Before [Steve] realized that it didn’t play nice with his network, he dismantled his Energy Detective TED 5000-G to see what made the device tick. He put together a nice teardown with high-res pictures throughout. Each component of the TED 5000-G is dissected, with the exception of the current transformers, which he claims are pretty boring anyhow. The gateway module is particularly interesting as it contains both an Ethernet interface as well as a 802.15.4 radio for wireless communications. While the device is still a bit expensive at the moment, the gateway module could be useful in projects requiring PLC or ZigBee communications some time down the road, once prices ease a little.
Is it really that hard to set up a separate subnet on your network? Sounds kinda like throwing away a car because it needs an oil change to me.
Joel: It’s not a subnet issue, it’s an HTTP request size issue which generally manifests itself when accessing the device via DNS.
“Wireshark showing the TED Gateway responding with a 414 Request URI Too Long when a cookie is sent to the device with an HTTP request. ”
overclock it? ;)
or try upgrading the firmware before sending it back?
http://www.theenergydetective.com/ted-5000/firmware-upgrade
both the forum posts mention having gateway v350 and footprints v185, latest is v383 and v210 respectively.
I had a strong suspicion these things where hugely over-priced. Now I’m sure. Steer-clear folks.
dmang: The firmware I have is already higher than those version numbers, so that won’t help. The tech support person I called also informed me that he doesn’t know if a fix for this problem is coming or not.
Drone: I think that if it worked as it should then it’d be just fine and worth the price. The web UI was the only real problem that I had with it.
Hmmm – have you tried replacing it the gateway with a really small shell script ? :D
The gateway could be a implemented as zigbee/usb hack and daemon on your open-wrt router. (And what self respecting geek doesn’t run open-wrt or dd-wrt on a router somewhere on their network?)
More detailed info coming soon but if your getting a 414 shorten the request. Flooding it will not get you anywhere, it’s not running linux as you may have already guessed. It’s a proprietary OS. The Zigbee module should be a lot easier to interface with. As you see the programming header is there so just wire in. But unless you write your own firmware PLC will be out of reach..sorry.
duane: that is pretty much what we did w/ our demo site (demo.theenergydetective.com). We use simple bash scripts calling wget to pull the data from an installed TED and cache it on the server. We then have an http server pointing to a directly w/ all of the cached files. We did it to handle the large amount of traffic going to the device (its not designed to handle hundreds of simultaneous connections). I’ll see if I can package that up and distribute it to anyone that’s interested. If anyone wants to roll their own, the data calls are documented in our 3rd party api documentation available on the website. I’ll also comment on Steve’s blog w/ comments on the breakdown and cookie issue.
Also as a comment: CT’s are boring agree just a coil so nothing fancy there.
But concerning moving away from PLC in favor of ZigBee, we would have a hard time getting 2.4ghz wireless signal out of a metal box (panel). But don’t give up on PLC yet :) it may yet surprise you.
Ye, It would be interesting to see it on a WRT, I think I have one laying around I can cannibalize to play.
I am ready to give up on PLC. I just had my house re-wired, new panel, new service, new breakers, etc. and a dedicated circuit with the MTU sending the signal only on that circuit, and the gateway on that circuit, and the gateway still gets poor readings. If its not PLC, then it’s TED’s faulty hardware. The TED people still insist its line noise in my wiring.
I had the exact same problem and did the exact same thing you did. Put the MTUs and ECC on their own dedicated circuit. I had trouble like you described, in 2017-2018 no less and they were sure it had never happened before had to be me. (I just lost a lot of respect for these guys reading your account from 2011) I even went so far as to get an X-10 signal isolator on their recommendation and put that between the breaker and the MTU+ECC junction box. still unreliable. Disappointing.
We just installed a TED5000 on a system in Kauai. Its running on a off-grid system, all solar with Outback Inverters. We have almost 200 Insteon Dual Band switches at this facility and everything works great. The TED5000 however was really picky about noise and we had to install an outlet off of the breaker box to get the gateway as close as possible to MTU. The MTU disconnects stopped and it works great now. I believe our 120VAC is pretty clean to start with but the Insteon provides a strong signal that doesnt play well with Ted. Even when it works, the footprints web interface is pretty lame, but the data is useful.
I can tell you from experience that the TED 5000 will not always work.
Yep, I have a large Insteon installation here and nothing I could do would make the TED play well with it… I’m not a big proponent of PLC! I went with the Green Eye Monitor and am very happy with it. http://www.greeneyemonitor.com/
I’ve been using TED systems for 5+ years at two different homes, one of which was filled with x10 and Insteon. For all its faults (and like everything, it has some) not only have they been ahead of the curve on these devices they have also remained in the business – more than can be said of the majority of companies that have tried to jump in to energy monitoring. I have found their devices to be very accurate and much more sensitive than some of the competing products which sure was nice to track down vampires before the advent of KillaWatt type devices.
Yes, PLC can suck. On the other hand wireless ain’t exactly at plug-and-play reliability in anything but “standard” environments either and there’s a whole lot more development $$ going in to that market. In the same homes where I’ve run TED I’ve had to run 2-3 devices just to get so-so wireless coverage. In my mind TED and PLC are far from perfect but both remain very viable when compared to other options.
TED and insteon user here. The combination of both is finicky. I went the route of a dedicated outlet box and breaker for the TED equipment. Power goes from the breaker out to the outlet box, and then back into the panel to power the MTU units. That way the ECC and MTU share the same breaker and are as close to eachother as possible.
I bought an x10 signal isolator with the intention of using it between the breaker and the outlet box. I did install it there at one point but ended up taking it out and it’s still working most the time so heck if I know. If I could do it over again I’d spend the little extra it is for TED’s ethernet-based ‘commercial’ solution, and be done with PLC.
The less things using RF (even if it’s over power line) and more things using IP, the happier I am generally.