ELA

Michel, I have always felt that the ISY was a good quality piece of hardware/software. Thanks for sharing the statistics that prove that to be the case. To be fair to the PLM it does have the nasty job of having to directly interface with the potentially harsh 120VAC power distribution environment in order to accomplish its PLC communications. I would love to hear the failure statistics along with the root case analysis of PLM failures. If mine were not under warranty I had intended to further diagnose why it failed. While I did test and confirmed no PLC transceiver ability I am guessing it had even an even larger problem in that it would not Factory Reset either. Maybe we need redundant PLMs to backup the most excellent ISY stats I am somewhat dreading replacing my PLM. I read several threads on how to Replace a PLM so that did not seem to be much of a chore. The ISY appears to do a great job of handling that task. But I had not considered the RF devices and now that bothers me. The need to get on a ladder and take apart (&wake up) 6 PIR sensors and two triggerlincs does not excite me one bit.

I am sure my ISY does get lonely since I moved it out into the garage. Especially now without the company of its tag-a-long PLM. I agree with you Illusion. I don't really want to have an inferior, or less feature full product, that I never use and its warranty runs out sitting on a shelf. That was kind of my thinking in keeping a 2413U on hand. I could either use it for test or as a replacement for the system PLM. I prefer the USB version when doing testing. I figured I could swap out the daughter board if needed to the RS232 version for the a system replacement. I thought the PLM would fail after the two year warranty and I could use the RS232 daughter board from the defective unit ( a slight gamble there as well). It is nice to have a spare PLM for testing so it can serve a dual purpose. (e.g.) It was great to be able to use the spare PLM to send a direct message to the suspect PLM in order to confirm it was defective. I also monitored the line with a OScope just to be absolutely sure the PLC transceiver section was not operating. A Pro model would be great. Or at least a more robust model for use with the ISY. I would certainly be willing to pay a little more for a reliability increase.

My ISY->PLM just died. I have a new 2413S on the way from SM under warranty ( only 1.7yrs old). I thought I had a spare but that was a 2413U. I had intended to swap the serial daughter board from the bad unit into the new 2413U if I ever had an issue ( assuming the serial brd was not the problem). However since they are sending me a new 2413S under warranty I was not able to swap the serial daughter boards. This leaves me down for a few days and I really miss my ISY!!! Over time I had migrated towards a heavy emphasis on ISY directed items such as PIR detection actions, curtain control, and some scenes. Not having my ISY now has me reconsidering some scenes that I might now reconfigure so they to not require the ISY in order to still be active. My point here is that if you are heavily reliant on your ISY you might want to be sure to keep a spare PLM on hand ( if you do not already). Of course the ISY is so awesome I am assuming it will never be an issue I know the older PLMs tended to fail with a limited lifespan. I was not sure if the new ones would last longer. 1.7 years was much too short and I will now be sure to have a compatible spare on hand.

Great write up Brian, Thanks for your attention to detail and for sharing.

I vote for "Spill the guts!" And let us see all the gory details! Really though ... if you return it wouldn't it be worth a call to alert them of the issue? I know how frustrating it can be trying to get them to understand there is an issue so no one could blame you if you simply returned it.

Hi Brian, I know this is probably not really important information with regard to the 8015 probe but I did measure the minimum response time of the 2450 to a pulse on the sense input and the unit tested would activate at 40 milliseconds. Of course this value my vary a little from unit to unit but just for anyone's future reference. At the 40 msec it would send all four messages. This is pretty fast although probably a little slow still for the 8015. Will you followup with Smarthome as to your findings on the 8015 and return the unit? I am still curious, did they forget to load some parts or what? A device that will track the 60hz ac input with a square wave out seems to have limited usage? Especially if it is not pulling below 1 volt?

Agreed that a interposing relay is always an option and a good suggestion. Brian, You had spent enough time investigating that I thought it might be interesting to understand more about how "someone at SM" thought that the 8015 could be used with the 2450. You stated that an AC input to the 8015 produced a square wave output. How close did the "low time" come to ground, was it below 1 volt? You also stated that a DC input produced a pulse that came close to zero. Do you have an estimate of the time that that pulse was below 1 volt? I would like to know if the short pulse of the 8015 could possibly be enough to cause the 2450 to at least send an on/off command. OmegaQuest did not "see" a transition in the ISY status but it might have been too fast for him to observe (that is if the 2450 sent a on followed immediately by an off). If there were no log entries then I guess that answers that. We are talking about milliseconds here and I have no idea if the ISY would be able to trigger in a program if the 2450 did send back to back on/offs. I am just wanting to get a little more understanding. I intend to run tests early next week when I have access to the required equipment to determine the minimum pulse input required to cause the 2450 to recognize a pulsed input ( both in terms of relay activation and sending an on/off command).

Thanks LeeG, I should have stated, "what is the reaction time of the 2450 input". Minimum pulse time required? I do not have much time this morning but I set up a quick test. When I pulse the sense input (as fast as I can with a quick touch of a wire to input) I see the led flash briefly and the unit does send a command over the power lines. Perhaps this unit reacts to a faster input (communications wise) than it takes to energize an output relay. It would be nice to know the actual time requirements to active the individual components (comm. vs relay out). I could do this early next week for interest sake if Brian does not beat me to it If there is an ISY involved in this setup maybe a program can be used to catch the faster input and drive the output relays with another command (to pulse the relays for long enough to trip the doorbell?

Excellent information Brian, Thanks for including that test. As always a picture tells a thousand words and whenever one has access to an Oscope that is the way to go. I have a 2450 I/O linc but have yet to play with it much. I bought it for possible future use and so there is much I do not know about its operating modes. Do you know what the minimum pulse time might be that is required to utilize the 2450 latching mode? Is the pulse time of the 8015 too short for that mode?

Hi Brian, Thanks for that investigative work and report back. Sounds like the scope would be helpful. One thought I had is that perhaps the 8015s output "sources current" rather than sinks it? PNP rather than NPN output? You noted earlier that there are three connections and perhaps they were designed for the other equipment interface you referenced and not really compatible with the I/Olinc? Would you consider connecting a dc source (+5) to the 8015, and then a resistor (maybe 10k) from the output terminal to ground and then measure the output across that resistor ( while varying the input)? This would be a test exclusive of the I/O linc, or just using its supply and not its input circuit.

I will be looking forward to reading about your findings Brian. You are such a good fellow to buy one to experiment with. You would think they had a bridge inside so that polarity would not make a difference? Maybe only half-wave rectification based on the manual listing a polarity? I had an issue finding the 8015 manual. Can you please direct me?

While monitoring a different IOLINC sensor post I had grounded my IOLINC input and measured 1.7ma. That implied an approx 3K pullup. That along with your 800 ohms pull pull down also indicate an approx. 1.0VDC threshold. I had been watching this post in hopes I might be able to offer something helpful. Ouch ... Who would have thought to ask if the meter was on AC or DC Now you might go back and do the 18 volt battery test you mentioned in the beginning once again.

Thanks for the redirection to this post TJF1960, It certainly appears that there is an issue with the V98 firmware given that Smarthome has released a V99 already. This is very upsetting to me. The fact that Smarthome can release what appears to be a buggy firmware version and we have to be fortunate enough to read a post like this one to know we puchased a bad release? In my case I have already disassembled that particular V98 PLM and made a slight modification anticipating making another test device using it. Looks like I am stuck with this turd. May still make an ok tester but hopes to use it as a backup for my system PLM seem to have just gone down the drain.

Hello Tim2U, Unfortunately I do not have a low voltage probe to be able to test and answer directly. As a follow up to Brian's comments on this probe being isolated: You said you measured 9V out of the alarm. Was that open circuit, before you connected the low voltage sensor, or with the sensor connected? The reason I ask is that Smarthomes specifications are incomplete and they do not list an input impedance or current for that sensor. Listing only a voltage does not assure it will work with any 3-24V source. To be sure you either need to know the current requirement or you have to measure with the probe connected ( at alarm side) to be sure you are attaining the correct voltage level. (sorry if you posted that result and I missed it). Also what is your result if you substitute a 9V battery in place of the alarm as input to the low voltage sensor?

Good question Teken, I had not thought about that before. Now that you asked I wonder if that might be a bit of a challenge since the triggerlinc runs on one 1.5V battery. I examined the Motion Sensor battery sense circuit recently. No problem there since it runs on 9V. There is a large margin from 9V down to 6.8V before a low battery is detected. It might be a bit tricky setting the threshold for 1.5V operation?

Have you asked him?

Hello Zick, It is possible for line voltage to be very rapidly changing, faster than the response time of your meter. An oscilloscope or a fast recording meter are required to determine that. Since you have a test cord would it be possible to try this in a neighbors house or at work? This might help you isolate/identify your power quality as the possible issue. I am curious about the yellow (hot) wire. Yellow seem unusual, unless it might be associated with a uninterruptible power source?

LeeG, After your reply I realized that my problem was that I did not have the "02,& 03" groups linked. I was not using the ISY or HL and so I had to use a terminal program with the "write all-link" and "manage all-link record" commands to link groups 2 & 3 into the PLM and the 2420M. Once I got this setup I could then easily reproduce the low battery warning. As you said the MSensor does not need to be awake. Also as you said the warning was only sent once and required power removal and re-application to reset. I performed the testing repeatedly and measured a threshold of ~6.85 volts. This level will vary a small amount from unit to unit with part tolerances. I traced the circuit and can see where a resistor divider from the battery input to a microcontroller pin is used for sensing. The sensor continued to work ok down to less than 4.5 volts where it became erratic. From this I feel comfortable that having ISY programs (w/notification) to detect that one message, when it is sent, should be a good way to detect the low battery in advance and to be able to avoid the nuisance on/off cycles that occur when the battery gets too low. Thanks again for your help.

Thanks for that additional information LeeG, I ran quite a few tests using a power supply with a variable voltage as well as a variable internal resistance simulation to no avail. I may have missed the message though as I was unsure what to look for. I am pretty sure all I got was group 01 on/off commands but I will try again now that you have showed me what to look for. Do you happen to know if the unit has to already be awake to recognize the low battery or if it can wakeup on a low battery condition? When I tested I made sure to wake it up by introducing motion just to be sure.

Thank you LeeG, That will be very helpful. Is the msg log information you provided from a past occurrence you saved, or can you readily reproduce the message via bench testing?

Does anyone know what the msg string is that is sent by the 2420M for the low battery indication? I purchased a new R2.0 , 2420M recently. Before installing it I have been testing it on the bench trying to get it to send the low battery msg. I have been unable to get it to send anything other than the on/off commands. When I lowered the supply voltage I would start to get erratic operation at just below 4.5Volts. At that level I could readily get it to begin sending repeated on/off commands (every 30 seconds, with the timeout set to 30 secs). I was curious if anyone knew the string to expect should it send a low battery indication?

Teken, I had a chance to run some quick tests using a DB lamplinc (LL), DB keypadlinc(KPL), and and ApplianceLinc(APL) for the 4 tap bridge test. It appears to me that things work but that the manual is incorrect. It should read: 4 middle keys BLINK when on SAME Phase LEDs that are already ON, go SOLID ON BRIGHT when on OPPOSITE phase. I got consistent results with the above criteria. Tested 10 different combinations between the 3 devices and the other two devices were as per our normal expectations. One very important point is that some (of the 4 middle) keys on the KPL have to be ON ALREADY for the opposite phase test to give an indication. Which is really screwy cause if you do not have any keys on, then you get no indication at all that they are talking and on the same phase!!! It happens that my KPL on (LED) level is not full brightness, so I am unsure what happens if your ON level for the LEDs if full bright. Hoping someone else will explain to us how this design actually does make sense somehow?

Hi Teken, As it turns out I was just running some RF attenuation tests to gauge the ability of the new ELAM function to measure attenuation. I put a Triggerlinc inside of a aluminum box about 10" x 2.5" x 5" with a hole drilled so the only thing exiting the box was a plunger of a switch I had connected the ext. input of the triggerlinc. I then had the ELAM monitor the RF and also provide the DB translation to a IOLINC that I linked to the switchlinc. This worked great to be able to measure when the RF signal strength was too low to turn on the IOLINC. The aluminum box has a lid that I used to vary the opening to free air of the RF from/to the Triggerlinc. It was very cool to see the RF theory prove out. From within about 4 ft of the ELAM I was able to completely close the lid ( complete the Faraday cage) and see you no RF on the ELAM background level of ~400 -500 millivolts and the IOLinc would not activate. All it took was about a very slim (1/8" slot) in the box opening and the IOLinc would receive. ELAM showed only a very marginal increase to ~550 millivolts. I found that as long as the box was open at least 1/2" (slot) it did not matter all that greatly the direction of the opening in the box to the receiver (in an ideal bench test). All these tests were 4ft from sender to receiver: Open air between = 1070 millivolts box closed on 5 sides = 800-900 when opening was facing toward the receiver. box closed on 5 sides = 700-800 when opening was facing away the receiver approx same results with a varied opening of 5" down to 1/2 inch. What I am afraid will make a large difference in the real world install is reflections that will be more greatly affected by the box opening restrictions. As one other point of reference I was measuring about 500-600 mv level at one KPL in my home when the other keypadlinc was the 4 tap initiator and they are about 30 ft apart (as the crow flies).

Teken, Seems we are in agreement on what the manual says, perhaps it is a miss print? From what you have said about the changing behavior, based on who is the initiator, makes me wonder if there could be a firmware error? I have never taken a close look at the 4tap test before this and the DB keypadlinc sure is not very user friendly in that regard. I do not have a spare 2487S to test with on the bench so my testing is limited to the house and my access to that may be limited. My bad for not ordering a spare of every device, just for testing purposes Hopefully a third person will be kind enough to offer their results or a clarification. My current understanding of the RF (radio wave) vs PLC (power line comm) is that the RF can be accomplished without regard to the zero cross for some messages, (e.g.) from RF only devices to a DB unit. To perform the 4 tap test, the messages used are "RF only", yet the units require a good clean zero cross signal in order to determine phase of the power line (for a correct LED blink result). So I can see where a noisy environment might cause issues in what on the surface sounds like RF only. I would not expect a signal sucker to affect the 4 tap test but I could envision noise affecting it in certain scenarios? If I can get access I will try to duplicate your results of the change in LED blink behavior based on who is the initiator.

Hi Teken, Will this Insteon fun never end In my earlier testing I was focused on evaluating the RF portion of my ELAM. I had to download and read the manual for the 2487S to figure out how to initiate the 4 tap test, because first attempts failed due to the difficulty I was having doing the 4 taps fast enough. I did not pay enough attention to the 2487S light sequencing. After re-reading the manual I see I missed the footnotes the first time. Now I am totally unsure of what I am seeing. Please help me understand how you interpret the manual? I thought is was saying this: "4 middle leds blinking = in RF range and opposite phase" " LEDs going solid BRIGHT = in range and same phase" " no LEDs = out of range"??? When I test using an access pt "for sure" on the same phase as the 2487S I get the 4 LEDS blinking. Am I understanding the manual incorrectly (error in manual), or am I just too old Then when you throw in: the Bright LED vs....out of range and no LED vs... your body interrupting the RF signal, it all seems very, very confusing.?