Interest for esp8266 sensor hardware?

[pmayer]

Hi guys,

we are a small german company dedicated to building things. We came from the web development industry but shifted our interests more and more to embedded development. We've developed different customer solutions like power measurement systems, digital signage solutions and even a scoreboard for a footbal club using mqtt and other techniques.
As a side project, we've developed and designed our own access control based on beaglebone black and AVR microcontroller devices using SHA1 iButtons, which is nearly finished now. This was done together with a PCB specialist which is one of our contractors.

More an more customers come to us and ask for integration of home automation devices. One of the biggest issues are the sensors you need to automate your home. Beginning this year we've discovered espeasy and use it in custom build sensors with mainly i2c temperature and humidity sensors, the DS18B20 is still in the business

We've also experimented with battery driven devices which use the espeasy sleep mode, as the biggest problem is to not dig up the walls to lay new cables. So why not use an existing infrastructure as easpeasy does? Wi-Fi!
The idea is to build buyable dedicated hardware with an integrated charging circuit for a rechargable and replaceable battery in a nice form factor with pluggable sensor modules. Maybe for the first run only one type of sensor. All this packaged in a nice case with a high WAF and without any need for a power adapter (but optional). An optimised antenna circuit and maybe an external antenna is one of our ideas. With a sleep time of maybe 15minutes this device should run 3-6 months.
Also we would like to open this up as much as possible. We are planing to integrate a button for flashing and pins for RX/TX so you can repogramm it or use your own firmware. As I understand it right the licenense of espeasy allows this.

My question now is, is there any interest for such a device? How much were you willing to pay?
I try to understand if it will be a good idea to invest time in this and start manufacturing on the long run.

Would be very kind to hear from you...

Cheers,
Patrik

[mrwee]

Hi,

This sounds very interesting and I think you're on the right track. In terms of pricing for such devices, you can look at ITEAD and see how the "competitive landscape" is. However I'm sure German Engineering will add true CE marking as well as better quality in terms of look & feel which rightfully can add some €€ to the price. Looking forward to see where this can take you. Good luck from my side

[pmayer]

Hey mrwee,

thank you for your kind words but we won't have the "cheap" prices of ITEAD
The PBC's would be manufactured in germany and we will meet EU-standards for safety. Of course we won't switch mains which will otherwise give us a headache to get verified by all authorities. Maybe later...

Cheers,
Patrik

[pmayer]

Guys,

could you give me some feedback here? We're already planing this but have to know if our effort wont be useless...

Thanks

so long,
Patrik

[dlefol]

Hello,

I think it can be a good idea as, for me, the main drawbacks of homemade system based on ESP are the lack of casing and the necessity to have a power supply. The main issue will be pricing I think. Most of the people on this kind of forum are happy to DIY their own sensors or to buy cheap ITEAD like sensor and mod them to their requierments (which for me is the main advantage of ESP based sensors). If you go higher in price for a sensor with only a specific set of functions you have some nice sensors already available from companies like fibaro or others. Personnaly I would not spend 40-60 euros on a ESP based sensor which I cannot modify as I would rather go for a z-wave one in that case (more suitable to be battery powered).

That s just my take on it so not sure it s representative.

[pmayer]

Hey dlefol,

thank you so much for your feedback.

Our idea is to deliver a base PCB with a charging/protection circuit for any Li-Ion 3,7V battery including an ESP-12E which will fit into a nice casing which we are evaluating right now. So you could use 18650 cell or any given 3,7V cell with a JST plug which could be exchanged if empty or which could be charged right in the casing. Also you could just run of a wall-wart if you want.

Further we want to have interchangeable sensor modules which plug into this "base". At the moment we think about a simple connector, like a pin header, for interconnecting the PCB's which exposes all GPIO's of the ESP usable by espeasy. So, if you want, you could hook-up your own circuits. Also we think of a pluggable proto-board where you can just solder your own circuitry and blug it into the "base".
So you could fully build your very own sensor and extend it for espeasy as you want.

The main advantage over a self built solution will be the integrated charging circuit and the finished - espeasy ready - PCB.
One of you problem on casing is, for example, how we could expose sensors to the outside of the case, e.g. a luminosity sensor. A version for outdoor use is another thing on our whishlist.

We hope we could get the pricing without battery but with case to something around the 25€ mark. But please don't nail me down on that

Cheers,
Patrik

[Stuntteam]

Maybe you should be more clear on what exactly you are trying to produce.
It sounds like a PCB with an esp and a single sensor like a ds18b20 (possibly on exchangeable modules) in a little case for around 25-30 euros and then batteries have to be purchased seperately..
Sorry, but why on earth would someone buy that?
Arguments that your PCB is of high quality are pretty much useless when you are using off the shelve products and are 10 times more expensive than other solutions ..

[pmayer]

Hi Stuntteam,

thank you for your feedback. I will try to explain it again:

We want to build a battery powered Wi-Fi MQTT sensor using espeasy with open source hardware, including the charging/protection circuit and an interchangeable Sensor module in a nice casing. As already said, the plan is that anyone can build their own sensor or use prefabricated modules and plug in into the "base". Every sensor espeasy supports should be buildable.
For the case we are evaluation but the ones with a high WAF are arround 8-10€, maybe cheaper if batch ordered. Something like this: http://thumbs2.picclick.com/d/l400/pict ... -KLIMA.jpg

Don't get me to wrong on the pcb's, that was only an example for fabricating lead free and ROHS compliant. Depending on the volume the pcb will be cheaper, too. So for the assembled pcb, including esp, charging controller/protection circuit, optimized antenna circuit, usb port etc. we aim arround 17-20€. Thats in the range of a huzza but fittable in a case to mount it in the living room without hotglue or something.

Please point me to any solution you know - I wanted to get feedback from you guys to exactly know if the effort is worth it. We don't need to build this if everyone is happy with the available modules/boards/cases

Cheers,
Patrik

[mrwee]

Hi Stuntteam,
Something like this: http://thumbs2.picclick.com/d/l400/pict ... -KLIMA.jpg

Sorry but to me, that box seems to have higher WTF factor than WAF . For me this http://www.energibutiken.se/en/temp-sen ... ensor.html looks much nicer. Don't know where to source that box .

But I'd say you've caught my attention . I'd like to see:
- Good form factor PCB which might fit into other boxes (bought from you as PCB, no casing) for misc. other usage.
- PCB could include step-down from 5V/12V with terminal block for fixed power.
- e.g. block terminals for GPIO
- Jumpers for e.g. sleep-mode on/off (GPIO 16?)
- one or two onboard LED which can be seen externally (tied to GPIO for custom applications)
- Possibility of using external antenna (for those poor coverage applications)

Looking forward to seeing where this will take you

Edit: Updated the link

[pmayer]

Hey mrwee,

thank you so much for your feedback, again. I think the casing itself is discussable or of course changeable. The biggest problem there will be the sourcing... We must get some of them in our hands to evaluate.

In theory you've just posted our whichlist The jumper for deep-sleep is a no-brainer.
Depending on the charging circuit it will be possible to use fixed power. We have to evaluate that.
The GPIO should/must be accessable but it should be possible to plug-in modules.
One of our bigger concern are the antennas. As we already did an evaluation in a HF testing ground and no tested module was realy good. Our PCB manufacturer optimized the antenna circuit and, if wanted, we can build in a connector for an external antenna. I have to admit we did not do long range tests with the PCB one. Any hints?

What do you thought about the two LEDs? Any special usecase?

Cheers,
Patrik

[mrwee]

Of course . I've tried to source the box i linked to, as it seems relatively small and has a clean design, but didn't find it .

I'm no antenna designer, but I agree that the antenna performance for a lot of the ESP8266 based products are not great. I'm lucky enough to have 3 AP's in our home so we're well covered, but I can see that placing an e.g. nodemcu in the garden is problematic as it almost doesn't reach, so I think a proper antenna design is quite important. Don't know if it's too much, but the PCB could perhaps have a design with allowed an external antenna like this one: http://www.rj45-jack.com/sale-5322498-i ... ector.html, mounted directly on the PBC with a hole in the enclosure for it. Just a though.

The two LEDS could be used for showing data transmission on one LED and e.g. setpoint reached yes/no (e.g. is the temperature below a set value), contact open/closed etc.

Actually the controller for our waterbased floor heating just broke down and the whole system needs to be replaced . Imagine having a WiFi based thermostat with LCD display, and keys for setting temperature . But battery life might be an issue.

[pmayer]

Hi mrwee,

the antenna you've posted is exactly how we think about it. Also this should be optional. We have to evaluate how it will be possible to use either the internal or external antenna.

The device is in deep-sleep most of the time. It only could give a short status, like espeasy already does with the internal LED of the esp2866 board. A self programmable LED would be nice and should be nothing that drives the price to high. As this will make the casing sourcing/designing more complex, though

I think one of the keys would be not to overengineer this - clean simple battery powered wi-fi sensor device. So, battery life would be the biggest issue. You could always expand the functionality when using direct power. Regarding your use case I think this will be the better way.
Even with a 3400mAh 18650 battery the device will last a few days when not using deep sleep.

Cheers and thanks again,
Patrik

[Tommi76]

Hallo pmeyer,
very great idea. Maybe you are working on what i do need.
So here are my wishes.

- The housing is not the problem.
- USB for loading the battery and programming the device, is a must be.
- (Optional) extra Board for supply voltage and loading the battery.
8 - 30 VAC, 11- 28 VDC
(230 VAC would also be very nice, but i think, that is too dangerous for most users)
- Multiple headers for SPI and I2C.
- Driver for white LEDs. I think 100mW to 300 mW would be enough. (Maybe as an extra modul)
So you could realize a small night light, with a PIR Sensor. The light should switch on instantly without waiting for a WLAN connection.
e.g. your child is waking up at night. A small light would be much more useful in this case, than a 10W LED bulb at the ceiling...
- RS232 to RS485 converter!

I have some main field of applications.
- Room sensor: Temperature, humidity, (pressure), PIR Sensor, and night light (and maybe an E-Ink display to show the measurements).
- Only Digital I/O: Look if a Window or Door is closed, and to switch several relays.

My main use would be a connection to the FHEM projekt.

Greetings, Thomas

[Drum]

In my opinion the nodemcu has more than antenna issues, but I don't have any problem getting 100 meters line of sight out out of a ESP-12F, or even a Olimex wifi-mod, with a good power circuit providing 3.3V.

I would love to see a well designed board and case where I could plug in sensors and battery. I think there are several people here trying to design that. What I want exactly is difficult to define. The capabilities of the software have improved so much over the last few months. 1 board is probably not enough.
Cases and good reliable standard connectors, indoor and outdoor for both.
Good quality components are a must. Using a garbage voltage regulator to save 0.25 / unit is a sin in my book.


The adafruit feather huzzah is the best dev board I have used so far, if they managed to add solar it would be my perfect dev board, or as close to perfect as I have seen.

[dlefol]

In terms of casing, I found that the idea used here :
http://www.nodeusb.com/

is quite nice as you can just add sensors via a simple connection and it s easy to plug in a charger to have power. The casing is not the prettiest but that can be improved on i guess

[pmayer]

This is by far our biggest concern. How do have a descent casing which supports all wanted sensors. Of course only sensors which make sense in a battery powered application.

Still evaluating... but thank you very much for your feedback and wishes. Keep 'em coming

I'm totally with you, Drum.

[Drum]

I kind of like the way some of the telephone network interface housings are made, from a conceptual side. Connectors are modular, added as needed, modular and separate, connection to PCB could be modular or fixed, would be easy to separate power and data. The board can be is a separate compartment which protects it from accidental damage when changing sensors, I have knocked wires loose when changing sensors. They are also weather proof, not water proof as the bottom is open, but this would be good for most of what I do.

Something else to consider, if you there is only 1 or 2 sensors, a ESP-8285 might be better. The PSF-A85 module is really tiny. I think you could make a pcb for it about the same width as a AA battery. Maybe a tube shaped case, cap screws off to replace battery(ies), sensor (s) in a opening or extending out and a small antenna (or PSF-B85 has built in ceramic antenna). 1MB internal SPI flash does limit firmware, but if it has a limited number of sensors, I think this is not a big problem. Sometimes smaller is a good thing. This is something I will certainly play with, just because I can make the whole package smaller, which means less obtrusive.

[pmayer]

Drum, thank you, we will look on every point you made. Right now we are evaluating like hell.
ATM we are trying out battery life with a 18650/2400mAh and a sensor delay of 15min. We have to know how long this thing could last.

Meanwhile a friend hinted me the Sensortag which is already built by TI: http://www.ti.com/ww/en/wireless_connec ... =sensortag

For that price the feature list reads like a dream:
- Humidity and Temperature
- Ambient & IR Temperatures
- Barometric Pressure
- 3 axis Accelerometer
- 3 axis Gyroscope
- 3 axis Magnetometer
- 2 push Buttons
- 1 Luxometer (CC2650 version only)

I'm curious for the Wi-Fi Device. Obviously this is not in the Arduino ecosystem but nevertheless interesting which way they go.

Cheers,
Patrik

[pmayer]

Hi guys,

some updates from my side. I've already got a ESP-12 with ESPEasy runnig for over seven weeks with one 2400mAh 18650 Cell. Right now I'm testing different battery capacities and different sensor combinations. Looks good on that

I've found an interesting kickstarter with a similar goal: https://www.kickstarter.com/projects/91 ... ity-sensor
It looks like they are using the TI CC3200, judging from the demo board they had lying on their table. Sadly this thing is missing mqtt but that is only a thing depending on software...

So, if this kickstarter will be successfull we have to think hard about putting any more work into our idea.

Cheers,
Patrik

[pmayer]

Happy Holidays

So, I think this idea is dead as zeroflow over at reddit is already developing a suitable hardware board: https://www.reddit.com/r/esp8266/commen ... tform_pcb/ - http://imgur.com/a/gZbui

Anyway, thanks for your feedback. We've learned much with this. Maybe there's something in the future where we can use this knowledge.

Cheers and a happy new year!
Patrik

[manjh]

A simple Chinese sensor: measures temperature and relative humidity, small LCD display, battery operated, transmitting the values into RF:ink/Domoticz. Can be ordered for $8.64, including shipping to EU. http://www.banggood.com/433MHz-Wireless ... ds=myorder
I know you are aiming at WiFi transmission, and I'm sure quality will be superior. But just to set some stakes, have a look at this price...
I have several in my home, quality is good enough for me.

[pmayer]

I'm with you manjh.

But the aim was to use a existing infrastructure and - of course - to have it enrypted. I've just don't want to sell a customer a chinese thermometer without any support or something. Support, availability and simplicity was on my mind with this idea.

Don't get me wrong. For the enthusiastic (home) user this is great

Cheers,
Patrik

[sovking]

Take care of two non chinese good European examples like:
- ESP8266-EVB
- ESP8266-EVB-BAT-BOX
They are manufactured in Europe and the quality usually is good.
There is a relay, a lot of pin for new connectors, a button for programming and a lipo connector and charger.... and all schematics and manuals freely available.
Drawbacks: a little bit pricey (but you can buy it on amazon), and the fixed voltage at 5v (I would prefer 7-30 Vdc and maybe 230 Vac version too).

However that are good starting point with can be improved providing variable voltage, more plugin-sensors ready to work, more plugin-actuators and a nicer case.
But the starting price for the base board with a single relay and a box have to stay not more than 15 Euro to be competitive.
Than if you provide software ready to use even the electrician will use it as standard domotic equipment (usually very expensive).

[pmayer]

Hey sovking,

that was the plan. First step: provide a reliable, well looking board with casing for the home automation enthusiast. Maybe, after that, provide professional solution for electricians and so on...

Right now the project is stopped from our side as there are alternatives (posted above) in the making.
Maybe we'll pick it up again.

Regarding the Olimex boards, you are right. Not for sensors though. But my complete test equipment are Olimex MOD-WIFI-ESP8266-DEV's. They have a good quality ^^

Cheers,
Patrik

[data]

How many of these devices do you operate? Do you know how many individual codes they support?

A simple Chinese sensor: measures temperature and relative humidity, small LCD display, battery operated, transmitting the values into RF:ink/Domoticz. Can be ordered for $8.64, including shipping to EU. http://www.banggood.com/433MHz-Wireless ... ds=myorder
...
I have several in my home, quality is good enough for me.

[manjh]

How many of these devices do you operate? Do you know how many individual codes they support?

A simple Chinese sensor: measures temperature and relative humidity, small LCD display, battery operated, transmitting the values into RF:ink/Domoticz. Can be ordered for $8.64, including shipping to EU. http://www.banggood.com/433MHz-Wireless ... ds=myorder
...
I have several in my home, quality is good enough for me.

Currently I have two operational. The third arrived broken, so is now on replace-order.
The device has a sliding switch marked 1-2-3, so I assume three devices should be possible, but I forgot how I set the switches on the first two...

[data]

Ok, thank you for the details. Good to know I can use only three of these at once.

[manjh]

Ok, thank you for the details. Good to know I can use only three of these at once.

Not sure about that. I think I can confirm that you can use at least three.
But I have had PIR sensors with a similar 3-position switch, even when two sensors were set to the same number they still produced different IDs...

[data]

Sounds good. Are you using RFLink to receive these? If so, which receiver module do you have installed?

[pmayer]

And here is something what provides the needs for sensors completely: https://www.itead.cc/sonoff-sc.html
As far as I know there is no custom firmware yet and without the battery option.

[LulaNord]

Hi everyone...i am a new user here. As per my knowledge in terms of pricing for such devices, you can look at ITEAD and see how the "competitive landscape" is. However I'm sure German Engineering will add true CE marking as well as better quality in terms of look & feel which rightfully can add some €€ to the price. Looking forward to see where this can take you.

turnkey pcb

[pmayer]

So, after a long time: Hi again

As we've needed a sensor board for a customer, we've built a prototype which - I think - could be a very good ESPEasy board.

The board has a SI7021 Temperature/Humidity sensor on board, wired to i2c. With ESPEasy in mind we've also added a proto area and tried to make every I/O accesible.
We've ditched the Li-Po charging circuit as we've thought that you won't take the case of the wall to charge the battery and it is simpler to replace the battery. You can use the JST 2.0mm jack to wire a 1-cell Li-Ion/Li-Po pack. GPIO16 is wired to reset to make deep-sleep possible.
Using DC Power with a power adaptor is possible via the integrated LDO (5-12V). The circuit is fused and protected against voltage reversal. For programming there's a reset and a flash button and of course header pins to connect an FTDI cable. VCC/CTS/DTR for FTDI are not connected.

The case is wall-mounted, 86x86mm.

Here are som pictures:

IMG_5426.JPG

IMG_5427.JPG

case 86x86mm

I am keen for your feedback... and please let me know if there is interrest of buying this. Sadly the initially targetet price could not be reached. In a small series we could be somewhere arround 65-70€ per board including the case.

Cheers,
Patrik

[LisaM]

Way to big for me, this size i can make myself.

What i'm interested in is a esp12e switching mains in this format:

micro module.jpg

30 euro is max pricepoint for me, more and i'll buy another z-wave module...

[mrwee]

Good functionality, nice looking enclosure, but quite a steep price. I think it is too high