FGD-212 question

[Tiwas 0]

Hi.

I wanted to use my Dimmer 2 to control my new LED strips. I have two 220V -> 12V 10A transformers, one to each strip. They'll be running at max 7.5A with the length of the strips I'm running.

 

For my first run I connected one of the drivers and had to use a bypass to get it to turn off completely. With the bypass it works excellent and after forcing a calibration dimming is perfect. Or - as close as can be expected. I tried the same setup on the other driver with the other strip and that works perfect as well. Now, if I connect both of them to the same FGD-212 (still with the bypass) the lights will come on, but it will then turn off again after dimming up to 100%. The load should be well within the dimmers capabilities with 0.78A on the 230V side or 180W if you will. The FGD-212 reports 78.4W, when one is connected.

 

Anyone know what this might be? Is there a max to what the bypass can handle?

[drboss 107]

the problem is connection of two transformers at the same time to one dimmer

[Tiwas 0]

48 minutes ago, drboss said:

the problem is connection of two transformers at the same time to one dimmer

 

Why would this be a problem? The transformers are exactly the same, the load is exactly the same and even the length of the led strips are exactly the same. Why would the dimmer differentiate between a single load of 160W or 2x80W? I'm not arguing, just want to understand the reasoning.

[petergebruers 1,265]

2-Wire or 3-wire setup?

 

What is the power factor (cos phi) of the driver? If it is less than about 0.95 you have to factor it in, it increases the load on the dimmer.

 

Please login or register to see this link.

 

If it is 2-wire, try to reduce parameter "2. Maximum brightness level" in steps of 5.

[Tiwas 0]

1 hour ago, petergebruers said:

2-Wire or 3-wire setup?

 

What is the power factor (cos phi) of the driver? If it is less than about 0.95 you have to factor it in, it increases the load on the dimmer.

 

Please login or register to see this link.

 

If it is 2-wire, try to reduce parameter "2. Maximum brightness level" in steps of 5.

I'm not sure how to exactly do the calculation, even after reading the wikipedia article, but the power consumption is only slighty higher than expected so I would say the power factor is high. I also bought over sized drivers (they can handle up to 120W even though I only need 75W max) by choice in order to put less strain on them. According to the manufaturer specification for the LED strips, 7.5m should amount to 75W and the dimmer reported 78W for one and 78.4W for the other. If I'm correct, I should then have a power factor of 150/166.4=0.9 ?

 

It's a 2-wire setup, but shouldn't auto calibration figure this out? I've set parameter 13 to 2 (as I have a bypass 2 connected).

 

But...with only 166.4W, should the power factor really matter as it's far away from the max currenct supported by the dimmer?

[petergebruers 1,265]

14 minutes ago, Tiwas said:

'm not sure how to exactly do the calculation, even after reading the wikipedia article, but the power consumption is only slighty higher than expected so I would say the power factor is high.

Thank you for reading that article!

 

The power measured by the FGD is "active power".

I am guessing, you do not have a tool to measure power factor or the "apparent power".

And "apparent power" is always higher than "active power".

 

I driver might have PF = 0.75 (if you cannot measure it, it should be in the specs.). In your case, the FGD-212 reports 78.4W. But it has to handle 78.4/0.75W  = 104 Watt. I do not have the exact details, but I bet they use a current limit, to shutdown in case of overload, so that is based on "apparent power" (*). With this number and 2 drivers, you still do not exceed the spec...

 

But anyway... If your controller does NOT report "overload" or "overcurrent" then that is not the reason for its shutdown.

 

14 minutes ago, Tiwas said:

It's a 2-wire setup, but shouldn't auto calibration figure this out? I've set parameter 13 to 2 (as I have a bypass 2 connected).

 

It is not perfect. Make a note of parameter 2 after calibration, then decrease it in steps of 5... It is worth trying, it has been done by many users.

You might want to test, in "lab conditions" if it works better in 3-wire setup. If that still does not work, the dimmer cannot handle your load.

 

EDIT

(*) For sake of completeness, PF and apparent power numbers work best if the load causes "sinusoidal" current. It was conceived in an era of iron core transformers, inductors, motors and capacitors. It does not give enough information to describe loads caused by "switched mode power supplies" like LED drivers.

Edited March 15, 2018 by petergebruers

[Tiwas 0]

40 minutes ago, petergebruers said:

...

But anyway... If your controller does NOT report "overload" or "overcurrent" then that is not the reason for its shutdown.

...

 

It is not perfect. Make a note of parameter 2 after calibration, then decrease it in steps of 5... It is worth trying, it has been done by many users.

You might want to test, in "lab conditions" if it works better in 3-wire setup. If that still does not work, the dimmer cannot handle your load.

 

Thanks a lot for taking the time

 

I will try changing parameter 2 and see what happens. There's no warning I can see about overload or overcurrent, which is why I was reluctant to accept that as a solution.

 

I will also try removing the bypass. Even if it was needed for only one driver it might not be needed for two.

 

Regarding the 2-wire connection there's little I can do about that. I need to run a wire with two leads and ground to where the drivers will be at.

[petergebruers 1,265]

2 minutes ago, Tiwas said:

There's no warning I can see about overload or overcurrent, which is why I was reluctant to accept that as a solution.

 

Exactly! I like your (healthy) skepticism

 

2 minutes ago, Tiwas said:

I will also try removing the bypass. Even if it was needed for only one driver it might not be needed for two.

 

I give you about 50% chance. Due to manufacturing differences, one power supply may start sooner than the other so one of them might glow faintly or blink (if you had that problem with one load and no bypass).

[Tiwas 0]

2 hours ago, petergebruers said:

I give you about 50% chance. Due to manufacturing differences, one power supply may start sooner than the other so one of them might glow faintly or blink (if you had that problem with one load and no bypass).

That's actually one of the real tests. This LED strip has awesome light (6000lm/m), but is pretty high wattage. Also, it cannot be longer than 15m in one strip or it will cause a voltage drop. If I cannot run more than one transformer off of one dimmer, I'm better off buying the 230V version which doesn't have as high lumen and light spread but can run up to 100m on one length without transformer. If I need 4 transformers+4 dimmers+4 bypasses and the synchronize them with scenes (which will cause delays) to run 50m (with enough light to burn out my eyes :p) or I can go for the 230V version, use only one dimmer+bypass and run it pretty much as long as I feel without voltage drop, with so little heat I can mount it with tape etc etc the choice is pretty clear.

[petergebruers 1,265]

I understand what you are saying...

 

13 minutes ago, Tiwas said:

If I cannot run more than one transformer off of one dimmer, I'm better off buying the 230V version which doesn't have as high lumen and light spread but can run up to 100m on one length without transformer.

 

I am not sure about that. I recommend a test... I've seen those 230V strips and I am not sure about the dimming range. I do not own such a LED strip so I cannot test this. But I have done tests with so called "driverless 230 V COB LED modules" (which actually contain a mosfet driver to dissipate the peaks of the sine wave). They have very narrow dimming range.

[Tiwas 0]

11 minutes ago, petergebruers said:

I understand what you are saying...

 

 

I am not sure about that. I recommend a test... I've seen those 230V strips and I am not sure about the dimming range. I do not own such a LED strip so I cannot test this. But I have done tests with so called "driverless 230 V COB LED modules" (which actually contain a mosfet driver to dissipate the peaks of the sine wave). They have very narrow dimming range.

I bought a 5m length just to test, and the dimming range with just the mechanical dimmer is awful. However, with the dimmer2 it's really nice. There's a point where the dimming isn't really visible, but there's no flickering light or anything and the lowest part is really dim and the highest is really bright. Keeping it in 10% increments was nice and fluid. At least for 5m

[petergebruers 1,265]

That is good news. Thank you for reporting back!

[Tiwas 0]

5 minutes ago, petergebruers said:

That is good news. Thank you for reporting back!

 

This is the 230V version, if you're interested: 

Please login or register to see this link.

 . The inline dimmer is cr*p, but the strip is pretty nice. And the price is definitely good

[Tiwas 0]

3 hours ago, petergebruers said:

That is good news. Thank you for reporting back!

 

Is there a good way to script a gradual up and down dimming from 0-100% I was thinking of making a video of the 230V strip if you'd like to see it.

[petergebruers 1,265]

Try "startLevelIncrease", I haven't used it in a while but I think it still works:

 

Example to dim slowly:

Please login or register to see this code.

Introduced in HC FW 4.080

 

"- New actions in Lua scenes:

    - fibaro:call(ID, 'startLevelIncrease', x, y)
    - fibaro:call(ID, 'startLevelDecrease', x, y)
    - fibaro:call(ID, 'stopLevelChange')

    These calls can be used to start increasing or decreasing Multilevel Switches (like Dimmer 2, RGBW or Roller Shutter 2 modules) value. 'x' and 'y' are optional parameters, used to set respectively time frame in which change should be applied and starting level. 'stopLevelChange' stops previously send action."

[Tiwas 0]

13 hours ago, petergebruers said:

Example to dim slowly:

Please login or register to see this code.

 

I tried this (and several variants), but I get this from luaTest

 

Quote

LuaTest 1.6

Lua file: /etc/cmh-ludl/luatest.lua

Results
Runtime error: Line 2: attempt to index global 'fibaro' (a nil value)

Locals
[main]

Print output
(none)

Code
   1   ID = 597 
   2   fibaro:call(ID, 'startLevelIncrease', 4, 0) 
   3   fibaro:sleep(4000) 
   4   fibaro:call(ID, 'startLevelDecrease', 4)

 

[petergebruers 1,265]

What is Luatest? Are you running this on your PC?

 

The error message suggests you either assigned something to "fibaro" or this is not a HC Lua scene, running on a HC...

 

I run this on my HC, firmware 4.162, as a scene, with no issues

 

[Tiwas 0]

6 hours ago, petergebruers said:

What is Luatest? Are you running this on your PC?

 

The error message suggests you either assigned something to "fibaro" or this is not a HC Lua scene, running on a HC...

 

I run this on my HC, firmware 4.162, as a scene, with no issues

 

Turns out the fibaro library isn't available on Vera, so I need to find another way to do it.

[petergebruers 1,265]

Those fibaro commands are wrappers around standard Z-Wave commands of the class "COMMAND_CLASS_SWITCH_MULTILEVEL" version 2, notably:

 

SWITCH_MULTILEVEL_START_LEVEL_CHANGE

SWITCH_MULTILEVEL_STOP_LEVEL_CHANGE

SWITCH_MULTILEVEL_SET

 

I did a quick search on the vera forum but I could not find anything relevant. Maybe Vera supports them, but uses a different name.