Iris V2.0 Hub Teardown
1 1

21 posts in this topic

Here is my V2.0 tear down.

 

Frist off you may ask why I did it. 

It's simple I wanted to know what makes it run and I know every else wanted that too.

 

To get into the hub is super easy, just get a small piece of plastic like a gift card. Work into the joint of the top and bottom of the case. Here is what is inside.

8A3zSKw.jpg

 

 

For everyone asking for local processing I don't think it will ever happen on 4GB of memory.

What is the Bluetooth for ? I would love to know why this is on the board.

Why use AA for something needing so much power ?

pavalov and loonytim like this

Share this post


Link to post
Share on other sites

I don't see 4GB of memory to be a bottleneck either.  This is just a home automation hub...  Many PC's still today have just 4 GB of memory and have far more complex calculations.  

scunny likes this

Share this post


Link to post
Share on other sites

4GB RAM and a 1ghz arm CPU seem perfectly suitable to handle things like local processing.
Assuming that "4GB memory" is actually RAM and not Flash memory (wasn't specified), thats considerably more RAM than most modern smartphones.
This actually seems like a pretty capable machine... but I do wonder how much actual non-volatile memory there is.

Compare this against the specs of the various raspberry pi models. Then look at all the things people are doing with the pi and it will give you an idea of how powerful this COULD be.

scunny, sparc and loonytim like this

Share this post


Link to post
Share on other sites

Okay, guys I am sorry about the confusion on my post. I am talking about flash memory(embedded memory) of 4gb. Because if the OS (Linux) uses about 600mb to 1gb then about 1gb of software. Then you leave 1gb for downloads like updates. You only have 1gb for databases and I can see that getting eaten up fast.

Share this post


Link to post
Share on other sites

There are some linux distributions that weigh in at 50MB or less such as puppy linux or DSL (about 12MB).
I'm sure they are using a stripped down distribution. Updates are not likely to ever come anywhere close to 1GB.
Back in the day when everything was CLI, no one even dreamed of 1GB hard drives. When a GUI is not needed, 4GB goes a very long way!

scunny likes this

Share this post


Link to post
Share on other sites

There's no question the box itself could do local processing.  I agree with what loonytim stated: It's a massive software overhaul on both hub-client side and on the back-end.  I'm sure it's a paradigm change in the software architecture in terms of server-client communications and synchronizations.

 

Is that a pin-header in picture dead-center?

 

Thanks for the tear-down pic.! :)

Share this post


Link to post
Share on other sites

dead center - resistor pack (you can kinda see the Rxx markings). But above CPU is what it looks like a serial console port (like an old hub had). I've seen forum posts where people were able to get a console output for the old hub. I don't think it's worth hacking though, if you are that level - you don't need Iris :)

 

But overall - pretty good components, next gen Z Wave chip (500 series), decent CPU and RAM. Just need a good software to go with it :)

Share this post


Link to post
Share on other sites

I'm betting they have it figured out but nonetheless way too delayed.

 

It really has been a while....at this point hopefully they're just working out the bugs and not just moving into the planning / implementation stage.  :huh:

 

 

dead center - resistor pack (you can kinda see the Rxx markings). But above CPU is what it looks like a serial console port (like an old hub had). I've seen forum posts where people were able to get a console output for the old hub. I don't think it's worth hacking though, if you are that level - you don't need Iris :)

 

But overall - pretty good components, next gen Z Wave chip (500 series), decent CPU and RAM. Just need a good software to go with it :)

 

Ack...duh.  I was looking at the standard image - can clearly see it when you click and blow the image up!

 

Agreed!  Pretty decent hardware all-around.  The developers really need to do some magic.

Share this post


Link to post
Share on other sites

"For everyone asking for local processing I don't think it will ever happen on 4GB of memory."

Why not? We went to the Moon and back with a lot less than that running on a 2 MHz chip!

https://en.wikipedia.org/wiki/Apollo_Guidance_Computer

"There are some linux distributions that weigh in at 50MB or less "

There used to be a stripped down version of Linux that could run from a 1.44MB floppy.

Share this post


Link to post
Share on other sites

dead center - resistor pack (you can kinda see the Rxx markings). But above CPU is what it looks like a serial console port (like an old hub had). I've seen forum posts where people were able to get a console output for the old hub. I don't think it's worth hacking though, if you are that level - you don't need Iris :)

 

But overall - pretty good components, next gen Z Wave chip (500 series), decent CPU and RAM. Just need a good software to go with it :)

 

I tried to hook up to the serial ports and found out that Rx is pin 2 (from the left) and Tx is pin 3 and Gnd is pin 4 but ive been trying every single combo of baud rates and nothings working, just getting loads of gibberish at all of the typical (9600, 115200, 256000 etc.) any ideas what it could be?

Share this post


Link to post
Share on other sites

Okay, guys I am sorry about the confusion on my post. I am talking about flash memory(embedded memory) of 4gb. Because if the OS (Linux) uses about 600mb to 1gb then about 1gb of software. Then you leave 1gb for downloads like updates. You only have 1gb for databases and I can see that getting eaten up fast.

 

The first Iphone came in a 4gb and 8gb model.  4gb is way more than you are giving it credit for...600mb is way more than what linux would actually use... and what software would they need that would be 1gb?

Share this post


Link to post
Share on other sites

Recently, I've come across more embedded processor software development efforts that actually use full-blown versions of Linux instead of their stripped-down embedded-purpose cousins.  No doubt simply because modern embedded processors are much faster and includes more RAM and FLASH and the need to strip down the OS becomes less of a priority.  Of course, it all depends on the application.

 

And I'd be shocked if the IRIS software running atop the V2 hubs eats up close to 1GB alone, with or without local processing....

Share this post


Link to post
Share on other sites

Seems risky to have a full install when you only need certain pieces. Too many potential backdoors or other security leaks but in a development environment not so bad. If they keep it all in the production version not so good.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
1 1