Remote Switch Devices Turn On By Themselves & Energy Savings

[jack0987]

I have a bunch of X10 modules (first image) like in picture and thought I would put them to use.

With some electronic devices plugged in, the modules would turn on by themselves after being shut off.
On one computer plugged into them, it turns on and then shuts off repeating this over and over about every thirty seconds.

Others have expressed this about other systems such as in second image.

To correct this problem, I added a double pole relay to the power circuit following the module to disconnect both line and neutral. This seems to work.

Three things I can think of:

1. Why does this happen?
2. Since there still seems to be a connection in the off mode (not to mention the added energy consumption of the module itself), are we really saving any power?
3. Other solutions?

 

[rjenkinsgb]

X10 is an early, fairly simplistic system, that uses high frequency bursts near the AC zero crossings for signalling.
A lot of things can cause electrical noise that may result in false on or off switching.

Noise suppressors can short out all high frequency noise, but that includes the X10 signalling and powerline data systems.
I have a load of X10 modules still, in storage as they are now not at all reliable.

Z-Wave is the best alternative at the moment, as that uses RF in a mesh network, rather than powerline signalling.

 

[jack0987]

Z-Wave is the best alternative at the moment, as that uses RF in a mesh network, rather than powerline signalling.

Please suggest a Z-Wave system(s).

 

[rjenkinsgb]

You can get direct remote controls, or USB interfaces to work with home automation software, or some automation hubs have Z-Wave built in as standard.

I use Homeseer with a zwave.me USB interface. I've built systems for relatives based on the same setup, in some version - you can get a version of Homeseer to run on a Raspberry Pi, which makes a very cheap and powerful system.

If you get the homeseer stuff during a sale, it's very cheap.

Virtually every Z-Wave device is compatible with every other, it is a very well supported and standardised system.

Just be sure to get the main interface and any remote devices for the correct part of the world, wherever you are - eg. EU and US versions use different radio frequencies and cannot work with each other, but most makers have versions for each region.

This site has a fair selection to show the type of thing available - but search elsewhere for prices as well!

Z-Wave

Z-Wave is an advanced wireless home automation technology that enables you to create a smart home environment where your electrical products, lighting, heating and blinds can work together as a single system. As Z-Wave is a modular system you can start as small as you wish and then expand your...

www.vesternet.com

I've use loads of TKBHome mini modules like the TZ74 dual switch, but it looks like they wen out of business during the pandemic, unfortunately.
Almost identical to these, in fact - but a third the price from Aliexpress..

Z-Wave Philio 2 * 1.5 KW Relay Inserts

The Philio Z-Wave Relay Inserts enable you to wirelessly control two independent loads.

www.vesternet.com

Bit cheaper here:

Relay Insert 2 * 1.5 KW

Relay Insert 2 * 1.5 KW

z-wavelab.com

 

[AnalogKid]

I've never had any problems with the appliance modules. The lamp modules, both plug-in and the light switch type, are TRIAC based. These leak (electrically) and cause some LED bulbs to glow in the off state. It is possible for a 0.5 mA leakage current to charge up the supply's main input filter cap to a voltage that is greater than the controller chip's UVLO (Under-Voltage LockOut) trip point, see what it thinks is a valid input voltage level, and start. The instant it starts, the drawn current discharges the main cap below the UVLO level, the chip shuts down, and the cycle repeats. Basically, you have a hysteresis oscillator.

The solution for a lighting circuit is to have at least one incandescent bulb in the circuit. In your case, plug an outlet strip into the X-10 module, plug the computer and a desk lamp into the strip, and see what happens.

RJ - if you were on this side of the pond, I'd gladly relieve you of your X-10 burden. That applies to anyone else - think of me as an X-10 rescue shelter.

ak

 

[jack0987]

The lamp modules, both plug-in and the light switch type, are TRIAC based. These leak (electrically) and cause some LED bulbs to glow in the off state.

Ah!. Leakage. So, even though it is off, it is not really off wasting energy. That's why my using a relay following the module most likely fixes the problem. It physically disconnects the line (no more leakage). And yes, I have experienced the issue where a string of LED lights would still stay dimly lit. I thought that by disconnecting the neutral as well was what was doing the trick but now I am not sure if that is the issue at all.

I can see where electrical noise can sometimes be an issue. For example, I think my xBox comes on occasionally when using my microwave. LOL

The only thing I can think of is to continue building my relay switches if I am going to use these things.

Energy saving is a big deal for me. You would think that practically living next door to Quebec, Canada would help. It does not. They have a 25% surplus of power and I have one of the highest rates in the nation. I have two computers that I use once in a while but are mostly shut off would not be a problem. When I physically unplugged them, the savings seemed to be quite significant. Do they use that much power even when off?

BTW: What do you do with rescued X10s?

 

[rjenkinsgb]

Energy saving is a big deal for me.

I recently got this - it could be of interest for energy monitoring & saving??

I've not installed it yet as I'm just updating my homeseer setup; the machine it runs on at present is an old desktop type PC, which itself uses more power than needed - I got a dedicated fanless mini PC in the Homeseer black friday sale, it's just arrived in the UK after a month with USPS! On sale, it's hardly more than the PC maker sells the machine alone for, though list price is a bit on the high side...

Aeotec are very helpful, I wanted the 100A 230V single phase version which no one stocks - they sent it to me direct, cheaper than most places sell the standard 60 for!
It's Z-Wave and fully compatible.

 

[jack0987]

I got a dedicated fanless mini PC in the Homeseer black friday sale

I am not a big fan of Raspberry Pi.

Will Homeseer will run on a windows pc?

I was thinking of using a mini ATX motherboard (fanless, of course) with a battery backup. This way I can program in C#.

How is the battery life of their wireless remote sensors? So far, on the Ardunio ones I built I can only get about 6-8 weeks using one 18650 rechargeable.

 

[AnalogKid]

Based on a very rough calculation, a leakage current of 0.9 mA would charge a 220 uF cap (power supply bulk filter cap estimate) up to 120 V (the peak voltage of an 85 V sinewave) in approx. 30 seconds. Checking a 2002 Teccor catalog, this is well within the range of leakage current values for room-temperature TRIAC devices. If this is in fact what is causing your oscillation, putting a light bulb in parallel with the supply should bleed of this leakage current and quiet the supply. It could be something as small as a 4 W or 7 W nite-light bulb. Note that the bulb is drawing that power level only when it is on.

Disconnecting the neutral certainly will break the circuit and stop the oscillation, but at what cost?. If there is no evidence of the relay chattering (indicating a significant trickle current through the coil in the off state), then it probably is the case that the relay's minimum operating current is much greater than the leakage current. But the leakage current still is there, it's just in a place where you cannot see any consequences.

1 mA at 120 Vac is 12 mW. That's not much. My guess is that your relay coil is drawing more than that when it is on. For more, post data (part number / vendor site / datasheet) for the relay.

Depending on the age of the computers, their power supplies might be "Energy Star" (or some Canadian equivalent) compliant. As such, the housekeeping supply will be below the max for the standard in place at the time of manufacture. For example, a 200 W supply might have a max. 5 W housekeeping supply. This is the supply that powers the control circuits to enable/disable the main supply. Any TV, DVD player, etc. that has an On/Standby switch rather than an On/Off switch must have a housekeeping or standby supply.

X-10's - If they work, I use them. If they don't, I cannibalize some to repair others.

ak

 

[Nigel Goodwin]

How is the battery life of their wireless remote sensors? So far, on the Ardunio ones I built I can only get about 6-8 weeks using one 18650 rechargeable.

Dump the Arduino - it's a development system, and is useless for low power applications, as there's too much stuff on-board wasting current. By using a PIC, suitable hardware, suitable software, and sleep mode you can drastically reduce power consumption - to years for an 18650.

If you like the Arduino system, then try the Arduino Pro Mini, which is essentially an Arduino compatible Atmel328 on a tiny board - and while it's not as low power as a PIC, it can still be used down to MUCH lower powers than the full Arduino boards.

But whatever processor you use, shut EVERYTHING down entirely (use the processor to switch the power off to everything), and just use the processor in sleep mode, waking periodically to read sensors. Generally I use a 32KHz crystal on TMR1 which runs in sleep, and is incredibly low power, and wakes the PIC up once a second to update an RTCC - use the RTCC routines to check if it's time to do something, once it is, then do it and go back to sleep.

 

[jack0987]

If you like the Arduino system, then try the Arduino Pro Mini

I am using an Arduino pro mini set to sleep all the time and wake only if a change in sensor state is detected. I did not disconnect the onboard LED. Going to try that next.

I do like the pics and have pic stuff but they are kind of expensive to work with unless you are a pro which I am definitely not. I need things mostly already designed and I can make minor changes.

 

[jack0987]

Disconnecting the neutral certainly will break the circuit and stop the oscillation, but at what cost?. If there is no evidence of the relay chattering (indicating a significant trickle current through the coil in the off state), then it probably is the case that the relay's minimum operating current is much greater than the leakage current. But the leakage current still is there, it's just in a place where you cannot see any consequences.

1 mA at 120 Vac is 12 mW. That's not much. My guess is that your relay coil is drawing more than that when it is on. For more, post data (part number / vendor site / datasheet) for the relay.

OMRON LY2 AC110/120 GENERAL PURPOSE RELAY 110/120VAC 10A DPDT PLUG IN 8 PIN

 

[jack0987]

If this is in fact what is causing your oscillation, putting a light bulb in parallel with the supply should bleed of this leakage current and quiet the supply. It could be something as small as a 4 W or 7 W nite-light bulb. Note that the bulb is drawing that power level only when it is on.

How about a neon bulb?

 

[jack0987]

If you like the Arduino system, then try the Arduino Pro Mini, which is essentially an Arduino compatible Atmel328 on a tiny board - and while it's not as low power as a PIC, it can still be used down to MUCH lower powers than the full Arduino boards.

My setup such as it is.

 

[AnalogKid]

How about a neon bulb?

That's a good question. I think I can argue it both ways.

If you can see it glow, it is drawing current. But most small neons need 60 V or so to break down (begin conducting and glowing). If you don't see little flickers, it probably isn't helping.

ak

 

[Nigel Goodwin]

I am using an Arduino pro mini set to sleep all the time and wake only if a change in sensor state is detected. I did not disconnect the onboard LED. Going to try that next.

I do like the pics and have pic stuff but they are kind of expensive to work with unless you are a pro which I am definitely not. I need things mostly already designed and I can make minor changes.

Disconnecting the LED is something you MUST do - as it uses massively more than the processor in sleep. However, you should also use transistors, or FET's to switch all sensors etc. completely off when not in use.

 

[Nigel Goodwin]

My setup such as it is.

View attachment 143898

Well you've got the LED nice and bright, plus the sensor (whatever it is) appears to be permanently powered. So battery life is severely compromised. Have you tried measuring the current it's taking from the battery?.

 

[AnalogKid]

View attachment 143897

At 110/120 Vac, the coil current is around 10 to 11 mA, considerably more than the TRIAC leakage current. In terms of system energy, it uses the same energy in 1 hour of on time as the leakage power during 10 hours of off time.

ak

 

[jack0987]

Disconnecting the LED is something you MUST do - as it uses massively more than the processor in sleep. However, you should also use transistors, or FET's to switch all sensors etc. completely off when not in use.

I am using a magnetic sensor that is normally open. When the magnetic field is applied, it is closed grounding the Arduino pin.

 

[Nigel Goodwin]

I am using a magnetic sensor that is normally open. When the magnetic field is applied, it is closed grounding the Arduino pin.

Exactly what are you trying to do?, to get decent battery life you need to power up sensors occasionally, take a reading, and then shut it back down. If it's something that needs continuous monitoring, then you can't do that, and you'll struggle getting good battery life.

For example, ESP modules can be put in sleep, and a slight mod (piece of wire) done for it to wake up periodically - but the current consumption is still really much too high, presumably due to the extra components on the module?. My solution is to have a PIC controlling it, by switching the power ON and OFF, and sleeping in between actions.

As a 'rule of thumb' it's usually considered that the average current needs to be less than 100uA, and as far below that as possible.