be somewhat price effective, particularly when they have their sales campaigns and include the dewalt router and free shipping or whatever. But I still think the Xcarve is a bit flimsy and mostly propped up by a successful campaign in social media where they gave away machines to well known youtubers.
In the end we opted for building a rack and pinion RawCNC instead.
http://cncmaskiner.org/
http://rawcnc.com/
Unfortunately the guy behind the project keeps changing his mind constantly about how to sell machines, kits, parts and plans so I don't know how practical it is to start a build right now. It all hinges on finding the parts like the aluminum extrusions.
In the end we didn't want a timing belt machine. And we wanted a vfd / spindle solution from the start not having to go through a dewalt / makita router first. These spindles are quiet and you get standard ER collets. But they are also really heavy and might not work on a standard Shapeoko without reinforcements. We opted for a 2.2kW air cooled model.
Here is someone who has custom built an unusually small RawCNC:
https://openbuilds.com/builds/rawcnc-1-5-desktop-edition.5771/
@NeverDie Perhaps but the design is a bit too simple and unoriginal. I'd have to clean up the files a bit and I haven't touched Eagle since I made this board several years ago.
Also pretty happy with Domoticz. Also installed Dashticz on one of my two installations (the other being at the local Makerspace). For me, at least historically, I needed something that could run well on a Raspberry. Domoticz is up and running in seconds obviously even on limited hardware. For me at least Domoticz is pretty flexible, with Blockly, LUA, Python and dzVents. I think there is also some sort of add-on for PHP. The interface and the handling of devices is a bit messy but still.
Haven't tried HA but I would certainly consider it.
Could we do what Moteino has done and drop support for HW/W radios and opt for HCW/CW/RFM95/96/RFM12 using double sets of pads?
https://lowpowerlab.com/shop/product/99
Would be great if the community could settle on just one radio or a common way of doing things.
Still very much interested in buying the board with RFM69 footprint! Just saying!
@scalz Would also be very interested in buying assembled modules. But why no more RFM69?
@Koresh Sounds great! Oh and btw, any idea how customs will handle the packages, any chance of taxes and customs fees? You're outside of the EU customs union right?
@Koresh Thanks! Great solution with the adapter board as well, you really do think of everything. Do you have the other boards without radio as well? Specifically the Switch Controller?
@gohan Thankfully I do have some CW 868 radios (~5 left) available that I bought ahead of time. I think everyone can solder the radio to the board by hand and so I personally would prefer that @Koresh would consider selling the parts separately.
@gohan Yeah I could but I'm just too lazy There are a lot of components to source from AliExpress before anyone can build these boards. It would take months even if there aren't any errors in the BOM or lost packages from China.
Great! One question though, is everything already soldered and if so is there an option to for 868MHz?
Just want to confirm that I also received a "VL-C702X-2 VER:C1" today in the mail so at least they haven't made yet another version.
Oh and another question, since I've never ordered anything from ITEAD before, how do they declare the customs value? Is it safe to order more than one gateway or should I order them one by one to be on the safe side? I'm talking about "SF E-Parcel" or "Registered Air Mail" of course, not TNT or DHL.
Which exact RFM69 is the footprint for? Cause it says "RFM69HCW" in some places, and "RFM69HW" in others. The footprint looks a lot like HW (https://c1.staticflickr.com/1/563/20328459340_62fa380554_b.jpg).
@Koresh Where do you get the rfm69(h)w rev 3.0 from? Or is it the hc model? I can only seem to find rev 2 (?) on Ali?
Fantastic! Finally some hardware debouncing! This would make a perfect mailbox alert system as well right? If a little overkill perhaps.
Do you know of any integrated hardware debouncer solutions besides going for schmitt triggers? (that are affordable!) I did look at the MAX6816/MAX6817/MAX6818 (https://www.maximintegrated.com/en/products/interface/circuit-protection/MAX6818.html) series myself, but I never got around to using them, and they aren't super cheap exactly and only operate from 2.7V. Surely there must be other manufacturers doing similar ICs?
@NeverDie Yeah you're probably right, but a CR123 is damn expensive from where I'm standing, compared to a single AA cell. I'm perfectly fine with changing batteries more frequently if it comes to that. The only thing that would sway me would be sub zero temperature performance. I'd sooner try a coin cell and see how far I'd get. I'm looking for compact solutions for mailbox alarms and the like where the sensor would need to be almost paper thin to stick to the lid of the box.
I love it! More boards like this!
I would concur with using a TQFP atmega if you're not doing manual assembly anyway.
As for other requests / suggestions .... perhaps you could make a separate version which has less pins broken out and instead has room for a flash memory. Also I am still very much in favor of more extreme step-up ICs, that can take say 0.6-0.7V from a single AA/AAA and boost it to 3.3V. For some reason no one is going that route on the forum anymore and I don''t know why? Has everyone switched to lithium batteries?
Really nice to see someone integrate the radio on the board, haven't seen many do that so far! Seems like the logical choice but for some reason very few have made that leap.
What kind of regulator do you use for step-up battery power? Could you shrink those oscillators down a bit?
Great idea for a compact button sensor such as a snailmail notifier or some such.
Any thoughts on hardware debouncing ICs, such as MAX6816, MAX6817, MAX6818? Not to be a backseat driver here but something like that would be cool for a dedicated button sensor.
Fantastic project! I also have a ton of these Wirsbo thermostats (from circa 1999) at home. Some of them have failed and have been bypassed. Remarkably expensive to replace and remarkably dull design wise. I always wanted to build something like this, possibly with the addition of a small oled display and a 3d printed or cnc machined (wood) enclosure.
Just a quick post to confirm I got the relays working as well (with Domoticz). I didn't read the schematic at first and just assumed how the mains and relay external wiring should be done. Still no idea about the smoking varistor.
@m26872 Indeed, mine is 5.08V (with no load). I ordered some 5.1V and 5.6V 1206 SMD diodes from Ali for future use. Plus some 3.6V. And a set of DIP diodes.
@m26872 You're probably right, I just followed the BOM. I'll see if I can stock up on some diodes from AliExpress for future use.
Hmmm. I got the 5.5V DC varistor from another AliExpress seller and I'm a bit puzzled as to why it starts smoking after a few seconds. Tried several of them. The HLK gives a solid 5V output. I'm assuming I got some lower voltage varistors by mistake?
@Dave-Dan Perhaps you are experiencing some sort of level issue, between 3.3V and 5V? But then it probably shouldn't work when connected to a computer. Still I'd try to use a 5V arduino -> 5V ring or 3.3V arduino -> 3.3V ring if possible. I think it should work on 3.3V without any issues.
Worst case scenario, use a separate 3.3V voltage converter or use a level shifter / converter.
I did Hek's Parking Sensor and has loads of issues with the ultrasonic sensor running on 5V and the Arduino on 3.3V. Signals wouldn't reach the sensor.
I don't have the sensebender, but I have a node with a regular 3.3V Pro Mini and a 501 that has been going for a year now. No regulator. Just 2 x 1.5V AA. Don't know how it is still functioning as I also read that I should expect false positives as well as false negatives. The sensor is outdoors so it is hard to say anything about false positives since there are birds and stuff passing by. But I have never ever known it to miss a beat as I walk by (which is something I do 3-4 times a day).
Just connect the 3.3V to the third / "H pin" on the PIR board.
http://forum.mysensors.org/topic/444/battery-powered-motion-sensor/13
Magnetic switch or tilt switch (ball or mercury) would be a safe bet. Other possibilities would be a mechanical switch either at the top or bottom (that reacts to the weight of the mail). A light sensor / resistor (which obviously wouldn't work at night) or perhaps some sort of IR beam (not very battery friendly I imagine).
Personally I have been using tilt switches for a couple of years. Very jumpy though when it comes to debouncing. Never tried with a capacitor as a simple hardware debouncer but rather just run the tripped message every loop and then sleep at the end of the loop, first for 10min and then until the next interrupt. Domoticz sets the sensor as "off" automatically after a time. A bit hacky but it works it seems. Reading the tilt switch state and getting it right every time became too frustrating.
Speaking of debouncers, I was almost desperate enough to look into debouncer ICs like the Maxim MAX6816/MAX6817. Almost.
A bigger concern is perhaps radio range. Despite earlier trials that seemed to indicate that the range was ok (and activating ACK), I had to set up a repeater node (with a pa+lna no less, the normal module wouldn't work!) and manually make sure the mailbox node always sends through the repeater. I was planning to set the repeater up eventually anyway, this just forced my hand.
@DrJeff I actually bought the 0.2A version myself before it was gone But I was surprised to find how hard it was to find axial ceramic fuses of the right size or at the right price. Maybe we could find some sort of small fuse holder instead?
The SSR has an integrated heat sink according to the specs. Don't know if this is a knockoff and follows the same standard of quality though. Given the low A rating, and the fact that most of us will be switching a 10-20W worth of lights, I think it will be fine.
I noticed the ceramic "Slow Blow Fuse (250v 0.3A)" was out of stock on Ebay. As was the 0.2A version. Someone is stocking up Would it have to be a ceramic fuse?
If I recall correctly, the step up module from Ebay (I think it is in the store?) ended up being pretty good after all. Someone on the forum tested it long term I think.
I picked Linear LTC3525 and Texas Instruments TPS61221 for my temporary modules.
http://forum.mysensors.org/topic/930/step-up-boost-regulator-pcbs
In the future though I hope to have a board with everything on-board. No add-on boards or anything. Except for the "smd" radio module. Look at the custom boards that people have started making:
http://forum.mysensors.org/topic/1857/my-sensor-node-motherboard-mysensorsnode
http://forum.mysensors.org/topic/1419/sensor-board-with-booster-and-supervisor/31
Very impressive! Particularly that BOM! A lot of money to save by avoiding Mouser in favor of AliExpress. And you didn't even factor in VAT? (at least in Sweden I've never paid VAT on anything from Ebay/Ali, the post office is probably swamped by these packages, which is probably the biggest selling point)
Overall, I'd like to say this is probably the best MySensors board to date. Of course, I'm partial since I like boost regulators, but still. Where can you go from here ... the only thing really is making everything even smaller, perhaps two mounting holes will be enough, and move to 0603 or beyond, QFN for the 328p etc. If it is worth the effort. The board is still pretty small as it is and should be hand solderable for most at this point.
Particularly like the way you've included both the RFM and NRF, and that you've used the the "SMD" version of the NRF. That old module was a real waste of space.
@Anticimex I can understand that. Plus the SOT23 is a tad easier to hand solder than the SC70. I guess I will have to get around to make my own sensor node eventually. Just gotta finish the reflow oven first.
@scalz Yeah, I've been looking at the 328p-mu (qfn) from AliExpress. If it is a counterfeit, so are all the other ICs on the Pro Minis we're all using probably. And those capacitors really are dirt cheap!
If you have the patience for it, you can probably source your components from AliExpress easily. I've been building up quite a stock of components and so far it has been smooth sailing (except maybe that one seller who sent the wrong item, money refunded, but time lost).
What sort of step-up regulator are you planning on using for the final version? Your schematic says TPS61097A? Any particular reason?
The sort of selfish reason I ask is that I have a whole lot of TPS61221 and LTC3525 already Could never find TPS61097A from AliExpress, only the older TPS61097.
Another vote for AA/AAA. I have one motion sensor node that has survived outdoors for almost a year now on the same pair of AA, and it doesn't even have a regulator ... it surprised me a lot. Go for alkaline. Low self discharge nimh may be better, but they cost a lot.
I'm going to to out on a limb here and say RFM69. The range and wall penetration is in another league altogether from what I've seen, given similar input current. And if I remember correctly you can force much lower bit rates than NRF24? Also, afaik, no knockoff issues. The issue for me is that I now have a ton of NRF24s and I'm reluctant to switch plus the RFM69s are hard to come by outside of the US, where the manufacturer has an official outlet. I've seen them occasionally on AliExpress but they have never seemed to catch on.
Would be great if we could get ITEAD to carry them perhaps?
@jerry
So, my reflow build is coming along ... slowly. Looking into plotting curves on a larger LCD, like the Zallus oven and using a small servo or stepper motor to open the door.
One thing that is bugging me though is the reflect-a-gold tape. Costs a fortune and I can't even seem to find any or even an equivalent product in Sweden. What do you guys think about the Ebay / Ali knockoffs? Are they good enough? I mean I don't care if they fall short a few percent in heat reflection, I'm not building a spaceship here, as long as the adhesive holds up.
@epierre I'm afraid I haven't gotten around to it yet. Partly because I have my CCTV that automatically feeds me images of whoever has the misfortune of stepping on the porch and triggering the motion detection so I haven't been in any great hurry exactly. I have all the components around here somewhere though.
Don't know if the ACS712 + logic level conversion is still the easiest approach or if we're over-complicating things here.
Nice design! Are you happy with the NRF24L01(SMD) or did you experience any range issues?
Just as a reference, the boost regulators I did, based on the TPS61221, cost something like $1.7, including the (very small) PCB. The TPS61221 is about $1 out of that cost. And I did not use the cheapest inductor exactly. Plus the cost of small volumes and ordering each component individually from AliExpress.
Don't have any testing equipment, and I haven't done any long term tests, but the prototype I did was perfectly able to run a mysensors node, including the motion detector module, and a small buzzer, on just one AA battery.
Not to say that I also wouldn't like pads for the Si7021 at least. It's a damn fine chip, hence the cost.
I think that in the long run, we should think about some sort of mini shield system where a base board can be extended, so (almost) everyone can get what they want. At the expense of bulkier designs, but it would be a choice for the end-user and everything comes at a cost anyway.
In all fairness I think it depends a lot on the seller. And it probably is safer on Ebay since there is PayPal acting as a buffer and a second tier for complaints if all else fails. I have only ever opened one dispute on AliExpress. Ordered 50x NR3015T4R7M 4.7uH inductors, got 1x "some other weird chip" I'd never heard of. Got a refund after a day or two without any complaints. Was probably like a $8 order or something. I'm sort of stuck with AliExpress since it's the only way I can buy components at a reasonable cost and avoid Digikey/Mouser.
Speaking of battery powered cameras, there already are commercial products such as Arlo and Homeboy:
http://www.cnet.com/products/netgear-arlo-smart-home-security-kit/
http://www.cnet.com/products/homeboy/
Not cheap enough, and battery times will surely vary depending on use, but they are a nice proof of concept.
I've only tried this module so far for prototyping. Worked as intended using the default code. Haven't made any measurements on it. But it ran fine in a 3,3V circuit. Certainly loud enough. Some current limiting would probably not be a bad idea.
Edit: The module I have is of course passive. Not sure if that is what you want or not. Active vs Passive.
Haven't tried Home Assistant, didn't even know it had MySensors support until now, but it does look pretty damn cool. Perhaps the best looking controller of all. And it has really useful plugins like open-wrt/tomato and sabnzbd. Not sure how well mysensors works with the controller though, supported sensors, and automatic discovery.
If you really need Python it is pretty much your only choice.
Can't really recommend any Java or Node.js solutions. Except for PiDome. Your other safe bet is to try Domoticz.
Regardless of what you choose you may want to install something like monit (example instructions) to make sure your controller software is running, otherwise restart the software and even automatically restart your Pi if the controller crashes too often or too quickly.
Interesting stuff. Would have bought a couple of them if it had another voltage regulator more battery friendly. And the authentication chip. Also, with all the talk of bad nrf24 chips, it is probably time to get around to making the radio integrated on the board. But just soldering the nrf24 smd board to the back is obviously so much easier. It's weird that no one has looked into an in-house mysensor pcb for the smd module yet.
@Mickey are you still happy with your SMD nrf24 boards? No issues with range or other weird issues?
Giles did something like that with his Micromesh board (https://hackaday.io/project/2492-micromesh-iot-mesh-network). Did not turn out well apparently.
Yeah, it is perhaps not the most economical solution. Probably closer to $100 per unit. Foscams have probably become better and cheaper since last I looked. Personally I'd rather have more control over my setup. Like how the housing is ventilated. Plus all the fun of building and figuring things out. Of course I don't get night vision, but I figured it is better to trigger a relay to turn on some floodlights or whatnot that will give intruders some warning. No use being sneaky about it.
But yes, it would be really interesting if someone could come up with a really cheap battery powered solution.
I have one (soon two) RPi with camera modules that take pictures and store / send to Pushbullet when a certain MySensors PIR is triggered. Has worked pretty well for the past year. and a Pi + Camera is still reasonably cheap plus it's a "real" computer so it can do anything with the camera and images. And the Pi camera is still pretty good, unless you happen to get a real bad clone from Ebay or something. Hardly worth it. I don't know how good these cameras in the store are, but I doubt they come anywhere near the simplicity and guaranteed quality of the Pi camera module.
If you like, you can also use the "motion" package on the Pi, but that will be a real hassle with false triggers and all. The cheap PIR motion sensors have worked surprisingly well. It's more a question of positioning them right so you don't have problems with sunlight, fireworks, curious magpies etc.
I have my RPi inside camera housings I got from AliExpress. There are a couple there that are pretty good. I think finding a bracket can be more of a challenge. I have J-brackets that fit under the eaves.
I have the raspimjpeg script running on the Pi btw. You could just use raspistill I suppose but it's nice to be able to manually view the camera feed from the browser and have remote control at your fingertips. Hell, you could even get one of those motorized RPi things that are able to turn the camera.
I have a sort of Power Over Ethernet going for my cameras btw.
In case anyone is wondering, shipping costs seem to be as follows, for example (to Sweden):
1 board: Registered Air Mail $4.52 or DHL $23.01
5 boards: Registered Air Mail $6.43 or DHL $23.01
...
I assume this is the standard "tax-free" Chinese packaging system when using air mail? (i.e. no invoice in the bag, not the real value on the bag)
So, the LEDs I ordered from Ebay got lost in shipping. What are the odds. Happened maybe two times out of 250 Ebay orders. I've been refunded of course but I now doubt I will continue down this track. Thinking of going back to 230V and relays instead + some cheap fixtures from IKEA.
@activemind Not a bad idea with the flood light housing, seeing as it is so difficult to find outdoor housings. I just hope they used stainless screws and bolts.
The Arduino Pro Mini could theoretically accept 12V on the RAW pin. If that is a good idea and a long term solution I don't know. Buck converters are pretty cheap and have worked with the Pro Mini and Radio Module when I've tested them. But I'm sure if I had looked at them using an oscilloscope I would feel different.
Interesting stuff. I have similar valves and thermostats to control them. It's an older floor heating system where the thermostats are starting to fail. I've been thinking of doing something similar since the thermostats, even ugly ones with just a dial and no lcd cost $50 a piece or something. It really starts to add up after a couple of floors.
Great work! Really impressive!
@Staringlizard Cool! Is it this oven you're using? 14L, 1200W? Are you happy with the temperature ramp-up time?
@GaryStofer Haven't used Ago Control in a long time. I wouldn't have recommended it in the past. Didn't like the AMQP deal. And there were other nuisances. However with version 1.0 coming up perhaps they have made enough improvements to warrant a new look. Support for the ImperiHome Android app is a big deal for instance since their own Android app was pretty basic. If you're not using 1.0 from their testing repository I can understand your frustration.
Personally I'm quite content with Domoticz for the time being. Easy to use and just works.
@pedrodova I think AliExpress has worked out nicely so far, though I can't be 100% sure I'm not getting counterfeit components from time to time (hard to tell with capacitors). Though it would seem unlikely with the rather uncommon components here, I mean if someone were putting out fake ICs they'd be all over Ebay/Ali and they're not obviously. There are only a few people selling LTC3525 and only one guy selling the inductor on Ali for instance, and prices are usually on par with Digikey etc if you look at volume sales and disregard shipping/customs. Considering the whole debacle with the NRF24 radio though I'll certainly be vigilant in the future in case things change.
I have ordered perhaps ~250 items from Ebay and ~25 from Ali and only ever had two things disappear in shipping (both from Ebay and one quite recently). Never had an issue with a seller not issuing a refund or re-sending a broken item. AliExpress packages are almost always sent as registered mail (as opposed to most Ebay packages) so I'll end up having to pick them up at my local post office as opposed to just being shoved in my mailbox. Shipping to Sweden takes about a month most of the time and some sellers are pretty slow to ship (though it is always stated on their item pages). If someone claims to take 15-20 days to even ship, you better believe it. Still haven't figured out why package delivery times differ so much either. Like I said, a month is what to expect but I've literally had packages (mostly from Ebay) arrive within a week.
Biggest issue with AliExpress is perhaps their search engine and the way some sellers list their items. Trying to find what you're looking for (or alternative products) can seem impossible sometimes.
Overall I'd say AliExpress is pretty safe and a nice way to save a few bucks on components - if you can find them, wont mind waiting and don't have a ton of components on your shopping list.
@shabba I guess any approach would work, if your sensor is awake and listening for incoming messages. I actually have a similar idea for the front porch where I will use an audio player module (WTV020-SD-16P - glitchy stuff but easy to find) and an amplifier (PAM8403) with a pair of 4Ξ© 3W speakers to get prospective visitors to crap their pants. It seems best to have the sensor do the triggering in that case. But why not have it set up to play another file as a warning or alarm when triggered from the controller.
Another vote for Domoticz. C++ ... need I say more? Now that it has both serial, ethernet (and soon MQTT?) support it should be the first thing to try. Not as nerdy as OpenHAB perhaps, but it just works. LUA scripting can do a lot if you need something that can't be set in the events editor. Like for instance how I run a python script every time a certain motion sensor triggers that allow for a camera to take a picture and push it to my phone.
Pidome was pretty cool last time I looked also. It should have progressed a lot since. Don't know about the platform compatibility though, I mean it is Java after all, but it was geared heavily towards the Pi.
These ovens are killing me here. No matter where I look I can't seem to find a convection oven at a reasonable price with quartz elements. It's either one or the other, or crazy prices or ridiculously large volumes that will take ages to heat without the addition of extra heating elements.
Here are my candidates so far:
http://www.amazon.it/Ariete-Cousine-Petit-Forno-Elettrico/dp/B004JLN2EW
10 litres, 1000w. Quartz. High W for the volume if it checks out.
http://www.amazon.co.uk/gp/product/B00K22YJEY
9 litres, 800w. Quartz. Looks a lot like the one Rocketscream used with great success.
http://www.harald-nyborg.se/wasco-miniugn-18l-1380w.html
18 litres, 1380W. Not quartz? But can be found locally. A bit low wattage for the volume.
I went and picked up one of these boxes. As suggested, the box fits perfectly, has clips (no screws), removable terminal blocks and has a nice stylish front that would probably work in most homes.
Thanks again for the tip! At 20 SEK ($2.3) it is quite affordable also. Makes me question 3D printing and CNC routing.
@kanchana Looks great! Have you had any issues with range etc when the radio is so close to and hanging over the pcb?
I've been wanting to move to the mini nrf24 module as well.
Powering a 1.3 inch OLED display (i2c).
Now if I could only put all of this together as a scene controller or whatever in a nice 3d printed box.
@funky81 If you look closely at the regulator, there should be some marking, whether it be a clone or not the regulator ought to be or appear to be a Micrel Mic5205. It should be either KB33/LB33 or KB50/LB50 for 3.3V or 5V.
Unsettling stuff. Thanks for the warning. Makes me wonder exactly how advanced the Chinese clone factories are, I mean, if they are making this relatively advanced IC, is any IC safe from China? Granted that only high volume stuff is worth cloning.
Another reason to move away from the NRF24 if opportunity arises. Genuine (?) RFM69 are looking better and better.
So, in conclusion I assembled the boost modules. I don't have an oscilloscope or any fancy variable load so I can't do much testing. But both the TPS61221 and LTC3525 seem to be working as intended, they power the NRF24 (not sensor in this particular picture below) without any hiccups at any rate.
I tried some ceramic power resistors to put some load on the regulators and got something like ~60-80mA out of one fairly run down AA cell if I'm not mistaken. Basically according to spec.
All I have to do now is design a compact sensor node where 1xAA actually makes sense. Kind of like LowPowerLabs' Moteino. Feels like it's time to take this to the next level. The only thing bugging me is the need to reflow instead of hand solder, and the massive task of sourcing components from AliExpress so I can avoid Mouse/Digikey. Perhaps I should give it a rest for now and see how the official MySensors modules turn out before I decide.
Another nitpick ... there are a lot of different Pro Mini pin layouts (http://arduino-board.com/boards/arduino-pro-mini). I don't know where you buy yours and if you can be sure of a certain pin layout, but I would probably not bet on anything other than the side rows that are present on all boards. A4 and A5 are fairly certain, but other than that ....
I think I looked at that box the last time I was at Biltema but I never got around to buying them. The ones I got from Ebay are ok, though a bit bulky in retrospect, plus the screws rust like crazy. So I'll pick up a couple of these next time. Can never have enough boxes.
Yeah, you're probably not saving a whole lot of money doing your own arduino-like board from scratch. Can't seem to find cheap ATMEGA328P-MU on AliExpress either, only ATMEGA328P-AU. But then to really save space you'd have to use the "mini" nrf24 module as well.
Yeah, my concern with the MCP1640 is that the link I posted to eevblog is on to something and that it is misbehaving. They never seemed to hit their uA targets, though it may be user error.
My main motivation for other regulators though, like the TPS61221, is that you need two less components, and thus less board space. I think the TPS61221 is about twice as expensive as the MCP1640. And the LTC3525 is about twice as expensive as the TPS61221. Those are the only regulators I have on hand, but there have been MANY more discussed in various posts. For me it is convenient that both the TPS61221, LTC3525 (and MCP1640) are available on AliExpress so I can avoid expensive shipping costs and probable customs fees.
@phil83 I think the 4.7uF electrolytic is a bad recommendation for this purpose. It has been discussed a couple of times but the recommendation hasn't been changed. Probably because no one has actually tested and come up with a solution that is proven. It doesn't really matter though, you could retrofit any cap you want later. I've mostly put 1206 ceramic no-name caps on mine, soldered between to the nrf24 pins themselves and it seems to work. But I have no oscilloscope or esr meter or anything so it is impossible to know what is going on.
@phil83 Well I don't know much about the MCP1640. It was tested here: http://www.eevblog.com/forum/projects/working-with-mcp1640-boost-converter-high-iq-at-low-load/
It is probably fine. Pretty cheap but it's annoying having those two extra resistors on the board. Other than that you could shrink your inductor and capacitors a little. They look a little big. And spaced quite far apart? Soldering shouldn't be difficult as long as you do them all in the right order so you wont obstruct yourself.
That surface mount electrolytic ... 470uF? Higher capacitance would yield lower ESR sure, but wouldn't a simple ceramic 4.7/10uF or whatever work just as well? Conveniently the same type also used for the boost regulator.
Interesting. Would be fantastic if MySensors would get more attention (and funding).
Classic click-bait journalism by IDG through. They make it sound like anyone with a breadboard is basically eligible for a stipend, while it is be targeted at Hacker / Makerspaces. I think there are 10-15 of these places in Sweden right depending on how you count? So if that is the criteria there should be plenty of money to go around. To be completely honest though I'm not sure most of these places are innovative enough to warrant any interest. But I guess that at the rate that governments are spending money anyway, it's great that Makerspaces get a slice of the pie finally. There are surely worse ways to spend taxpayers' money.
@Anticimex How do you manage SPI_EN from software? I seem to have run into the same problem with SPI and the Ethernet gateway (despite having run the MQTT gateway for months with no issue). Thinking of doing the hardware hack on the ethernet shield to avoid soft spi (better to solder one more wire than de-solder and re-solder a bunch of wires).
@slarti I see now that at least GRBV doesn't render the output gerbers correctly. Weird. Oh well, mine worked fine and so will yours probably.
I think I deleted the .GML. I also recall some sort of error if I didn't. "Board outline not found in GML/GBR/GKO file" or some such.
@slarti Yes, that happened to me as well. My guess would be that you have your outline layer in just about every gerber. Some board houses (like DirtyPCBs) like it that way or have it in their CAM anyway. The panelizer however only works if you include the outline layer ONLY in your .GKO.
If you send the files as shown above, your board house will probably figure things out, but then again they may not.
I just switched to the ethernet gateway to try out native Domoticz support and quickly experienced the SPI issue that everyone has been talking about. Can someone explain how the MQTT gateway was able to run for several months with no issues, but the moment I flash the ETHERNET gateway on the same hardware everything goes belly up after a few hours?
@tbowmo The last time I needed a non-standard thread (for a camera tripod-like mount) I had to go to WΓΌrth. They charge about an arm and a leg for a few screws, but on the other hand the metal is probably very high-grade and resistant to corrosion.
