Insteon Review

I need to correct something I posted a few weeks ago _before_ I had some hands on experience with Insteon.

It is possible to set an Insteon lamp dimmer module to respond both to Insteon _AND_ X-10. This wasn't clear from the online documention.

This means the lamp can be controlled by the Insteon tabletop controller (in Insteon mode) as well as by a palmpad + transceiver (or any other X-10 controller).

For those interested in the Insteon PLC signal, I've published some scope screenschots at

Thanks for the scope shots Dave.

I received the Insteon starter kit too. Have a few Smarthome and X10 products in the house including their KeyPadLinc, and active booster/coupler that plugs into the dryer outlet. The later works in compliment with the Insteon RF repeater/coupler which is active ONLY on Insteon. The X10 booster on the dryer plug must remain for repeating X10 signals.

The tabletop controller has a neat feature where it blinks the white LED very brightly when receiving X-10 or Insteon confirmation of a command. It does not blink if X10 activity is initiated elsewhere.

The response time is much better than X10.

I'm waiting for the new Insteon enabled PowerLinc and Keypadlinc which will make better use of the Insteon 2 way functions and inherent reliability. The present suite of Insteon offerings is still basic. Nice, but no big deal.

RF Dude

I can't agree about the response time. I've read where Insteon claims a 0.2 seconds response time vs. 0.6 seconds for X-10 but I really cannot see much difference between pushing a button on the tabletop controller and pushing a button on something like my mini-timer. Most X-10 modules will respond to the first copy of the X-10 code which takes 11/60 (0.183) seconds to traverse the powerline. IOW, I think most of the speed claims made by Insteon (and UPB) are bogus.

I find RF range to be abysmal (~20 feet). That's not a problem where only the two SignaLinc RF modules are used to bridge the phases since most people will be able to find opposite phase outlets that are within range but it may mean you'll need a lot of the RF modules once they have RF remotes (which I assume are coming).

It's disappointing that they didn't design the RF units to act as an X-10 bridge but I can understand why - it would really complicate things. The PLC protocols are just too different and the RF frequencies are also different.

Still, in my limited testing, I'm impressed. The interoperability with X-10 is a big plus although the fact that the more Insteon units one has (all are two-way) the more the X-10 signal will be attenuated, tends to negate that.

Given the terms of the Insteon SDK license, I did not buy it. I would have liked to use the PC interface to create some more rigorous tests but the license would prevent me from publishing any results. :(

So far, the reliability has been 100% with no missed signals. Interestingly, things like the CM11A, CM15A and RR501 will sense collisions with the Insteon PLC signal. I will try to devise a way to test how Insteon fares in a noisy environment.

Dave:

You can try to do various tests, see how fast Insteon compares with X10 over a bridge, etc. But from a purely subjective perspective, Insteon seems a lot faster than X10. Under ideal conditions, I can press a button on my palm pad and have the outdoor motion on 0.6 seconds later. Realistically, some devices take longer if they are on the other 120V leg of the house and have to wait for the repeater to do its thing.... or not work at all if there is interference. For instance, it took three ATX power supplies on my old computer before I found one that didn't cause interference. How about coming home and finding a light turned on for no apparent reason... in the middle of the day. So I got into the 2-way stuff. I noticed that it takes my KeyPadlinc almost 1 second to turn on a Switchlinc Relay, the LED gets the status back about 2 seconds from the initial key press and keeps flashing for 3.5 seconds after key press until the active repeater finishes doing its thing. Thats a lot of slow data moving around on my power lines.

I programmed one key on the Insteon Tabletop controller to send both an Insteon command and an X10 command. Why? A larger room with table and wall lamps all controlled together. Really neat! Press the button to turn on the lights and the Insteon LightLinc is very quick (like you say, 0.2 seconds) followed by the X10 wall switch/lamp shortly thereafter.

Insteon will likely shine when applied to macro's or automation. While on a one for one command, the difference isn't big, delay would grow significant with X10 if trying to do a series of commands in rapid succession. Now add

2-way and it is no contest in favour of Insteon.

I opened up one of the units. Quality looks an order of magnitude better than X10.

So I have expectations for a Insteon version of the PowerLinc, and KeyPadLinc followed later by other useful modules. When more gadgets are available, the power of Insteon will be much more useful. Until the new products come, I agree that the differences aren't too exciting.

Keep testing and let us know what you find.

RF Dude

I had also made that same setup but I can see no difference in the time it takes the lamps to turn on and off. That may be because the Insteon lamp is on the opposite phase. I'll try later with both on the same phase.

I'm sorry but there is no way to make that comparison by just looking at the insides of the module. All of the SmartHome devices are of more recent design than are the X-10 devices but the fact that they use SMD components and look prettier doesn't necessarily translate into better. I bought a very early PowerLinc - it had design flaws. I bought one of the first two-way LampLinc modules - it had design flaws.

The SDK already includes a PowerLinc.

I'm impressed with the concept and, so far, with the execution but am still leery because SmartHome has such a poor track record. I also do not have problems with X-10 but do understand that others would probably prefer something that is more reliable (in their view) and easier to install so I hope that Insteon will succeed. The prices are attractive.

X-10 requires an address message (2 copies) and a command message (2 copies) to be sent totaling 44 cycles. The soonest an x-10 module will respond will be the first command message which is 33/60 of a second (0.55 seconds). The soonest INSTEON will respond is after the first complete message which is 5 zero crossings or 5/120 of a second (0.042 seconds). 0.2 seconds is actually a very conservative estimate. If the INSTEON device doesn't hear the message until it has been relayed 3 times (the maximum number of 'hops'), it will respond in 0.2 seconds. To truly evaluate the response times of the system, make sure the fade rate is set as fast as possible. As a dimmer fades up, the lamp is usually not visible until the level reaches 15 or 20 percent. A slow dim rate would add a significant delay to the perceived response. Better yet, use appliance modules; you can hear the click of the relays and instantly determine the response.

The license terms of the SDK does not prevent one from publishing their results, findings, or efforts with the SDK kit. Try it, use it, talk about it.

The only restrictions are on the publishing for sale software developed with the aid of the kit. Smarthome has made a short FAQ about this subject....

Will I be able to sell the product(s) I develop with the developer's kit?

Yes, you will be able to sell your software product(s) developed from the INSTEON software Developers kit. Prior to offering for sale, you must submit the product to Smarthome for INSTEON compatibility testing and INSTEON certification. This is a free sevice that Smarthome provide to insure the software complies with proper Instoen protocals. Once you are at a stage where the product can be tested contact Bob Cusey (bobs email goes here).

How do I sell my software or hardware I created using the developer's kits?

Once the product passes compatibility testing, you sell the software and/or hardware through any distribution channels. Additionally, the product may also be submitted to Smarthome category managers for consideration to distribute through Smarthome web and catalog channels.

I moved the lamp module to the same phase as the controller and made sure the module was set for the fastest ramp rate. I still see no big difference between the time that an Insteon lamp turns on from the tabletop controller and an X-10 lamp turns on from a mini-timer button. Insteon is marginally faster but it's really not all that obvious to an uncalibrated eyeball.

I have explored a bit more. If the Insteon controller does not receive an ACK from the Insteon module, it will resend the command several times. The documentation says up to five and it looks like it's about that long using the ESM1. I don't know whether the number of retries is a user setting. The LED on the controller flashes rapidly for several seconds after such a missed command. I tested across phases by unplugging the target module. The result gives a rough indication of PLC signal reliability and I think the PLC signal reliability will prove to be very good since it's two-way with ACK/NAK. Only experience will tell whether the hardware/firmware reliability will be as good.

BTW, it's possible to disable the controller's "beeps" and even the LEDs on the modules. It's amazing what one can learn from reading the manual. ;)

I have one of the RF links near a CRT type monitor. It's LED flickers faintly constantly. The LEDs on other RF module and lamp module on that same phase are steady except for occasional brief flickers off (once or twice per minute). This reminds me of the flaw in the X-10 protocol LampLinc two-way module where the LED would flicker at random and the module would go deaf to the powerline.

The LED on the other lamp module is rock steady except it will also flicker off briefly. When it does, I see a brief flash of the ESM1. It's shorter than the normal Insteon messages. I do not know whether its noise or some brief housekeeping signal from the module or controller. Tomorrow I'll move the ESM1 to the other phase to see if there's similar activity when the LED flickers.

Troubleshooting when problems do arise may be difficult. The ESM1 will indicate PLC signal presence but when it's absent, finding the cause may require dedicated test gear to measure/analyze PLC and RF signals.

The hell it doesn't.

1. CONFIDENTIALITY. The Kit and all related information are confidential and proprietary to SMARTHOME. You agree not to disclose any information relating to the Kit (including without limitation the results of use or testing) to any third party without SMARTHOME?s prior written permission. Unauthorized use or disclosure of such information would cause irreparable harm and significant injury to SMARTHOME that would be difficult to ascertain or quantify; accordingly you agree that SMARTHOME shall have the right to obtain injunctive or other equitable relief to enforce the terms of this Agreement without limiting any other rights or remedies.

The above is a direct quote from the license and would appear to be legally binding whereas the postings here or elsewhere of some anonymous person who claims to represent SmartHome and who may or may not be authorized to speak for SmartHome on this topic are not.

If you really mean it, change the terms of the license to allow dissemination of test results.

Who, in their right mind, would spend weeks or months developing and testing application software or Insteon hardware when you will have the right to prohibit them from selling it?

I would assume that since you've designed the hardware and API that it would be nearly impossible to create anything that was incompatible.

So far, I like the Insteon technology. It appears to work as advertised. But I won't buy the SDK or waste my time developing anything under the terms of your license. I doubt that anyone who consults their attorney will do so either.

It would be interesting to learn how Insteon performs in a purely X10 environment. Perhaps they have a better X10 implementation than many of the existing X10 products, and would be attractive even for those with primarily X10 controllers.

I swapped locations for the two RF modules. The one with the rapidly flickering LED is no longer flickering. The one that was stable is now flickering. So there is some environmental cause. I suspect some noise source on the phase with the flickering LED. I'll try to isolate the source later. It may be the pulses from an unknown source I've shown in the screenshots at

which although present on the other phase are very low amplitude or it might be noise from an ATX power supply that's on the phase where the rapid flickring occurs.

Interestingly, while testing that all modules still worked, I saw my first "missed" command. The module on the phase opposite the controller took a few seconds to respond to an OFF command. The LED on the controller flashed on/off for several seconds as it does when the target module is unplugged. Note that the module did respond after a longer than normal delay - it apparently sent a NAK (or failed to send an ACK) and the controller tried again. I am more and more impressed with the Insteon technology.

I also moved the ESM1 to the other phase. It shows brief activity when the LED blinks off periodically. The time between these events varies from 10 to

30 seconds. The powerline activity is not present if the RF unit is unplugged so I speculate that it's either some Insteon housekeeping message or maybe some other 900MHz device in the neighborhood. In a mixed environment it presents a potential for collisions with X-10 PLC signals.

I isolated the cause of the rapidly flickering LED to the PC that's plugged into the same circuit as the RF unit. When the PC is off, the flickering stops.

Except that the RF units intended as phase couplers do not repeat X-10 nor do the individual Insteon modules. I don't know the capabilities of the Insteon PowerLinc unit (now available only with their SDK) - it may or may not send/receive X-10.

Historically (e.g. SwitchLinc, PowerLinc, LampLinc, BoosterLinc), SmartHome has not done a good job with X-10. I doubt that will change in their Insteon devices. After all they want to switch the market to Insteon. I think Insteon will be able to coexist with a limited number of X-10 devices but the more Insteon devices installed, the more trouble you will have with any remaining X-10 devices.

Insteon devices, being two-way, attenuate the X-10 signal so you may need their BoosterLinc to compensate. You will need an X-10 phase coupler.

The Insteon specs and concept look very good. My trepidation is solely because it comes from SmartHome and I have little regard for their engineering prowess as demonstrated for the past several years. I suspect that Insteon came from different roots.

I still want to try to generate some noise on the powerline to see how Insteon handles that but so far, in my rather limited tests, Insteon looks very good. Those who have troubles with X-10 might want to consider it. My X-10 set up is reliable so I have no reason to switch. If I were starting anew, I'd look hard at Insteon.

Now, if they were smart, they'd create a 418MHz RF to Insteon bridge and publish the protocol so that Pronto owners could control it wirelessly. I don't think they're smart. :(

Does filtering the PC with an X-10 filter help at all? I assume that if the transmissions show up as noise on the ESM-1 that an X-10 filter might have some effect.

-- Bobby G.

My enthusiasm is waning.

I have now seen a few complete failures - cases where the target module never responded and the controller LED flashed for several seconds. There have been several other cases where the target responded but only after 3-4 seconds.

The ESM1 was plugged into the target circuit and it showed nothing during the failures which was an indication that the RF link wasn't working. It turned out that I had repositioned one unit (to avoid the bright LED which really is annoying when looked at head-on) and had the antennas oriented badly but this is with the RF units only 8' apart. I think this will be tough to troubleshoot without dedicated diagnostic tools to indicate both RF and PLC activity. In fairness, in a permanent installation the antenna orientation is unlikely to change - but I suggest the RF units be within a few feet of each other. And, it may well be that they have diagnostics in the firmware that can only be triggered by a PC interface.

My test setup is less than ideal as the only way I have to send a signal is by pressing buttons on the tabletop controller. I could devise more rigorous and automated tests with a PC interface to send/receive and with a 900MHz RF receiver to indicate whether a failure was related to RF or PLC. But then I would be the only one to ever know the results given their license terms. :(

Over the week-end I will try to come up with a way to generate some noise on the powerline and see how the system handles that. Then, I'll be returning the kit to the dealer who loaned it to me.

It doesn't show up on the ESM1 nor can I see anything on the scope. That may be because it's outside the passband of the ESM1 and of the ACT Scope-Test2.

Anyway, I only have one X-10 filter and it's on my VCR. I'd have to reprogram it if I stole the filter. ;)

I haven't found a way to inject high level noise on the line. A low level (~45mV) 131.7kHz CW signal does not hinder Insteon operation but a higher level might.

All in all I'm impressed with Insteon and think it is worth consideration with two caveats.

One: The weakest point is the 900MHz RF links which serve as a phase bridge. They need to be fairly close together and are subject to interference from other 900MHz devices.

Two: SmartHome has a poor track record so other flaws may appear under more intensive use.

The two-way protocl is reliable (subject to caveat one) and reliability should get better as modules are added since each module acts as a PLC repeater.

Signal suckers will still require filters. Noise sources may not be as much of a problem as they are for X-10.

The initial price point is competitive with X-10. If that remains true as they expand the number of modules and devices I think it has a chance to displace X-10.

We're having the same issues with this license as Dave Houston, in particular the actual license is much more restrictive than the quoted paragraph:

This is saying that applications to be sold must be according to the terms of an agreement that Smarthome writes, otherwise you cannot sell them with the Insteon label, or maybe at all. So what's to prevent Smarthome from charging a substantial license fee for every copy sold? This is really completely open ended.

We have no problem with a compatibility suite test, but aren't quite ready to sign a blank check.