Constant Current Driver

[wquiles]

Thanks Will,

I have another question. I've been using another branded driver for my brothers car for a few months which is very similar in specs. it's been working perfectly. One driver for each LED unit. But each LED driver has a 12v regulator circuit with some capactiors wired to the input. But recently one outputs dimmer than the other. And after about 40minutes or so the dim one shuts down. I am assuming it might be the regulator or even the driver that is going wrong. Just wondering if this is normal behavour of a failing driver?

The unit shutting down could be doing so when it activated its thermal shutdown function (assuming it has one). And the shutdown could have been in either the regulator or the driver itself. It is nearly impossible to diagnose remotely - too many unknowns.

But being in a car environment, I would expect that spikes/noise in the power to the module, and heat on the driver and/or the LED is what is causing the premature failure. Out of these heat would be the most likely problem.

Will

[john2k]

Thermal shutdown is what i was suspecting but then after cooling down for a long period of time a thermal shutdown should reset but this doesnt. This starts off slightly dimmer than the other driver and then 30-60mins later goes very dim that it seems like it is turned off.

[john2k]

On my brothers setup, difference is i have the oem driver also plugged in with diodes to make current flow 1 way. so it basically looks a bit like this x2

[john2k]

Forgot to mention Both drivers do not run at the same time but just incase they somehow did there is the diodes.

[john2k]

Quick Question: if my LED driver has a input rating that says input 12Vdc. And I used lets say a L7812CV on the input of the driver which has a typical voltage drop of 2v that means the input of the LED driver can drop lower than 11v. Does provide less voltage effect the driver? because although input says 12v i've managed to use the driver with a 9V low drop regulator and that works fine. Just wondering if a driver has a input rating of 12vdc if providing less voltage would effect it? i am assuming worst that can happen with less voltage is it won't work, but in this case it works with a 9V regulator as well.

[john2k]

I have purchased a new driver that has a 370mA CC output, when I measure current draw from the output of the driver to the LED's it shows correctly at 0.37A. But if i measure the current draw from the battery to the input of the driver. the current draw reads around 0.21A. Is this strange or normal? i tried with 2 other drivers made by different suppliers and they all seem to read around 100mA less on the current draw on the input that the output.

Can anyone help?

[charlie_r]

Is it a buck converter? According to Ohm's law and Watts law, when you drop voltage, current rises. So ahead of the converter you will see higher voltage at lower current, behind the converter you should see lower voltage and higher current. As explained elsewhere in this thread, any complex LED driver will adjust the output voltage to levels that will push just the right amount of current to the LEDs. You are essentially trading voltage for current.

[john2k]

Thanks Charlie,

My driver has a 12Vdc input and if I use a regular 12V regulator the dropout voltage on those are usually 2V. I am told that for the 12V regulator to work it will need minimum 14V but in most cases a car voltage can vary betwen 13V - 14.4V. So during the times when the car is outputting less volts does that mean the regulator wont work at all and just output whatever is being inputted? or will the regulator output whatever is being inputted -2V. So if for example car was giving 12V the regulator would give 10V? Or will it fail or start giving strange outputs that can damage the driver? I am thinking maybe I should use the 12V LM2940CT which is a low dropout 0.5V version. I need to try and keep the input voltage on my driver as close to 12V as possible.

I was considering using a even lower regulator because the driver works on lower voltages however with a 9V regulator the power loss in watts at highest would be 1.68W which would need heatsinking and because my enclosure is plastic it would be difficult to work around that. So I am trying to work with as less heat as possible. With a 12V regulator the highest power loss in watts will only be around 0.84W

Can anyone offer any advice or help as to what I can do? I've used a regular 12V regulator up until now and it has been fine and does not heat up too much. But problem is one of the driver is malfunctioning now and I am suspecting it has to do with the regulator.

[Mike S]

There's going to be a voltage drop through the regulator regardless of the input voltage. So if the input voltage drops to say 13V, then the voltage regulator might output something like 11 or 11.5V. Keep in mind that the dropout voltage on some regulators can change quite a bit as the load and/or the heat increases. It could be 1.3V with a 100 mA load and 2V with a 1500 mA load. I've also found in my experience that some regulators absolutely require filtering caps to output a constant voltage, while others do just fine without them.

If the LED's were bright at first and then dimmed and continue to stay dim even after a cooling period, it sounds like the driver was damaged. Built in thermal protection doesn't always save the chip. I would try it again with a 9V regulator, but this time monitor the chips temperature.

[john2k]

If the LED's were bright at first and then dimmed and continue to stay dim even after a cooling period, it sounds like the driver was damaged. Built in thermal protection doesn't always save the chip. I would try it again with a 9V regulator, but this time monitor the chips temperature.

Yes that is what has happened, even after days of cooling down the LED's with one driver come out much dimmer than they used to and gradually dim over 30-60minutes period until its very dim to notice that the led's are actually on. The other driver on the other side is still working fine.

Just wondering what could have caused the driver to damage. Rather than changing out the driver and then waiting months or weeks for it to damage and replace it, if i can prevent the premature failure or find out what is causing it to try and prevent it that will be great. The setup is pretty simple as pictured below, and the custom 300mA driver to provide less brightness is connected via a regular normal dropout 12V regulator with a few capacitors.

One possible cause I am thinking is that the regular regulator might have caused the driver to gradually fail. Because from what I have just learnt, a regular 12V driver has a voltage drop of 2V which means it needs 14V to operate. And on average when I check the car there is mainly 13.8V reading out. So what I am thinking is that maybe the regulator was not operating properly during the times when the voltage dropped below 14 and it let through all the voltage e.g. 13.8V to the driver. And because the driver is rated at 12Vdc, thats why maybe the driver has failed.

Other thing is that the driver details says that the led lamps should be wired in series, but mine are wired in parallel and the voltage does adjust accordingly depending on series or parallel. If i wire it in series, the voltage output adds up per LED. whereas on my parallel setup the voltage output from the driver stays at 3.2V which means it is adjusting the output correctly. So I dont know if that may be something?

Here is a quick sketch of how the 300mA driver is hooked up with the LED unit and the OEM 7W driver.

[Mike S]

You mentioned earlier (post #66) that the output of driver is 0.37 A at roughly 3.2V which is 1.184 watts. Were you using a 12V regulator at that time? If so, the input was 12V at 0.21A which is 2.52 watts. That would only be 46.98% efficient, so that poor little chip is probably producing some heat. If you weren't using a 12V regulator, it would be even worse than that.

Does the drivers PCB have a metal core or thermal vias right below the chip? Something that could help transfer and spread the heat?

If a 9V regulator were used there wouldn't be as much loss in the driver, but I still don't think it would survive. Maybe try an even lower value voltage regulator in a TO-220 package. They typically have good thermal properties and there are dozens of different heat sinks available for them. If you're stuck with the LED's wired in parallel, then sharing some of the losses with a voltage regulator might be the only option.

[john2k]

Does the drivers PCB have a metal core or thermal vias right below the chip? Something that could help transfer and spread the heat?

Nope. The driver is housed inside a small plastic box and is a sealed potted box which has 4 wires coming out of it.

If a 9V regulator were used there wouldn't be as much loss in the driver, but I still don't think it would survive. Maybe try an even lower value voltage regulator in a TO-220 package. They typically have good thermal properties and there are dozens of different heat sinks available for them. If you're stuck with the LED's wired in parallel, then sharing some of the losses with a voltage regulator might be the only option.

When I test it with a 9V regulator, the regulator gets rather hot to touch and if i touch the plastic housing of the driver its also a little warm when I touch it. But as far as I remember when I used a 12V regulator the regulator was much less hot and the driver box hardly produced any heat. But I will test this again later today.

So do you think heat is what has caused the driver to fail? are you suggesting that the lower the input voltage I provide using a regulator to the driver the better it would be for the driver? but the lower the input voltage the higher the heat would be. For example if i was to drop the input voltage to 5V. that would mean the regulator needs to drop from a max of 14.4V all the way to 5V. the regulator will need to bring down the voltage a huge amount and I am pretty sure that would generate a huge amount of heat and not sure how long the regulator itself will last that way. Also, I need to house the components inside a box so heatsinking becomes an issue aswell because most boxes are plastic.

instead of using 2 seperate drivers to drive the two seperate LED units that are wired in parallel. will powering both the LED units with the same driver help at all and be better for the driver? so for example if I split the output from the driver and then connected it to both the units. (i guess in parallel)

also, come to think of it, the driver is designed for a 12Vdc input. and the output voltage adjusts depending.. So as long as the output voltage is within the range that the driver is rated at (which is 3-10Vdc) then shouldnt the driver be just fine? because the driver says its input is 12Vdc and output is 3-10Vdc and it also says 1-2W.

[SemiMan]

You WILL NEVER EVER read 13.8V on a battery that is not charging. NEVER. You will never ever read 13.5 either. You may read 13.0-13.1. I don't care how new the battery is.

Your supposed OEM driver is Phileds, not Phillips and your "new" driver is obviously the same brand ... easy to tell from the wording you posted:

http://www.phileds.com/upload/iblock...957801a15f.pdf

You claim you get 280mA, +/-2. That is within 0.7%. These cheapy units are highly unlikely to be that accurate and nor is your meter.

You have not posted a picture of your actual units, test set up .. though asked.

You did not post the name of your driver .. which would have made sense.


Everything in this post screams BS ...... Initially I wanted to help, now I just think you are a troll .......

[Mike S]

So do you think heat is what has caused the driver to fail? are you suggesting that the lower the input voltage I provide using a regulator to the driver the better it would be for the driver? but the lower the input voltage the higher the heat would be. For example if i was to drop the input voltage to 5V. that would mean the regulator needs to drop from a max of 14.4V all the way to 5V. the regulator will need to bring down the voltage a huge amount and I am pretty sure that would generate a huge amount of heat and not sure how long the regulator itself will last that way. Also, I need to house the components inside a box so heatsinking becomes an issue aswell because most boxes are plastic.

I can't be certain, but I suspect that's what it is. Either that, or the driver was faulty before you ever applied voltage to it. A well designed switch mode driver should have no problem dropping 9 volts with a 370 mA output. Even the standard SOIC-8 with its high thermal resistance can handle that without breaking a sweat.

Actually, I'm not even sure about the voltage regulators. I went back and read this thread again and there's definitely something wrong. Whether it be the LED drivers or the vehicles electrical system, it's hard to tell. A 5 volt model might be too low for the driver to function, or at least function properly. That would only be about 1V overhead considering the output diodes are dropping ~0.5 volt.

instead of using 2 seperate drivers to drive the two seperate LED units that are wired in parallel. will powering both the LED units with the same driver help at all and be better for the driver? so for example if I split the output from the driver and then connected it to both the units. (i guess in parallel)

I may have missed it, but are the two drivers in the diagram ever on at the same time? I've never tried that with switching drivers before, but will do a little experimenting when I get home.

Using a single output to power all those LED's in parallel isn't ideal, but I suppose it would work. Try it out and see how well the LED's share the current. Can you afford to dim the LED's that far?

also, come to think of it, the driver is designed for a 12Vdc input. and the output voltage adjusts depending.. So as long as the output voltage is within the range that the driver is rated at (which is 3-10Vdc) then shouldnt the driver be just fine? because the driver says its input is 12Vdc and output is 3-10Vdc and it also says 1-2W.

Yes, it should be fine, but it doesn't sound that it's working correctly in this case.

[john2k]

I may have missed it, but are the two drivers in the diagram ever on at the same time? I've never tried that with switching drivers before, but will do a little experimenting when I get home.

They are never on at the same time, the second one powers up when the first one shuts down. But I put the diodes in there just incase something goes wrong.

Using a single output to power all those LED's in parallel isn't ideal, but I suppose it would work. Try it out and see how well the LED's share the current. Can you afford to dim the LED's that far?

Depends how much they dim, i'll try and experiment and see how dim they get. If i had 2 sealed units with LEDs wired in parallel and then I wire them to the same output of the driver, will the driver treat all the LED's across both units as one big parallel setup?


Thanks