Top 10 Features I Would Like To See In A Lighting Protocol

[Brightan]

Getting sick of UPB deviecs that aren't in the mode they say they are. Also second the no neutral. A nightmare rewiring old three and four way switches that had no neutral at one end.

here's a question... The neutral is required because the switch itself needs power and so it has to get power between the hot lead and the neutral lead, but, in N.A. wiring (and presumably elsewhere) the ground wires and the neutral wires are tied together in the breaker panel. So, the question then becomes, can a switch be run by connecting the neutral connection to the box ground? I know somebody is going to say that this would be a violation of the electrical code, but my point is, despite the code, why not? The tiny bit of current required for the switch shouldn't be a problem. The light itself would not be using the ground as the return path because it has its own neutral.

Ok, so I figured I better try it out before posting, just in case I was missing something, and, it does in fact work as normal. The amount of current draw from the switch is the same when wired through the neutral as it is through the ground and stays the same whether the light is on or off, meaning the current for the light doesn't go through the switch neutral. What I found more interesting was that this was tested on the new Jetstream light switch and the switch current was 100mA which is 11 watts of power! Now I needed to know what was going on inside the switch... I tested the current requirement for the electronics themselves, before the 3.3v regulator, and it is less than a 1/2 watt. Where the other 10+ watts are going, I don't know at this point.

So, maybe I answered my original question about the ground as the switch return, maybe 11 watts is just to much to "sneak by"

 

[upstatemike]

Getting sick of UPB deviecs that aren't in the mode they say they are. Also second the no neutral. A nightmare rewiring old three and four way switches that had no neutral at one end.

here's a question... The neutral is required because the switch itself needs power and so it has to get power between the hot lead and the neutral lead, but, in N.A. wiring (and presumably elsewhere) the ground wires and the neutral wires are tied together in the breaker panel. So, the question then becomes, can a switch be run by connecting the neutral connection to the box ground? I know somebody is going to say that this would be a violation of the electrical code, but my point is, despite the code, why not? The tiny bit of current required for the switch shouldn't be a problem. The light itself would not be using the ground as the return path because it has its own neutral.

Ok, so I figured I better try it out before posting, just in case I was missing something, and, it does in fact work as normal. The amount of current draw from the switch is the same when wired through the neutral as it is through the ground and stays the same whether the light is on or off, meaning the current for the light doesn't go through the switch neutral. What I found more interesting was that this was tested on the new Jetstream light switch and the switch current was 100mA which is 11 watts of power! Now I needed to know what was going on inside the switch... I tested the current requirement for the electronics themselves, before the 3.3v regulator, and it is less than a 1/2 watt. Where the other 10+ watts are going, I don't know at this point.

So, maybe I answered my original question about the ground as the switch return, maybe 11 watts is just to much to "sneak by"

I'm not going to get into the old code discussion but I will mention that my Levton motion sensing switches do exactly that. The run off the hot and ground and use a relay to switch the load. They have all the UL and other certifications so if they can do it legally with those, why not with UPB or Insteon?

 

[123]

Search the web for "purpose ground neutral wires". In brief, the described scenario causes a difference in the ground potential among outlets (i.e. ground is not "zero volts" when measured between outlets). It can cause an electrocution hazard in a special situation (i.e. neutral line is broken at one outlet and current flow, by design, returns via ground). Plus, the inconsistent ground potentials can cause headaches for distributed AV equipment ... ground for outlet A is not at the same potential as ground for outlet B.

 

[Digger]

Wouldnt this also cause havoc with a GFCI?

 

[Brightan]

Wouldnt this also cause havoc with a GFCI?

I think a typical GFCI will trip at something like 10mA, so yes, as soon as you have more than 10mA of current going out and not coming back through the neutral, the GFI would trip.
In my testing it was 100mA which would definitately have tripped a GFI. I guess the bottom line is that even a couple of mA would be too much as it would create a ground loop. I am still surprised they would draw 11 watts just sitting there...

 

[Steve]

I am still surprised they would draw 11 watts just sitting there...

Is that with the LED off? How does it change with the LED on or off?

 

[Brightan]

I am still surprised they would draw 11 watts just sitting there...

Is that with the LED off? How does it change with the LED on or off?

it is the same with the LEDs on or off but that is because my meter is not accurate enough. A 10mA LED at 2volts is would only account for 1/50th a watt.

 

[hucker]

You guys sure about 11W at the switch? That is a lot of heat, the switches would be quite warm. My UPB switches consume very little -- it is hard to measure on a kill-a-watt meter even if the switch LED is on.

 

[hucker]

Back to the protocol question, these are sort of aimed at UPB and I suppose is expanded a bit to look at a system level than just the communication layer.

1) It would be nice to be able to control the LED via the protocol rather than the device configuration.

2) Tapping multiple buttons/rockers on the same switch body should do what you would expect. If I hit both the bathroom fan and the bathroom light they should both turn on or off, instead of doing nothing (SAI are you reading this!). It would be easy to handle in the firmware -- have an link triggered when both toggles are pressed.

3) More work on reliability for wired solutions.

4) The protocol should support timers. I should be able to send a command that says turn this light/scene on for X minutes. Also build in sunrise/sunset stuff. I know you can pack a lot into the flash on the processors they use. They could even advertise it as Green!

5) PC Based software should be available that allows the stability of the network to be tracked and verified so warnings could be provided to the owner/installer about degradation in performance. Basically a PIM plugged in to a USB port on the automation PC.

6) More Speed, ACK on the firmware upgrades, ideally over the power line but I'd be OK with plugging in a USB cable. I'd spend a lot of time walking around my house to upgrade to the faster response time on my old UPB switches.

7) Dimmers with LED status.

8) I imagine that with LED lights coming out over the next few years that color should be an option in the protocol. Ditto for fan speed control. If the 'switch' was a speed controller it should respond to dimmer like commands, just don't call it the 'dimmer' level since I'm told by EE's that you can't dim a fan.

9) The dimmable CFL's that I've played with have extremely non linear response. It would be very cool if I could program a curve that linearized the switch action. For example, my CFL's do amost nothing as you start dimming, then it moves pretty quickly then at about 30% it is all flickery. If I could say dim_max=85, dim_min=30 that would go a long way to increasing the WAF. Bonus points for having a light sensor on the switch to auto generate the table

10) Support for broadcasting power consumption in Watts, or kWhrs. Most PIC's have AD's so this shouldn't require any extra 'expensive' hardware and could just be extra status bytes on state changes. Perhaps not as fancy as the whole house monitors like TED and Bruhltech but data is data.

11) How about a soft on during power recovery for lights configured as dimmers.

12) How about switches with IR blasters on them so you can power up and down AV stuff. (Most likely just OFF).

 

[Brightan]

You guys sure about 11W at the switch? That is a lot of heat, the switches would be quite warm. My UPB switches consume very little -- it is hard to measure on a kill-a-watt meter even if the switch LED is on.

I was thinking the exact same thing last night, where's all the heat going? As it turns out, I pulled a total newb and was measuring AC current on a multimeter that is not a true RMS meter. The switch has a power factor of .14 which means my meter reading was completely skewed . The real value is more like 2 watts.
Sorry for the hijack, back to your regular programming...