Constant Current Driver

[wquiles]

The LED lamp unit does have a driver unit that connects to the LED unit itself. So there are basically 2 parts, one is the LED unit itself which consists of the 4 LED's wired in parallel and the other part is the driver to power it which has a label saying 8W 12 on it. But what i've been trying to do is under-drive the LED's because the original driver that comes with the LED unit makes the LED's too bright and they do warm up a little. I want them much dimmer. So thats why I bought a 280mA constant current driver.

So you are basically replacing the circuit that came with this kit with a different circuit? Do you know what current was being fed to the LED's with the original circuit?

How does the driver detect if it's parallel or series and give it either 3V or 12V? Just wondering my led's are in parallel and the driver should i'm assuming give no more than 4v max whereas if it was in series it should give 12v.

Usually there is a small-value, current sense resistor, either inside the driver IC, or external to the driver IC. Each driver does this a little bit different, but they have a feedback loop in which they:
- measure current on that sense resistor - basically a voltage is generated as current goes through a known, high precision resistor (most of the time with a value less than 1ohm, like 0.1 to 0.01 ohms).
- if the current is still too low, it increases the voltage
- measure current again, keep adjusting the output voltage
- if the current is too high, lower the output voltage (there should be some hysteresis as well)
- etc.

Some drivers also have provisions for open circuit on the output, short circuit on the output, thermal overload, etc..

Will

 

[john2k]

So you are basically replacing the circuit that came with this kit with a different circuit? Do you know what current was being fed to the LED's with the original circuit?

Yeap that's exactly what i am doing. all i know is that the original circuit has a label that says output 12V 8W nothing else.

they have a feedback loop in which they:
- measure current on that sense resistor - basically a voltage is generated as current goes through a known, high precision resistor (most of the time with a value less than 1ohm, like 0.1 to 0.01 ohms).
- if the current is still too low, it increases the voltage
- measure current again, keep adjusting the output voltage
- if the current is too high, lower the output voltage (there should be some hysteresis as well)
- etc.

So from what you explain above it seems like the as long as the constant current output on this new driver does not exceed the max current the led's are rated for (in this case 8W which will be over 600mA) then technically it should be fine because the driver will slowly raise the voltage until the desired current is met in this case the desired current for the LED driver is 280mA. But can 280mA not be delivered in many different voltages? like for example lets say the current reaches it 280mA and the voltage gets to 3V, and still continues to climb up but the current stays at 280mA, will the LED technically speaking still be ok?

 

[john2k]

I have just come across an article that says the following:

If you are planning on using the driver to power an individual LED, or to run numerous LED's in series, you would need to ensure that you get a constant current driver.

If however you are planning to run multiple LED's in parallel, you would require a constant voltage driver.

Does this mean I am using the wrong type of driver all this time ? :hairpull:

 

[alpg88]

that is wrong, to drive leds you need cc driver, period, burn that article.

 

[wquiles]

Yeap that's exactly what i am doing. all i know is that the original circuit has a label that says output 12V 8W nothing else.

Do you know if the new LED driver is rated/designed to work in auto/car environments? Voltage spikes, temperature, etc.?

So from what you explain above it seems like the as long as the constant current output on this new driver does not exceed the max current the led's are rated for (in this case 8W which will be over 600mA) then technically it should be fine because the driver will slowly raise the voltage until the desired current is met in this case the desired current for the LED driver is 280mA.

The constant current LED driver does not care nor knows what is the current rating on the LED's. It only knows to try to regulate current by varying the output voltage, until the current is about 280mA, or until an internal limit is hit, or the device over-heats, etc.. You can actually use that LED driver, connect it to a 5mm LED (which a rated current of about 20mA), and pretty much kill it instantly. OR, you can use that driver with an SST-90 (current rating of about 9Amps), and the LED driver will simply under-drive the LED and provide about 280mA - even if the LED is rated to 9Amps.

But can 280mA not be delivered in many different voltages? like for example lets say the current reaches it 280mA and the voltage gets to 3V, and still continues to climb up but the current stays at 280mA, will the LED technically speaking still be ok?

No, that would be violating Ohms Law:

Voltage = Current * Resistance

In the case of the LED, the LED represents a variable resistor (as its vf curve varies with the current and temperature), but the law still holds true.

I have just come across an article that says the following:
---Quote---
If you are planning on using the driver to power an individual LED, or to run numerous LED's in series, you would need to ensure that you get a constant current driver.

If however you are planning to run multiple LED's in parallel, you would require a constant voltage driver.
---End Quote---
Does this mean I am using the wrong type of driver all this time ? :hairpull:

In general, and taking those sentences out of context, without knowing more about the article and other information, I would say the article is wrong. LED's need a current regulated supply, regardless of how they are wired. Period.

But the real problem is really driving LED's in parallel: It is generally not a good idea, as the vf of the LED's is usually not the same, so the current flowing through each LED will be different, even as the voltage is held constant across all LED's. The way around this problem is to test and match the vf's of the LED's as best as possible, so that they draw about the same current (which in the case of 4x LED's, each LED would draw about 1/4 of the current).

Will

 

[alpg88]

what i,m confused about, is why your leds are in parallel, your driver can drive them just fine in series, and they'll be brighter too.
unless they are soldered to a heatsink\star by heatpad and negative(or positive, doesn't matter really) common terminal, but in that case why it is marked 12v280ma??? you will burn them if you put 12v to them.
was the light new in the box when you got it? or it was used, and someone might rewire or put different leds???

 

[wquiles]

what i,m confused about, is why your leds are in parallel, your driver can drive them just fine in series, and they'll be brighter too.
unless they are soldered to a heatsink\star by heatpad and negative(or positive, doesn't matter really) terminals, but in that case why it is marked 12v280ma??? you will burn them if you put 12v to them.
was the light new in the box when you got it? or it was used, and someone might rewire or put different leds???

+1

Same thoughts/worries that I have. Something is not quite adding up

Will

 

[john2k]

what i,m confused about, is why your leds are in parallel, your driver can drive them just fine in series, and they'll be brighter too.
unless they are soldered to a heatsink\star by heatpad and negative(or positive, doesn't matter really) common terminal, but in that case why it is marked 12v280ma??? you will burn them if you put 12v to them.
was the light new in the box when you got it? or it was used, and someone might rewire or put different leds???

no they are brand new phillips brand, the driver that comes with the unit is not marked as 12v 280mA its marked as 12v 8W but i am trying to under-drive the unit with my 12v 280mA driver

 

[john2k]

Will, I think I am now understanding a bit better. I've got a few of those 280mA led drivers in my car and a multimeter so had some time to pop into a local electronics store nearby and bought myself a few 1W LEDs to do some testing with my multimeter and the driver. I carried out the following tests:

First I wired the 3 led's in series, connected them via my 280mA driver and added the multimeter in the current line to measure the current draw, it read exactly 0.28 (280mA). I then removed the multimeter so that while the 3 LED's were lit I can measure the voltage and the voltage read just above 9V so thats around 3V drop across the 3 LED's.

Then I decided to remove 1 LED from the series so that 2 were wired in series and then the voltage read 6V which means the driver is correctly adjusting the voltage. And because it's wired in series the voltage is adding up per LED.

I then decided to wire the 3 LED's in parallel and connect them via the driver. And first i tested the current draw and it was also exactly 0.28 (280mA) I then measured the voltage while the 3 LED's were lit in parallel and the voltage read 3.2V. So i think this test basically answers my questions.

So it seems that with this 280mA driver my parallel wired LED unit will get 280mA with a voltage of 3.2V because it's in parallel, but if the exact same unit was to be modified into series then the same driver would deliver 280mA with 9V.

So I am assuming that my concern about the fluctuating input voltage from the car power source will not be a problem because firstly it's within the drivers input range and secondly the driver will only output the voltage as tested above?

 

[wquiles]

Will, I think I am now understanding a bit better. I've got a few of those 280mA led drivers in my car and a multimeter so had some time to pop into a local electronics store nearby and bought myself a few 1W LEDs to do some testing with my multimeter and the driver. I carried out the following tests:

First I wired the 3 led's in series, connected them via my 280mA driver and added the multimeter in the current line to measure the current draw, it read exactly 0.28 (280mA). I then removed the multimeter so that while the 3 LED's were lit I can measure the voltage and the voltage read just above 9V so thats around 3V drop across the 3 LED's.

Then I decided to remove 1 LED from the series so that 2 were wired in series and then the voltage read 6V which means the driver is correctly adjusting the voltage. And because it's wired in series the voltage is adding up per LED.

I then decided to wire the 3 LED's in parallel and connect them via the driver. And first i tested the current draw and it was also exactly 0.28 (280mA) I then measured the voltage while the 3 LED's were lit in parallel and the voltage read 3.2V. So i think this test basically answers my questions.

So it seems that with this 280mA driver my parallel wired LED unit will get 280mA with a voltage of 3.2V because it's in parallel, but if the exact same unit was to be modified into series then the same driver would deliver 280mA with 9V.

So I am assuming that my concern about the fluctuating input voltage from the car power source will not be a problem because firstly it's within the drivers input range and secondly the driver will only output the voltage as tested above?

That surely behaves as a buck-type, constant current LED driver. That is good

As to whether it will survive in the harsh environment of a car? Typically when designing circuits, you design for a particular worst case scenario, in a known type of environment. For small electronics, the auto/car represents a tough case, specially with regards to temperature, inductive voltage spikes, etc., so who knows how well that LED driver will work/last in a car. That I can't answer.

Will

 

[john2k]

That surely behaves as a buck-type, constant current LED driver. That is good

Thanks Will your advice and help is very much appreciated.

As to whether it will survive in the harsh environment of a car? Typically when designing circuits, you design for a particular worst case scenario, in a known type of environment. For small electronics, the auto/car represents a tough case, specially with regards to temperature, inductive voltage spikes, etc., so who knows how well that LED driver will work/last in a car. That I can't answer

What would happen if the driver lets say was to fail in the harsh environment of a car? is there any fail-safe options to prevent blowing the LED's if the driver fails? like cut out power completely etc.? I'm planning to use the power source from a light bulb circuit so would that not be a good stable source of current from the car?

 

[alpg88]

you make things look more complicated than they are, if driver blows, leds will still be fine, never seen a driver burn and take leds with them, it is theoreticly possible, but i have not seen that in real world.
flactuation of current\voltage isn't that big in a car, if it was every cars lights would noticably flicker most of the time.

 

[john2k]

you make things look more complicated than they are, if driver blows, leds will still be fine, never seen a driver burn and take leds with them, it is theoreticly possible, but i have not seen that in real world.
flactuation of current\voltage isn't that big in a car, if it was every cars lights would noticably flicker most of the time.

LOL thanks alpg88, thats a relief i guess i'm just trying to think of the worst case scenario LOL

 

[wquiles]

you make things look more complicated than they are, if driver blows, leds will still be fine, never seen a driver burn and take leds with them, it is theoreticly possible, but i have not seen that in real world.
flactuation of current\voltage isn't that big in a car, if it was every cars lights would noticably flicker most of the time.

+1

It "should" be OK, but one will never know until you try it out

The worst problem would be that the new circuit/LED/lamp assembly draws too little current and the "smart" computer in the car might sense this as being an open circuit - thinking the bulb blew off and needs replacement

 

[john2k]

thanks guys really appreciated

 

[john2k]

Managed to find the full specs for the 280mA driver i am using. the specs are as follows:

• Self adapting output voltage to suit for wide range Vf LED (Vf = 3.2V).
• Input voltage 12V AC / DC
• Constant 280mA output current
• Output short circuit protection
• Self over temperature protection
• High conversion efficient (up to 90%)
• IP65 Design for indoor or outdoor installations.

Maximum Ratings:

• Supply Voltage (Vin): 24VDC / 15VAC
• Max Operatiion Current: 1A
• Operation Temperature Range: -10℃ to 50℃
• Storage Temperature: -20℃ -

 

[wquiles]

Managed to find the full specs for the 280mA driver i am using. the specs are as follows:

• Self adapting output voltage to suit for wide range Vf LED (Vf = 3.2V).
• Input voltage 12V AC / DC
• Constant 280mA output current
• Output short circuit protection
• Self over temperature protection
• High conversion efficient (up to 90%)
• IP65 Design for indoor or outdoor installations.

Maximum Ratings:

• Supply Voltage (Vin): 24VDC / 15VAC
• Max Operatiion Current: 1A
• Operation Temperature Range: -10℃ to 50℃
• Storage Temperature: -20℃ -

Looks good, specially the short circuit protection and the operating temperature range

Where did you but the driver at?

Will

 

[john2k]

Thanks Will, i bought it from my local electronics store. I forgot to mention that this driver also says the following: Vout =< Vin -3V (Depending on Load). I'm assuming this is just basically saying that output voltage will always be lower than input voltage minus 3V. So if input voltage was for example 14V the max voltage the driver could output would be 11V. I guess this isnt a concern for me as my led unit will only use 3.4V anyway.

Thanks again for your help Will, very much appreciated.

 

[hank]

Suggestion: start by questioning your first assumption.

Sounds like you assume the voltage available in the vehicle tops out at 14-something.
Where are you getting that information?

Try searching, for example

https://duckduckgo.com/?q=automobile+voltage+spike

example of results:

Suppression of Transients in an Automotive Environment
www.industrologic.com/autotransients.pdf
"... Independent studies by the Society of Automotive Engineers. (SAE) have shown that voltage spikes from 25V to 125V can easily be generated [1], ..."

Voltage Spiking and V-LEDS « V-LEDS BLOG
v-ledsblog.com/2011/03/03/voltage-spiking-vleds/
Mar 3, 2011 – "Chances are they were the helpless victims of voltage spikes. These voltage spikes are found in every car or truck...."


I use LEDs in my dome and convenience lights. But I never, ever, have them switched on when starting or stopping the engine, to avoid transient voltage spikes.

(if those links are forbidden and disappear, look for "Society of Automotive Engineers" and "autotransients" along with the other search terms)

 

[john2k]

I use LEDs in my dome and convenience lights. But I never, ever, have them switched on when starting or stopping the engine, to avoid transient voltage spikes.

My LED driver wont be powered on instantly when car is powered on, the input power I am providing to LED driver is going to be a feed from the side lights of my car. The sidelights will turn on either when I turn them on or if the headlamps are in automatic mode in which case there will still be a delay.