Today, October 31st is, at least in the United States, the day that the doorbell rings the most. I am happy to say that iDONT was active and ready for the test.  While we did not have as many doorbell pressers as normal (it was literally freezing here), we had over 10 presses by kids dressed in everything from lady bugs to giraffes and every one was sent to my phone. Great job iDONT!  

The main motivation of the iDONT is not allowing the doorbell to wake my family from sleep. Yesterday, we were blessed with a new addition to our family, as my wife gave birth to a girl we named Truly. I am looking forward to using iDONT now that she is here.

All said, it has been a while since I have had an update.  There has been a lot of other things going on (the house has never looked better), but I have been able to get some time in on iDONT.  Over the next few days, I will share the final prototype and how it works.  Until then, I am going to enjoy holding my daughter.

One of the important details in making a product for the consumer market is the look.  Without spending a lot of time and or money, premade enclosures are a good option to get close.

For this project, I am planning on starting with this enclosure:

In addition to just a box, Digikey can customize itwith holes and add things like silk screen.  Nice.

With over 12 years of designing electronics for manufacturing, I know I can build the hardware. On the embedded software size, Joel has 13 years of experience including the critical security training.  Together, we can take iDONT from concept to production ready. Here is how the business will run:

1. Design a prototype circuit board. This means that the ESP chip is not soldered directly to the board, but “connectorized” to an ESP module board with headers. This allows for people with minimal soldering experience to build a working iDONT.
2. Build, test, and debug the prototype iDONT using the prototype circuit board. This is both hardware and software. Shipping a robust working product from day one is key to success.
3. Cerate detailed instructions.
4. Sell 20 iDONTs on Tindie. Some would be fully assembled, programmed, and ready to go. Some would be a kit. And, finally a few would be just the circuit board. If the sale of the 20 units goes within two months, I believe that I have enough market/interest to…
5. Create a production circuit board for manufacturing iDONT. This circuit board would have as few through-hole/hand soldered connections as possible to reduce cost. Additionally, using a SMT ESP module that has been FCC certified is about half the cost of a dev module that the prototype would use.
6. Repeat steps two and three.
7. Pursue Crowd Supply to get even bigger discounts by buying in quantity. Selling on Adafruit, Seeedstudio, Sparkfun, or Tindie (with the help of a contract manufacture) would be another option if Crowd Supply did not work out.  I want to help others with their doorbell woes, but need support to do so.  All these avenues offer so much behind the scenes help: setting up payment, shipping, making secure webpages, asking detailed questions that lead to better planning and ultimately success.

I asked a buddy Joel, who is an experienced embedded programmer, to help with the project.  He came over this weekend and helped me get a captive portal working!  This is a huge step in making iDONT a consumer product rather than requiring programming the ESP module for a person's Wi-Fi credentials (SSID and password)  and upload the code.  With a captive portal, the iDONT broadcast it's own SSID and once you connect with your computer or smartphone, the iDONT asks for the credentials and flashes itself.  Great work tzapu and Joel!

That said, this only works on an ESP8266, not on a ESP32.  Joel thought that with enough work, we could get it going on an ESP32, but given our requirements are small, everything will run just fine on a ESP8266.  As a bonus, the ESP8266 is also less expensive.  So, time to make a change.  

I was honored to have a Hackaday front-page article of iDONT! Much better description of the project than I could ever come up with...

I have been playing with a Non-Invasive Current Sensor that works with AC currents. Along with the ESP's built-in ADC and a conditioner circuit described here, doorbell presses when the doorbell is on/active can be detected. This would be good for logging and maybe additional smart home features which are beyond the scope of this project.

A week ago, I was one of the winners of the Internet of Useful Things Hackaday 2017 Contest! It is an honor to be selected. I was even written up in the in a HACKADAY.IO BITS....

We've found useful Internet of Things things

By Brian Benchoff

To our shock and amazement, the Internet of Things can be useful, and it's more than just tweeting toasters and Internet-connected spy cameras. Check out the winners of the Internet of Useful Things phase of the Hackaday Prize which we announced yesterday.

Just look at entries like the iDONT -- the Internet Doorbell On/Off Trigger. This is a project to connect your doorbell to the Internet, giving you the ability to turn the doorbell off for meals and naps, and send doorbell push notifications to your phone. [David Spinded] created this device with a relay and an ESP32 microcontroller. It's so simple and so brilliant we're shocked this isn't a commercial product yet.

---

Thanks Brian for describing my project better than I did!

Enough celebrating....get back to work....

One thing that I have been struggling with is what relay to use. I have been running into an issue with the relays freezing up and I think I know why now. I have been using two relays (one for the line and one for the neutral) and controlling them with two different microcontroller pins. While in the code, the command to change the IO state happens one after another, it is not instant. I am thinking now that the relays were freezing because half of the doorbell was on AC and half on DC for long enough that the load side of the relays were seeing some "crazy" things and not switching.

By using the same microcontroller IO the relays will be switched at the same time. To address the 12mA drive limit of the ESP32 a transistor will be used off the IO to switch the relays.

So, I believe now that I can use relays that are inexpensive and widely available, for example the Songle SRD-05VDC-SL-C. These relays are in so many relay modules that are under $5 on ebay, or in piece parts for under $0.25 with free shipping. This certainly beats paying ~$10 for a high power SSR. While I have done a bit of testing of this, I need to completely prove this out.

Until next time, happy hacking.

My wonderful wife gave me some extra time this weekend to work on iDONT. I was able to put all the modules together and now it is time for documentation. After being awakened by a doorbell last week myself, I have some good motivation to get this project closer to the finish line. Here is a video of where the project is at:

This block diagram gives a good overview.:

One big thing to note is the constant current DC converter. I plan on setting the current limit low so that when the iDONT is set to turn off the doorbell, the low voltage converter is powering the doorbell. When the doorbell is pressed, the resistance changes and we can monitor if the doorbell has been pressed.

The ESP32 software is coming along, but not ready for prime time. Here is a screenshot of where I am at:

Here is the link to the code for the above software: GitHub. I have been able to get time as well in order to set a auto turn back on function, but still it needs work.

Here is my scanned in log book over the past few weeks:

Until next time, happy hacking!

After lots happening in my personal life, I am back with an update on the iDONT: it works!

The basic function of a doorbell disconnect using a relay that can be controlled via a web browser. The page is simple enough that it works fine on a smartphone.

The code is located here: https://github.com/SpinInven/iDONT/blob/master/ESP32_iDONT_basic_webserver.ino

I am shocked how easy it was to develop the software. There is a growing number of examples and tutorials and I hope the above example helps someone else. I am pleased with the progress with everything except for the security. It is just whatever comes with the basic Arduino wifi example...less than ideal. I don't fully understand what the webserver is doing. This means that I have a long road head for security...no wonder so many products are lacking security, deadlines don't seem to allow for such dives. While I don't mind if people know the status of my doorbell, it is something I plan on addressing.