0
\$\begingroup\$

I'm revisiting an old project, and I want to see if there is anything I could have done differently, or if I missed any important details or considerations.

Basically, I used a constant current driver (designed for LEDs) to power and dim a halogen bulb. The driver is a MeanWell HLG-100H-20AB, and the bulb is 35W, 12V halogen. This constant current driver can be controlled via variable resistance (10kOhm at min. brightness and 100kOhm at max. brightness), and its max current output can be adjusted as low as 3A. I calculated that the bulb needs 2.917A to shine at its maximum, so I set the max current on the MeanWell to ~3A and arranged my potentiometer and resistors such that the max resistance delivered would be ~97.5kOhm, meaning, theoretically, that the Meanwell unit should not output more than 2.917A.

Here are my questions:

  1. This worked well from what I saw. The light dimmed as I expected, and the bulb didn't burst. But, I'm wondering if there's anything crucial that I could have missed. I don't generally see people dimming halogen or incandescent bulbs with constant current devices, and there must be a reason for this. Any ideas?
  2. I'm assuming that if I had used the MeanWell to output, lets say 4A, the bulb would have failed, since my understanding is that the MeanWell unit is capable of "force feeding" current. Is this actually true? Or, does the unit just modulate voltage to indirectly control current?
  3. I noticed that the spec sheet has a "constant current region" listed between 10V and 20V. I don't understand what this means, or how it would affect the halogen bulb. For instance, when running only 1.5A to the bulb, for example, its a safe bet that the voltage across the bulb's termninals is less than 10V. Yet, the MeanWell unit still supplied steady current. What gives?
  4. Am I correct in thinking that this setup negates voltage drop considerations for the bulb over a long cable?

Thanks!

\$\endgroup\$
7
  • 1
    \$\begingroup\$ There is no way to "force-feed" current other than by "modulating the voltage". \$\endgroup\$ – Eugene Sh. Dec 18 '20 at 18:49
  • 1
    \$\begingroup\$ @EugeneSh.,...and that is exactly what a constant current supply does. If you connect a two-terminal device (e.g., a halogen bulb) to an ideal constant-current supply, then the voltage across the terminals will be whatever Voltage it takes to drive the set amount of current through the bulb. \$\endgroup\$ – Solomon Slow Dec 18 '20 at 19:01
  • 1
    \$\begingroup\$ @SolomonSlow What is ideal current source? There is no such a thing. Real current sources have a control loop which is monitoring the current and increasing/decreasing the output voltage to keep it at target setting. \$\endgroup\$ – Eugene Sh. Dec 18 '20 at 19:02
  • 1
    \$\begingroup\$ Don't know what the 10V lower limit means, but the 20V upper limit is easy enough to understand. As, @EugeneSh. said, "ideal" current sources only exist on paper. Real current sources have limitations. The 20V upper limit means, that if it takes more than 20V to drive the set amount of current through the load, then the constant-current circuit will fail to meet the goal. One possibility is that the voltage will be clipped to 20V, and the current will be less than the set amount. That's how the bench power supply on my desk behaves. Another possiblity is that the power supply shuts down. \$\endgroup\$ – Solomon Slow Dec 18 '20 at 19:09
  • 1
    \$\begingroup\$ It's the voltage at the output of the power supply that's going to be limited to 20V. The Voltage at the other end of the "long cables" will be less. But yes, If 20V is enough to drive 3A through the cables and the bulb, then you'll be guaranteed to get 3A. \$\endgroup\$ – Solomon Slow Dec 18 '20 at 19:18
2
\$\begingroup\$

I don't generally see people dimming halogen or incandescent bulbs with constant current devices, and there must be a reason for this. Any ideas?

Price. It is much easier to power a bulb with a battery, mains, transformer.

Or, does the unit just modulate voltage to indirectly control current?

Correctly, almost all current sources do it that way. Practically, there is no such a device like intrinsic current source, just theoretical. But it could be true somewhere in the universe, so we're ready to use our equations.

Am I correct in thinking that this setup negates voltage drop considerations for the bulb over a long cable?

Yes.

\$\endgroup\$
1
  • \$\begingroup\$ Re, "Modulate voltage to...control current" Typical residential "dimmer switches" control the intensity of an incandescent bulb by using inexpensive circuits based on Triacs or SCRs to "chop" the AC waveform. Such dimmers may or may not be capable of reducing the brightness of the bulb all the way down to zero. en.wikipedia.org/wiki/Dimmer#Solid-state_dimmer \$\endgroup\$ – Solomon Slow Dec 18 '20 at 19:12
1
\$\begingroup\$

I'm assuming that if I had used the MeanWell to output, lets say 4A, the bulb would have failed,

Most likely.

I noticed that the spec sheet has a "constant current region" listed between 10V and 20V. I don't understand what this means, or how it would affect the halogen bulb. For instance, when running only 1.5A to the bulb, for example, its a safe bet that the voltage across the bulb's termninals is less than 10V. Yet, the MeanWell unit still supplied steady current. What gives?

The unit is designed to provide constant current if the voltages do not need to go outside of that range to do that. If you tried to drive a 1M\$\Omega\$ resistance with your unit set to 3A, the voltage would probably rise to 20V, or somewhat above and give less than 3A current. If you short circuited the outputs of your unit, the voltage would probably drop somewhere below 10V, and the output current would be something we cannot determine from that spec.

Am I correct in thinking that this setup negates voltage drop considerations for the bulb over a long cable?

Yes.

\$\endgroup\$
1
\$\begingroup\$

What you did seems fine, but your fear is well-founded.

Tungsten (and other wire) filaments have a positive temperature coefficient of resistance. Running them at constant current may not quickly achieve a stable final temperature, power or brightness.

Instead, the filament temperature and lamp brightness may continue to rise for a long time in a positive feedback loop. To be safe, it is better to use a voltage supply rather than a current supply. This will insure that the system rapidly converges on the final steady state level. The final filament temperature is relatively insensitive to voltage, and relatively hyper-reactive to current.

It is probably safe to use a current source as long as the current is low enough (as you did) OR if the supply also has a safe voltage limit so that it can never go over the rated voltage.

\$\endgroup\$
0
\$\begingroup\$

Nothing wrong to power incandescent bulb from that power supply. Just set max voltage to 12V. In CC mode setting current as in normal condition may cause only slow start. Cold filament has low resistance, it rises when heats. It even improve life span of bulb. Dimming circuit with PWM is OK. Resistance of wires should be taken in consideration, but if it is known value, you can rise output voltage to compensate wires drop voltage. For fire safety current should not exceed max current for wire gage.

\$\endgroup\$
1
  • \$\begingroup\$ @solomonslow. Theory is good when applied to ideal objects. Power supply will be in CC mode for start and going to CV in continuos. Can you describe what do you mean by bad positive feedback. The circuit gonna oscillate? What frequency? You comment need more clarification. \$\endgroup\$ – user263983 Dec 21 '20 at 16:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.