Jump to content
Forum sending old emails Read more... ×
Sign in to follow this  
jdownj

Suggestions on MCU/Chip Selection

Recommended Posts

I am banging something together to provide a hardware solution to a service problem. Due to unreliable connections, my cable modem drops offline every 24-36 hours, requiring a reboot. I want to build a device that will:

 

-Connect via Ethernet.

 

-Ping a known server every 5 mins.

 

-If server is unreachable, pulse an output high, resetting the cable modem.

 

I am thinking G2553/Wiznet 5100, but if anyone has any ideas on how to lower the BOM cost or explanations on why I need to reconsider that, please let me know.

 

Currently I implement this functionality on my router(OpenWRT), but it is possible for the router to freeze as well. I think a hardware watchdog like this is the most reliable solution.

Share this post


Link to post
Share on other sites

I think Wiznet is the best way to go, but as you already know, modules are kind of pricey, ~$20 + s/h. Maybe we could start a project here to design a W5x00 based board (reverse engineering is another option.) Wiz chips are ~$3-$5, other parts <$5, I think we could build a web server board for under $10. BTW, I need one for my home automation project.

Share this post


Link to post
Share on other sites

Well, your router should have a hardware watchdog already (if not, time to get something made in this millenium :)) and at the very least some unused GPIO ports as well.

 

In this particular case, unless you have some other prohibiting factors (distance between cable modem and router or whatnot), I think building a separate device is the least practical thing to do (not counting the fun factor, of course ;)).

Share this post


Link to post
Share on other sites

I have been following the wifi booster project. I don't think much from that will apply, except for code. I am shooting for a total BOM less than that wifi module. I am not a coder, but I'm doing my best to learn. My plan for this board is:

 

MCU/Ethernet Chip

802.3af PoE

2-3 GPIOs with transistor to drive relay.

Other GPIOS available on header.

 

I am aiming for a total BOM less than $20. If others are interested, I am certainly open to making kits or selling bare boards. Fair warning, SMD doesn't scare me;). Wiz5100 is 80-LQFP, and I will likely go QFN on the MCU. Most PoE chips are also "effing tiny". Board space is expensive... I hope to beat both the size and BOM cost of the Dangerous Prototypes Web Platform, about the closest available device.

Share this post


Link to post
Share on other sites

@gordon,

I currently implement that on my router, but my issue is that my router runs many services, and some have their own reasons for locking up. OpenWRT is stable... CUPS/Samba/Asterisk not so much... Because I am away from home often, and I access stuff on my network via VPN, I am really chasing brute reliability here. It is annoying to have to wait for an acceptable hour to call the neighbor to reboot the router/modem.

Share this post


Link to post
Share on other sites
@gordon,

I currently implement that on my router, but my issue is that my router runs many services, and some have their own reasons for locking up. OpenWRT is stable... CUPS/Samba/Asterisk not so much... Because I am away from home often, and I access stuff on my network via VPN, I am really chasing brute reliability here. It is annoying to have to wait for an acceptable hour to call the neighbor to reboot the router/modem.

 

Turn a msp430 into a watchdog, combined with a script on the openwrt router. Router script pings server, if good, toggles gpio. This triggers an interrupt on the msp430, which resets its countdown timer. 3 missed resets, it reboots the router. No need for a ethernet/wifi connection on the msp430's part, and allows for it to be small and internal to the router, and powered from it too. Minimal coding required.

 

Alternatively, instead of just a gpio, you can use one of the serial ports on the router instead.

Share this post


Link to post
Share on other sites

Thanks!

 

Damn I hate it when a perfectly good project is shot down by superior thinking;)

 

I may continue the idea of an Ethernet dev board, but you guys have definitely provided a better solution to the problem at hand.

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
Sign in to follow this  

×