LED MR-16 Replacement Dimming issues

Status
Not open for further replies.

Matt N

Member
Location
Alameda, CA USA
I've put Green Creative LED MR-16 replacements in some line-voltage track lights. Surprisingly, the 50-watt transformer does power up the 8-watt lamp, but dimming is a real bummer (DVLV no, DVCL no, DVELV no). Has anyone tried this and come up with a solution?

Thanks in advance.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
180518-2050 EDT

Matt N:

Tell me more about these bulbs. I quickly looked on line. No useful information.

Apparently these are 12 V. Note 10 W at 12 V is about 0.8 A into a resistor, AC or DC.

What does 12 V mean? Can you apply a 12 V RMS sinewave and not damage the bulb? This will relate to LED PIV (peak inverse voltage rating), and what exists between the bulb terminals and the actual LEDs.

Does the bulb only conduct in one direction? Meaning it probably should be powered from a DC source.

Internally does the bulb have a bridge rectifier? If so, then the internal LEDs (LED chips are inherently a diode, that is, a unidirectional conductor) could be effectively run from either an AC or DC source.

If we assume nothing but diodes and resistors are in the bulb, then varying the average DC current thru the bulb will give you good dimming.

Strip LED lighting is basically resistors and LEDs. Most designed for 12 V DC operation.

An Armacost LED Dimmer 12-24 V 8 A DC only from Home Depot is basically an adjustable pulse width on-off modulator that has a several thousand Hz switching frequency. Correct polarity must be observed.

What is a 50 W transformer? An iron core with two coils, or some electronic device? Is it a 60 Hz sinewave voltage source, a DC power supply, or some kind of modulated waveform?

.

.
 

Matt N

Member
Location
Alameda, CA USA
180518-2050 EDT

Matt N:

Tell me more about these bulbs. I quickly looked on line. No useful information.

Apparently these are 12 V. Note 10 W at 12 V is about 0.8 A into a resistor, AC or DC.

What does 12 V mean? Can you apply a 12 V RMS sinewave and not damage the bulb? This will relate to LED PIV (peak inverse voltage rating), and what exists between the bulb terminals and the actual LEDs.

Does the bulb only conduct in one direction? Meaning it probably should be powered from a DC source.

Internally does the bulb have a bridge rectifier? If so, then the internal LEDs (LED chips are inherently a diode, that is, a unidirectional conductor) could be effectively run from either an AC or DC source.

If we assume nothing but diodes and resistors are in the bulb, then varying the average DC current thru the bulb will give you good dimming.

Strip LED lighting is basically resistors and LEDs. Most designed for 12 V DC operation.

An Armacost LED Dimmer 12-24 V 8 A DC only from Home Depot is basically an adjustable pulse width on-off modulator that has a several thousand Hz switching frequency. Correct polarity must be observed.

What is a 50 W transformer? An iron core with two coils, or some electronic device? Is it a 60 Hz sinewave voltage source, a DC power supply, or some kind of modulated waveform?

.

.I regret to tell you that I don't have any knowledge to add to the engineering of these LED MR-16 lamps. I can say the existing Nora track lights have a 50-watt DC electronic integral transformer. Thank you for suggesting the Armacost Dimmer; I will certainly check it out and see if that works.
 

ActionDave

Chief Moderator
Staff member
Location
Durango, CO, 10 h 20 min from the winged horses.
Occupation
Licensed Electrician
I've put Green Creative LED MR-16 replacements in some line-voltage track lights. Surprisingly, the 50-watt transformer does power up the 8-watt lamp, but dimming is a real bummer (DVLV no, DVCL no, DVELV no). Has anyone tried this and come up with a solution?

Thanks in advance.

I can't give you any good solution, but I can tell you from similar experiences that dimming leds with anything other than 0-10V fixtures is journey into madness and fast track to an early grave.

Get rid of the low voltage track fixtures and your odds of getting some sort of dimming that works will improve. But then again finding somebody that wants to spend the $800 before labor charges for all new track heads and led lamps is rarer than a live unicorn.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
180518-2453 EDT

Matt N:

10 W at 10 V is ballpark 10 ohms. Putting 100 ohms in series with your bulb, and connecting this series combination to a 12 V battery may be safe. If the bulb is polarity sensitive, then it would light with one polarity and not the other.

A 15 W incandescent might be in the range of 100 ohms at 12 V. You could experiment and find out. A 100 W bulb at 0 V is about 10 ohms, and about 144 ohms at 120 V.

Once you know if the LED is polarity sensitive, then you could try it with the Armacost dimmer and any reasonable 12 V battery. This dimmer requires correct polarity.

If the LED light is polarity sensitive, then I would expect some sort of keying of the bulb to get correct orientation. That would also mean the track is identified in some way, and a DC power supply has to be connected correctly.

.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
180518-2549 EDT

A 100 W 120 V incandescent at 32 V is about 0.3 A or 96 ohms. At 16 V about 0.2 A or 32 ohms. So possibly try a 25 W.

ActionDave:

DC control of current to a DC LED gives great dimming control. 0 to 100 %, and smooth, and done correctly no vissible flicker.

.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
180522-2423 EDT

Matt N:

I picked up a packazge of GE MR-16s today.

These are not polarity sensitive. I can apply a DC voltage of either polarity and all 4 LEDs in the bulb light light with either polarity. Threshold of turn on is around 4 to 5 V DC.

Somr likely initial implications:
1. A bridge rectifier exists at the input.
2. Two LED chips are in series with some current limiter, probably at least a linear resistor.
3. There is more than just a linear resistor in series with the LEDs because one has to exceed the dropout threshold to get illumination started.
4. Voltage adjustment (current) changes brightmess.

I have run no other experiments yet.

.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
180523-1146 EDT

Matt N:

Doing some more playing with the GE MR-16 I get very unsatisfactory results.

The package says dimmable. With my tests so far not very dimmable.

With AC sine wave input the light pulls in at about 3.1 V and drops out about 2.3 V. These are peak voltages of 4.4 and 3.3 V. At time of pull in and with no input voltage change there is a brightness change over a short time.

With DC from a very large filter capacitor 3.1 and 2.7 for pull in and drop out.

Close to full brightness is around 8 to 9 V.

The Armacost dimmer doesen't work real well. It appears there may be a capacitor in the MR-16.

I don't know if phase controlled drive near 60 Hz would produce better results.

The bulb does not look good for dimming. Nothing comparable to variable DC current or pulse width modulation to a 12 V LED DC strip light.

This is not close to a simple resistor for current limiting in series with the LED.

.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
180523-2210 EDT

Matt N:

I did some more experiments with the GE MR-16.

This time I did pulse width modulation at a much lower frequency, 120 Hz.

Yes, to some extent I can adjust intensity. But I really don't consider this a satisfactory dimmable bulb. To label it dimmable is misleading in my opinion. Nothing comparable to what you can do with an incandescent, a plain LED chip, 12 V LED strip lights, or some CREE 120 V LED bulbs.

I don't know what kind of dimmer you had tried, and how it works, but I think my varied tests probably cover (simulate) the different dimmer drive modes that might be used.

If your existing dimmer works well with incandescents, and dimming is important, then I would continue using incandescents.

.
 

ELA

Senior Member
Occupation
Electrical Test Engineer
Maybe like this?

Maybe like this?

Sorry picture not the best
 

Attachments

  • tranny.JPG
    tranny.JPG
    29.7 KB · Views: 0
Last edited:

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
180524-2345 EDT

ELA:

What is the device? What is the output? What does the output look like with different types of loads? With a phase shift dimmer at the input what does the output look like? Same with a Variac input?

.
 

ELA

Senior Member
Occupation
Electrical Test Engineer
Electronic Transformer/Converter

Electronic Transformer/Converter

Gar,
As I have heard you say to others, " you should to some experiments to learn".
Get yourself an Electronic Transformer specifically designed to drive this LED bulb and answer the questions you seek answers to.

I imagine there are many flavors of Electronic Transformers. One I worked with, too long ago to recall exactly, was a switching supply that converted the 120Vac input to a variable DC current source or PWM. Designed specifically to track the phase angle of the dimmed input in order to be compatible with existing phase fired/cut dimmers. Not the absolute best method to dim LED but compatible with more common or existing dimmers.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
180525-2431 EDT

ELA:

I agree with your comment. However, there is a limit to how many different things I can experiment with that are of no direct concern to me. Right now I am hung up enough on the GE MR-16 bulb.

Very few posters on this site do any experiments themselves to determine the fundamental operating characteristics of things they ask about.

My experiments with the GE MR-16 to date do not indicate that it is very dimmable. So far it looks like this bulb uses the peak input voltage, so long as it recurs fast enough, that is above some threshold, about 6 V, to supply power to the bulb and limitedly control the bulb's intensity by the peak value of the voltage magnitude.

Below a fairly high threshold level there is no light output.

Pulse input duty cycle seems to have little affect on light output.

To date I would not classify this bulb as dimmable. Rather it has slight dimming characteristics.

What is dimmable needs to have a clear definition.

.
 

ELA

Senior Member
Occupation
Electrical Test Engineer
Hi Gar,
I agree that different peoples interest in different things are greatly varied. Many people here are more interested in getting the job done in order to move on to the next job. Others with time to spare may want to experiment driven by a need to know.

Here is some information that may help if you want explore further.

https://www.maximintegrated.com/en/app-notes/index.mvp/id/5372
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
180526-1009 EDT

I tried to look up information on the Osram ET-MZ 60/110-130 shown by ELA.

Anything I could find was basically useless garbage.

Relative to the low voltage LED bulbs it seems they are generally rated 12 V AC/DC. One bulb was listed 12 V AC and 20 V DC.

Lots of so called compatibility charts, but do they really mean anything. One has to understand the basic operation of the devices, then it is possible to make some reasonable judgements as to what will really work with what. Lots of claims on dimmability, but no definitions.

Any LED light bulb or strip that works on DC, has no internal chopping, and basically uses LEDs, resistors, and diodes internally will produce no RFI, flicker, or audio noise when supplied from a pure DC adjustable current source (can be an appropriate voltage source), and can be dimmed from 0 to 100%.

The GE MR-16 is not such a bulb.

What is the Osram ET-MZ 60/110-130? I do not know. And I did not find a useful description.

There are descriptions of parallel and series wiring. Clearly wrong. Parallel as used meant a star wiring topology. Yes that is parallel. Series as used meant a linear bus distribution means, but is really a parallel electrical circuit, and not a series circuit.

.

.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
180528-0854 EDT

ELA:

I agree that different peoples interest in different things are greatly varied. Many people here are more interested in getting the job done in order to move on to the next job. Others with time to spare may want to experiment driven by a need to know.

The problem presented in post #1 related to getting the job done.

I've put Green Creative LED MR-16 replacements in some line-voltage track lights. Surprisingly, the 50-watt transformer does power up the 8-watt lamp, but dimming is a real bummer (DVLV no, DVCL no, DVELV no). Has anyone tried this and come up with a solution?
I don't know what this statement means.

I believe it means the track is a 120 V 60 Hz power source, a 120 V dimmer of some sort feeds the track, and that there are fixtures in the track with a built in 50 W transformer (what kind of transformer?) (what output voltage?) (what kind of voltage, AC or DC?) (if AC what frequency?) (with some defined dimmer, and an incandescent lamp load, what varies at bulb terminals, and what is the dimming range?).

Stated dimmers don't work well with the Green Creative LED MR-16 bulb.

Data sheet for said LED is at
http://gc-lighting.com/wp-content/uploads/GREEN-CREATIVE-LED-CRISP-SERIES-MR16-7W-High-CRI.pdf
States 12 V 60 Hz. Nothing about means to dim, voltage or phase shift. Or whether it works on DC. Probably works on DC but dimming may not be good.

Dimmer compatibility chart is possibly useless, but is at
http://gc-lighting.com/wp-content/u...MR16_7W_High-CRI_LED_Dimmer_compatibility.pdf

What happens if a 60 Hz voltage source adjustable from 0 to 12 V is applied directly to the LED bulb? No discussion. This is an easy experiment. Use a Variac into a 120 to 12 V ordinary iron core transformer to the LED bulb. I have no idea what will happen. The GE MR-16 has a very poor dimming range. On DC 100 down to 80 %. AC is not much different.

Instead use a suitable phase shift dimmer at the said transformer input. Means it must be designed to work with an iron core transformer load. In turn this means essentially no DC component into the transformer primary, otherwise core saturation occurs. Now what is the satisfactory dimming range?

I can not assume the GE MR-16 works anything like the Green Creative LED MR-16.

The Maxim chip you referenced appears primarily to be designed to make the electronic transformer think there is a heavier load than what is present.

In a different datasheet I found a driver chip for use in electronic transformers that produces an output frequency of about 35 kHz. So this is possibly what might be a typical AC output frequency.

.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
180528-1720 EDT

Matt N:

Connect one of your MR-16 bulbs to your car battery. The bulb should light.

Assuming the bulb worked, which it should. Then get 8 D cells. These can provide a variable DC voltage source in 1.5 V increments. See how the bulb works by varying the voltage. I will guess drop out between 3 and 8 volts.

If you have a Variac, then you could power a 12 V bell transformer from the Variac and adjust the AC output voltage to the bulb and see how well it dims.

.
 
Status
Not open for further replies.
Top