Final Goodbye

[Last_Mile]

Well, its been a 6 month or more trail with the Oak.  Sometimes it was an adventure in learning, other times it was just hell.

I started out buying three Oaks, some level changer shields and other misc parts on Kickstarter.  Now I’m free of the Oaks, as they have been de-Oakified.  They are all now deployed in my home automation system and are working fine, although non of the features that set Oak apart from the other development boards are being used.

From my point of view here are the Pros and Cons of the Oak.

Pros
1.   Small size.
2.   Lots of performance in a small footprint.
3.   A decent amount of memory.
4.   A decent number of ports for its size.
5.   Can be programmed using the Arduino IDE
6.   Over the air programming is fast when it works.
7.   On board voltage regulator.
8.   I2C, SPI, serial and analog ports available.
9.   Convenient micro USB power connection.
10.   The Oak voltage regulator is robust!  I was able to power the Oak AND an Arduino!  (The Arduino was NOT able to reliably power the Oak.  Go figure…)

Cons
1.   Over the air programming using Particle.  Works sometimes.
2.   Over the air programming using Particle.  One out of 10 or 20 times works.
3.     Over the air programming using Particle.  This sucks.
4.   Should have waited and not assembled all of my Oaks with the female headers.
5.   Pins out the bottom would have been better for some applications.
6.   Micro USB only supplies power.

Overall the Oak is a good compromise of features, size and the number of ports.  The Oak is head and shoulders above the ESP-01, but at $3 each they are hard to beat for very small projects.

If you want to go three times the size an ESP8266 NodeMCU is MUCH easier to develop on, program and deploy.  The USB connector supplies power and provides serial access.  The availability of a program and reset button makes it convenient to download using the Arduino IDE.  (Haven’t tried the Arduino auto reset mods yet.)  Costing up to $30 each these are definitely at a higher price point.

Spending a little more money was worth it for me.  If you want to, you can still use OTA, but the sheer increase in productivity was the tipping point for me.  Instead of one or two code, build, download cycle in 10 minutes, you can probably do a code, build, download cycle every 25 - 30 seconds!

So, all I can say is with Kickstarter there is always some risk.  Otherwise, you should just order off of Amazon, Ebay or Aliexpress and receive known quantities and capabilities.  The Oak was a good learning experience.  This was my first adventure away from the traditional Arduino form factors and into the ESP8266 territory.  Now, my eyes are open to a whole new ESP8266 universe.  I’m sure if we wait long enough there will be an Oak clone addressing all of the cons addressed above.  In the mean time there is a robust community of ESP8266 developers and hardware suppliers.

Goodbye and thank you for the opportunity to learn with the Oak.

[PeterF]

Sorry to see you go, and good luck with your projects!

I'd just like to add that if you want to keep the Oak's with Particle, but the OTA programming is screwing up, there is a serial upload option, where you can load your program over the serial. But to be honest, I'd only be going that way if you wanted the keep the Particle infrastructure/link. If not,  to be honest, I'd just put the ESP8266 Arduino OTA stuff on and be done with it... OTA updates with that are SO fast And the ESP8266 core over there has come a long way (the OTA stuff wasn't around when Erik started on the Oak, which is why it didn't leverage that at all), so might as well enjoy all those benefits.

[wickwire]

Hi,

First of all thank you for sharing your point of view. I found myself in a similar situation, however with the Particle Cloud instability issues and some other limitations on specific stuff I was doing (MQTT over SSL) I ended up trying out the ESP8266 Arduino solution on the Oak.

Pleased as I am with it - even the OTA update mechanism seems robust enough - how would you say the Oak fairs against an ESP8266-12 + IO Adapter Plate Expansion, considering the use of the ESP8266 Arduino firmware on both?

(see image in attach for a clearer view of what I mean by ESP8266-12 + IO Adapter Plate Expansion)

The ESP8266-12 doesn't have the micro-usb port for power, however it is significantly cheaper than the Oak. On the other hand, maybe the Oak has better electronics circuitry than the ESP8266-12 with the IO adapter plate - something I know nothing about since electronics is not my background.

I'm considering this comparison because comparing the ESP8266-12 alone against the equivalent Acorn by Digistump does little for development in terms of accessibility over pins and power, on this particular thread discussion.

All that said, my point is, if using the ESP8266 Arduino firmware, would the Oak still have advantages over the generic ESP8266-12?

As for the points you've mentioned, I'd just like to mention that to me:

4) in the cons section shouldn't actually count as it was your decision to do it I guess, whereas you could have just work with one module first and then broaden the setup to a few more modules later on, like some people do, me included.

By then perhaps you would have realized if soldering female headers was the way to go or not for whatever project you were working on, before doing it to all the modules.
The same should be said about OTA updating: regardless of the infrastructure used, update the production line only after lengthy and successful testing with the development units and do the production updates in a phased manner, splitting the production range into several groups.

As for point 5), I'm not sure I understood it correctly, but you can always solder the pins the other way around and even remove them later on, if you don't want to use them. Both pin headers and Oak or ESP modules should be re-usable, it just boils down to soldering skills and not the module architecture/design, in my opinion, having done that myself.
Again, should not be seen as a specific limitation on the Oak board itself.

[PeterF]

lol... yeah, I think 4/5 go together... and i'm also guilty of that... so one or two of my oaks have been relegated to never be used outside a specific project because of that It did make things awkward when I decided to make a bit of a test jig so I could easily serial program/flash/debug my oaks when fiddling with them... had to use stable headers so I had both female sockets on the top and male pins sticking out the bottom

[DrJFM]

I started to go the "Generic" ESP8266 route with my Oaks and must say, I was generally pleased with the results vs the Particle stuff.  I had no real problem converting my DHT22 Blynk app to the ESP8266 Oak format.  I had not really needed to get into the OTA aspects of the conversion -- I am comfortable with the serial adapter uploads, and I don't need remote updates.
The WiFi works fine with Blynk and I can view my sensor history using Blynk on my phone. HOWEVER, when faced with the prospect of really learning and getting into the ESP8266, I discovered the new big brother to the ESP8266, the ESP32.  This already has a fairly robust community of makers, hobbyists and professionals pumping out new and improved code, discovering its rich hardware attributes and moving to bluetooth as well as WiFi.  If I had to learn details of a new system, why not a really new system.  Priced like the Oak by mail from China (Not yet ESP8266 cheap, but not expensive), equipped with WiFi, bluetooth, AND a built in USB programming port (and more memory, speed, etc) whats not to like.

So I am moving to the ESP32 platform.  It uses onboard flash to store my SSID and password and recovers from power outages etc much more gracefully than the Oak ever did.  OTA is supported although I haven't needed it yet.  I moved my Blynk/DHT22 app and my Blynk IR control app to the platform readily.  Updates have been flowing (still fairly early days) but the Arduino based programming is stable and supports most hardware features (w more being added).

If Erik is serious about a new WiFi internet based dashboard system, he may want to see if the ESP32 would not be either a better alternative or a modern, second addition to any new Oak like efforts for IOT controls.  Espressiff makes both the ESP8266 and the ESP32 and the ESP32 specific forum has over 11,000 posts (a few more than the Oak part of this forum).

So my future will include some generic ESP8266Oak work (I own a few) but new stuff will be on the ESP32.  Just saying. 

[Z69]

I will second that...

I have 2 Oak's running alternative firmware - "LetsControlIt" and they have been running for a year or more with "Zero" crashes.
The Oak hardware is solid, it is the firmware that is very bad.

OTA or Serial code upload, the Oak firmware is bad.
This is a pity as i had high hopes for it.

[PeterF]

Yeah, the hardware is one of the best, if not the best, ESP8266 boards you can get, but the firmware, particularly the particle side of things has rendered it quite unstable. Having said that, I ran one as a simple temperature sensor with deepsleep, with manual wifi config, and didn't have any issues for a good six months. I'm actually having more trouble with another one that is running the current stable ESP8266 Arduino core, TFT display and thinkspeak connection... I'm certain the thingspeak library is a POS and keeps locking up on the ESP8266 every 2-3 weeks :-/ Just haven't had a chance to work out what my other options are for reading from thingspeak.