# [PARTLY SOLVED] T5 dimming using 0-10V

Hello all,

I really need help on this one.

I've got an electronic dimmable ballast with 0-10V control, the light rating is from 1% to 100%.

Measuring the voltage on ballast's control PINS gives 10V, so the ballast itself is producing current, the light output is at full power.

I've got a circuit using a LM324N OpAmp: - Receives on PIN 3 and 2 the +5V PWM signal from arduino - Receives on PIN 4 and 11 the +12V - Outputs on PIN 1 0V to +12V

Then the signal from the OpAmp passes trough a v-reg that brings it down to +10V.

Ardunio PWM PIN is at value 0, measuring the v-reg output PIN will give me 0V.

When I connect the ballast's control PIN+ to the v-reg output and PIN- to common circuit ground, the ballast will automatically dim the light to a certain value, but for sure it's more than 1% because the light is still intense.

If I measure the v-reg output PIN I read something like +1.5V, but I still have the arduino PWM value at 0.

I assume that the ballast is injecting 1.5V back into the circuit and that's the reason I cannot full dim down.

If I start increasing the PWM value then the ballast dims up to the full power.

I've got two questions for those HW gurus out there. - (The most important question) Is this the correct way to dim an electronic ballast using the 0-10V control ? - How do I bring that +1.5V to 0 when the ballast is connected ?

Thanks for any help provided.

(The most important question) Is this the correct way to dim an electronic ballast using the 0-10V control ?

No

How do I bring that +1.5V to 0 when the ballast is connected ?

Have a lower impedance driving the input pins. It looks like the input pins need to sink current. Could you post a schematic of the op amp and voltage regulator as from the description it sounds very wrong.

If this is not the correct way, what should it be ?

Removed the faulty schematic.[/edit]

What do you mean by a sink current ?

What do you mean by a sink current

Provide a low impedance to ground.

The voltage regulator on the end is just wrong. This takes a variable input and produces a fixed output. This is exactly what you don't want. What is more when you provide less than the required regulation voltage it no longer acts as a voltage regulator.

Basically ditch the voltage regulator and feed the output of the op amp directly into the base of a transistor. The collector to 12V and the emitter to a 470R resistor to ground. Also connect the resistor to the input pin of your controller.

Maybe I'm thinking this wrong.. but the v-reg in this case is acting as a protection to the ballast, it will not allow more than 10V to pass, since the opamp is feeded by 12V, but will allow less than 10V to pass.

What would be the role of the transistor on the scenario you've described ?

We're talking about a Helvar EL239sc with two pins for the 0-10V control.

I was able to get the following info from Helvar: - The computer must have an analog output 0 - 10 VDC. Buffer should be a "sink"-type buffer. The current should be taken from the HFC-ballast. The HFC-ballast is a current source. - Active light control range is 1 - 10 VDC. These are the limits when the light level is changing.

What do they mean by "buffer" and "sink" ?

This a post back in 2008 and it work. http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1216381158/all

An op-amp is capable of sourcing and sinking so it will work as a "buffer" as you require. I think you want something like this:

I would find out just how much current the T5 is sourcing at its control output. The TLV2371 can only sink a few mA.

-- Check out our new shield: http://www.ruggedcircuits.com/html/gadget_shield.html

Maybe I'm thinking this wrong.. but the v-reg in this case is acting as a protection to the ballast,

Buffer should be a "sink"-type buffer. The current should be taken from the HFC-ballast. The HFC-ballast is a current source.

That says it all, you don't need to supply a voltage to this controller at all you need to sink current from it. Sinking current is draining it to ground.

What would be the role of the transistor on the scenario you've described ?

To act as a low impedance sink. An op amp can only sink a small amount of current, the transistor would act as a buffer to allow the controller to sink current. (a buffer is something that cushions two signals. In this case it allows you to turn a high impedance output into a low impedance one.) The transistor would allow current to flow out of the controller pin until the voltage across the emitter resistor was equal (plus 0.7V) to the voltage on the base, thus it would allow control of the current sink. For a proper design we would have to know what amount of current the controller sinks. That way we could get the emitter resistor the exact right value.

@BillHo: the dimmable circuit was taken from that post. But the user never told if it really worked nor told about the voltage injection from the ballast.

@Richard: You may find all the information I've got on this page.

@RuggedCircuits: The difference I see on your schematic is the C2 capacitor, what is the objective of that capacitor on the circuit ?

is the C2 capacitor, what is the objective of that capacitor on the circuit ?

Simply supply decoupling.

For a proper design we would have to know what amount of current the controller sinks. That way we could get the emitter resistor the exact right value.

I can do a real world measurement later today, but I think I've read somewhere that the control current is at max 2mA.

@Grumpy_Mike: So basically will reduce the ripple effect ?

Well ripple is not quite right but close. It is noise on the power lines that make chips not work correctly. See:- http://www.thebox.myzen.co.uk/Tutorial/De-coupling.html

If it is 2mA then that is fine to sink with a simple op amp, although the transistor with 470R emitter resistor is what I would use.

Should it be something like this ?

Removed the faulty schematic.[/edit]

No.

You didn’t say you had two control pins.

Put Ballast C0 to ground. Put Blast C1 to the transistor emitter. Put the transistor collector to +12V.

Removed the faulty schematic.[/edit]

• Is the resistor 470R still required ?
• The transistor must be any "special" one ?
• The transistor should be PNP, NPN or it really doesn't matter ?

Is the resistor 470R still required ?

yes yse yse yse yse yse yes. Otherwise how the chuff would it work.

An ordinary NPN transistor.

Thanks for the patience and also for the contribution, but I’m a newb at electronics.

Did I make it this time ?

I’m now wondering the working principle, after reading the Wikipedia article about transistors does the following sentence descrives it ?

“In a grounded-emitter transistor circuit, (…), as the base voltage rises the base and collector current rise exponentially, and the collector voltage drops because of the collector load resistor.”

From my understanding, when the opamp signal starts rising on the base, the ballast control voltage (+10V when base=0V) connected to the collector will start to drop.

Although in dutch, but the schematic speaks for itself: http://aartsite.nl/licht/o_5vhrfi.htm

Greetings, EriSan500

@Erisan, @Richard: Thanks for the contribution ! So for this solution the LM324 VCC and VDD is not fed at all ?

I don’t know if I have a BC547 at home but I’ll check, I also give this one a try and post back feedback.

The +12V power source is not a problem to me, because I’m regulating from +12V to +5V to feed the arduino… so the +12V line is there anyway.