Retrofit LED project 0-10vdc dimming questions

[gwright]

Hello lighting pros,

I am an in-house electrician at a convention center. We are replacing our inefficient and costly-to-maintain 750 watt pulse start metal halide high bay lighting with LED. Dimming system will be 0-10vdc. We had hoped that dealing directly with the manufacturer would provide us quick, knowledgeable responses to our questions, instead we get a cut sheets and promises that "Everything will work out fine", "Don't worry so much", and "Someone will get back to you with an answer". Hoping someone here can let us know if we are on the right track, as well as answer a few questions.

277v
400 fixtures
Run lengths do not exceed 300' between drivers and dimmers

We started running control wires in same conduit as supply conductors, then realized we were asking for trouble doing it that way due to noise/interference and possibility of inducing voltage in control wires. We started over and will run the wire outside the conduit. Planning to use 16/2 stranded/shielded cable. Should it be Plenum-rated? Since we are not running the wire inside the conduit, can we use 300v rated cable or better to use 600v? Expect to pay about $.40/ft?

Dimming question. The drivers are 4 wire. We are using 0-10vdc. My understanding is this: The controller sends 10v in order to get 100% light. If the dimmer control is used to 'request' lighting at 20%, the controller would pull down the dc voltage to from 10vdc to 2vdc. Switching lights off would require the controller to sink the voltage to the 'low-end cut off' level, or just short the wires to get zero voltage. Correct so far? Last thing - If zero voltage on the signal wires is required to turn the lights off, how can the lights work at 100% when the dimmer wires from the driver are disconnected? Isn't that zero voltage?

Sorry this is so long. I just don't want to be the guy remembered for screwing up the $500k lighting project because he didn't ask a question. Many thanks!

 

[gar]

170404-0951 EDT

gwright:

As an answer to your question on 0-10 V control.

I conjecture that the LED driver has a common line and a signal line for the 0-10 control. These are probably totally isolated from the AC power to the driver, from the EGC, and real earth by at least a several thousand volt insulation breakdown value.

I believe that internally there is a 10 V DC voltage source with a resistor of possibly 10 k between the 10 V source and the signal input terminal. Probably the negative side of the 10 V source is connected to the common input terminal. From the signal terminal to common is a voltage sensor that controls the lamp brightness. Connect a high impedance meter between signal and common with nothing else connected and you should read about 10 V.

With nothing connected between signal and common the voltage is 10 V. Thus, full brightness. With signal to common shorted the signal voltage is 0 and 1 mA of current flows thru the short. Put a 10 k resistor from signal to common and 1/2 mA flows and the signal voltage is 5 V relative to common, and light brightness is whatever is produced for a 5 V signal input.

Thus, one could simply use and an adjustable resistor for brightness control. But better control is achieved by an adjustable voltage source of 0 to 10 V applied to signal and common. The source impedance of this adjustable voltage needs to be small compared to the internal impedance of signal to common. For my above suggestion of 10 k for the input impedance of the signal to common terminals an adjustable voltage source of less than 100 ohms is probably adequate.

On the subject of noise into the signal terminal. An RC low pass filter at the signal to common terminals might eliminate the need for shielded input wiring. No matter what --- you want a twisted pair for the control wiring. A twisted wire pair reduces magnetic coupling of noise. What it really does is cancel magnetically induced voltage from one twist to the next twist.

.

 

[Fnewman]

The 30 W LED floodlights I just installed have 4 wire 'drivers' (looks like a small ballast). Two deliver 120 v to the driver while the other two control the current to the LED. In my case, 10 v is 700ma = 100 output while minimum on the slide seems to reduce output to about 10%. Luton dimmer was used.

 

[gwright]

170404-0951 EDT

gwright:

As an answer to your question on 0-10 V control.

I conjecture that the LED driver has a common line and a signal line for the 0-10 control. These are probably totally isolated from the AC power to the driver, from the EGC, and real earth by at least a several thousand volt insulation breakdown value.

I believe that internally there is a 10 V DC voltage source with a resistor of possibly 10 k between the 10 V source and the signal input terminal. Probably the negative side of the 10 V source is connected to the common input terminal. From the signal terminal to common is a voltage sensor that controls the lamp brightness. Connect a high impedance meter between signal and common with nothing else connected and you should read about 10 V.

With nothing connected between signal and common the voltage is 10 V. Thus, full brightness. With signal to common shorted the signal voltage is 0 and 1 mA of current flows thru the short. Put a 10 k resistor from signal to common and 1/2 mA flows and the signal voltage is 5 V relative to common, and light brightness is whatever is produced for a 5 V signal input.

Thus, one could simply use and an adjustable resistor for brightness control. But better control is achieved by an adjustable voltage source of 0 to 10 V applied to signal and common. The source impedance of this adjustable voltage needs to be small compared to the internal impedance of signal to common. For my above suggestion of 10 k for the input impedance of the signal to common terminals an adjustable voltage source of less than 100 ohms is probably adequate.

On the subject of noise into the signal terminal. An RC low pass filter at the signal to common terminals might eliminate the need for shielded input wiring. No matter what --- you want a twisted pair for the control wiring. A twisted wire pair reduces magnetic coupling of noise. What it really does is cancel magnetically induced voltage from one twist to the next twist.

.

Much appreciation for your response. Very helpful. Completely understand your points.

 

[gwright]

The 30 W LED floodlights I just installed have 4 wire 'drivers' (looks like a small ballast). Two deliver 120 v to the driver while the other two control the current to the LED. In my case, 10 v is 700ma = 100 output while minimum on the slide seems to reduce output to about 10%. Luton dimmer was used.

Thanks for responding.

 

[Fulthrotl]

Hello lighting pros,

I am an in-house electrician at a convention center. We are replacing our inefficient and costly-to-maintain 750 watt pulse start metal halide high bay lighting with LED. Dimming system will be 0-10vdc. We had hoped that dealing directly with the manufacturer would provide us quick, knowledgeable responses to our questions, instead we get a cut sheets and promises that "Everything will work out fine", "Don't worry so much", and "Someone will get back to you with an answer". Hoping someone here can let us know if we are on the right track, as well as answer a few questions.

277v
400 fixtures
Run lengths do not exceed 300' between drivers and dimmers

dunno where you are at in this process, however.... there are some
considerations. you may want a level of control beyond 0-10.

know how you sometimes have to pull a lamp out above a booth,
and charge the exhibitor for it? 35 years ago, it was $150. god only
knows now. you can't do that with 0-10 and led very well, unless
you pull a cord connection out, and you still have to get up into the
overhead.

something like nLight would allow full control of each fixture. you
could do anything with controls you wanted to.

now, is LB CC permitting this? do you have title 24 compliance to
consider?

a five minute phone call might bring up some stuff worth considering.
PM sent.

 

[ronaldrc]

dunno where you are at in this process, however.... there are some
considerations. you may want a level of control beyond 0-10.

know how you sometimes have to pull a lamp out above a booth,
and charge the exhibitor for it? 35 years ago, it was $150. god only
knows now. you can't do that with 0-10 and led very well, unless
you pull a cord connection out, and you still have to get up into the
overhead.

something like nLight would allow full control of each fixture. you
could do anything with controls you wanted to.

now, is LB CC permitting this? do you have title 24 compliance to
consider?

a five minute phone call might bring up some stuff worth considering.
PM sent.

Do they make dimmers for 10 volts?

Ronald