larryllix

I recently thought about switching my Asus to Merlin to resolve dropping the WiFi frequently but after reading user complaints many have reported it did not help the problems. However I did find one thing that has solved my router drop out problems. With the Asus router you can reserve too much QoS bandwidth and when to many QoS app are running the router just shuts down. It runs much better with the automatic QoS (oxymoron) or just shut the nonsense off completely for everything except your VOIP, with minimal bandwidth reservation, as needed.

I have a Netgear Nighthawk AC1900. The ventilation with them sitting horizontal on a table doesn't seem to be adequate for summer months and mine overheated. This caused me a lot of grief for a long time trying to find this problem. I finally placed a CAO Tag on the surface of the router, and every time the temperature hit 42C my LAN started crapping out. My printer would stating warming up doing it's self-checks and devices would disconnect and heal etc.. If you can keep it cool enough it's a great router. I also found the antennae are not sufficient to turn on the 5Gz and 2.4GHz to 100% simultaneously. It now works much better at about 25 or 50% with much better range on the 5GHz, strangely enough. From what I can figure the signals must be overloading the capabilities of the antennae system. I had so much trouble with this router I bought an Asus to replace it. I thought it was the power supply crapping out but when I would change PSs it would cool down and work for a few days again. I have it back working again now and it works fine in the winter or if I keep a stream of air through it in the summer. It really needs to be mounted vertically due the bad ventilation design when horizontal.

Mine does that very frequently if I have two devices with the same words in the name. It seems if Alexa can't make out the second word absolutely it favours the first one only. Usually saying exactly the same phrase gets it.

With 18 of them operating here, I have no Insteon problems and no filters. If you don't want RGB, then just turn on white. You can get JUST white strips and you might save $1-2 on them. A smaller PS might be used but I doubt they are any cheaper. I never turn off my power supplies. I would have to reach behind cabinet or climb on counters to get at the plugs. They are all switching supplies so they draw next to nothing when off. I measured all the power draws for this and posted it about a year back. If you want neat appearance and waterproof, I wouldn't cut these strips into segments. Just bend them around the corners and tie wrap them to the frame. Some rounding will be necessary to not break the PCB inside.

admin/admin should work after the Factory Reset you just did. Try powering down your ISY and PLM again. Power up the PLM first and then after about 15 seconds power up the ISY. Try again. You programs and setting will be gone. I hope you have a backup.

I play with my programs frequently, but in view of the 15-25 minutes to boot the admin console and run the backup, I try to backup after major changes and maybe once per week, if I am in that computer location. UPDATE: OK I exaggerated for my faster 8 core machine. I timed it this time... ... it took 45 seconds for admin console to load, and 3 minutes 20 seconds for backup to save the file, total 4:05 I don't use my slower machine for backups as the times listed initially are just too brutal. Love that XML! I have been caught with an out of date backup once and it wasn't fun.

ISY firmware v5.* has other options to generate a date without year internally and self dependant. However v5.* is beta software yet. http://forum.universal-devices.com/topic/17157-more-flexibility-in-schedule-setup/

Sounds like you might have missed the "Controllers' part of the setup advice.

http://forum.universal-devices.com/topic/19913-magic-ufo-controller-for-rgbw/page-1?do=findComment&comment=189572 See new post #199

Two Outback 3.6kW inverters with no problems.

Use the old cable installers trick. Coat the splice point with silicone grease or caulking and slip heat shrink over it. Now apply heat and the silicone oozes out for a hermetic seal that is good underground for...well they haven't failed yet. Many of the phone companies splices come that way. they crimp, snap together and bury it.

IP 67 rated RGBWW LED strips are completely sealed in a rectangular crosssectional silicone tube. Run them into a sealed wiring box with some hole sealant and run the cord out of the box with a cord sealing grommet. Many of these controllers are WiFi now and that would give you some festive options from a cell phone app or your ISY NR module. RGBW strip (5m) $12 Controller $12 Connectors $3 Power Supply 12v 3-5A $12 Electrical box $25 Grommet $1 Prices are aproximate.

Yes, I agree it is a very cool project. A little envious, as I did many projects like that in my past, and have a lot of respect for what he did.But these things are changing very fast and you can end up re-inventing the wheel. Been there done that too. Ramping and other features would be cool but, as I have found in the 5 channel units the little CPUs don't have the horsepower to do too many fancy things. The 5 channel units have a slight ramp built in, but when it is happening, the controller CPU does not answer the http: call and your supervising software now has little delays in it. This really destroys the smoothness of animations going on. Better hardware inside the controllers (like timer interrupts) would be nice to accommodate these drools. I am doing some ramping in my software for throbbing effects but with my test 10 units and Ethernet WiFi I see some strobing going on, where they get out of sync. Not a bad effect, in itself, but it demonstrates the delays can add up. Inside each controller is definitely the place to do it = distributed intelligence, if it can be done. The 5 channel units have built in effects...about 22 of them. Strobing with about 5 single colours, seven colour strobing, fading, and some fades between the major colours. You can see the jags in the ramp, somewhat. In the end, after thinking I was going to do all these eefcts in my house I realised I do none for HA. It's annoying. But I have another 18 of them I am creating animations with, not for my house though. Cripes! I don't even mix colours after replacing all my MiLight controllers fof not being capable of mixing pastels. As you develop you question yourself and sometimes it become clearer you are wasting your time... but having fun.

That's really cools stuff but no IoT is required. Most HA people are trying to avoid that option. Programming microcontrollers is not required. ISY is already all that and then some. When my garage door is open one RGBW flashes red and returns to whatever it was once the garage door is closed. I have five other bulbs across the wall behind my television that flash red in a wave across the front wall at midnight each night, if we are home, as a reminder to go to bed. When I ask my Alexa boxes for Movie Lights multicolours all appear around the room as well as some white overhead ones, very dimly. These are all http: protocol items driven by ISY programs sending http: protocols via ISY NRs. This is what ISY can already do. @Scott. NRs send http, https, TCP, UDP and a few other protocols with option variations for each. NR is the firmware module that can do all these. You are already sending http: packets to your UFO RGBW strips controllers from your ISY as well as your other system and the handheld remote? and mobile phone app.

Oh I agree we have a lot of fun challenging ourselves with programming, but http: is the protocol the UFO and 5 channel LEDenet units use already. For somebody that doesn't have anything yet, and wants general lighting with White. I love these controllers. They blast WW (3000K) in my room like a hospital operating theatre, and yet dim down to 1% (barely visible) in any colour I could want and very rich colours, not like my first $200 worth of Hue bulbs.

Nice project but.... This seem like an awful lot of work to save about $2-3 for a 4 channel unit instead of a 5 channel unit that already accepts http: commands directly. Not that anybody would use the 5th channel for RGBWW. http://www.ebay.com/itm/4A-5CH-5CH-RGB-WW-CW-LED-WIFI-Controller-IOS-Android-Smart-Link-Timer-DC-12-24V-/232301057166?hash=item361638d48e:g:ZSgAAOSwls5Y7x0L I bought many of mine for $11 Canadian, about $8 USD.

Just send whatever R/G/B/WW/CW bytes you want it to be. There are effects and a speed control for them but mostly useless. I have discovered the brightness levels work well with the squares of 0-10 for a nice small stepped ramp. I am using python3 so don't try this at Home ISY. In view of the constrants you shouldn't need more than off, dim, med, bright anyway.

That's exactly what I described and almost every effect / action can fall into. Program Water.initiater If ...trigger conditions (could be a list of daily times) Then ....Wait a while to be sure ....run program water cycler (if) ....disable self (optional with same below) Else .... Program water cycler (disabled) If ....lots of water needed (based on brightness and/or wind/temperature conditions etc. ............................. Outside humidity maybe best) Then ....disable Water.initiate ....turn on water ....wait 5 minutes ....turn off water ....wait 55 minutes ....enable Water.initiate Else ....disable Water.initiater ....turn on water ....wait 1 minute ....turn off water ....wait 4 hours ....enable Water.initiater

What connection method and software are you using to flash the eprom inside? This unit appears to be the same unit Scott (Mr. Cool White) has been playing with lately. His supports WiFi and iR input and was really cheap.

The usual and safe way to circumvent triggers (the we need to star a process) from stopping the process is to creat two programs. -Use one (enabled) to see all the triggers and call the second program from it. -Use a second program with any further wanted conditions (maybe twoo different timers or processes) and a timer that ensures the job will complete. -Disable program Two from any trigger influence.

The logic is being continuously evaluated. If it ever become False Then stops running and Ese starts running. You could set a variable true/false and test that with Repeat While $sVariable = $cTrue You could use one program to initiate on a cyclic basis the other with conditions. You cannot have more than one condition is a Repeat While, though. BTW: The Repeat Every is a bt of misnomer. Your loop will happen every 2 hours and 10 minutes. 10 minutes is a lot of water once the ground gets wet! I did this years ago with an X10 system and used 3, 2 or 1 minutes, depending on the brightness in the house. Then I didn't cycle at all at nights. Grass grew so fast that a neighbour commented on the speed.

ewwwww. That does break the trigger concept for programs. Now we would have to write ISY code to compare parameters with a known/old value in a variable to detect a change. This doesn't sound like a workable solution for ISY logic. Does that mean every program based on a node will notify the user on Query All every night? In my case the value was triggering programs from REST input to those variables. Without a value change programs were running every time a REST value was inserted. That means with every usage of a REST/VAR/ command a series of programs in ISY will evaluate and run Then or Else, and no change is detected. EDIT:Read Michel'sl logic very carefully. (I didn't at first). This is actually a good testing feature. We can initiate programs manually right from the horse's mouth.

I found the same problem with State variables updating to the same value, back about v5.0.2, UDI reported to be fixing that one. What firmware version are you finding the status false trigger with?

OK. One thing I have found about these LEDenet controllers you should know up front. They take TCP protocol and that is not good with real time. That means that your router and/or driver software may do things like, - let packets passed each other so they get out of order, (this seems to be OK at higher speeds) - clump packets together to increase the throughput, (destroys smoothness of animation) - disconnect sockets, based on a timeout, and they take a part second (0.1-0.7) to reconnect again. They have to be constantly poked to remain fast responding (connected). I have found high speed animations look smooth and really slow animations are OK but medium speed animations can get out of order and sporadic. I suspect these controller tiny CPUs are busy doing some ramping etc. and they don't respond to the TCP requests fast enough. They should have been UDP protocol. I doubt these controllers will satisfy what you have done with DMX.

PDP-11? No PDP-8? Just a newbie!